From bussonniermatthias at gmail.com Sat Nov 1 06:12:15 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Sat, 1 Nov 2014 11:12:15 +0100 Subject: [IPython-dev] Notebook format "incompatible" changes Message-ID: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> Hello Jovyans and other beings from Jupyter galilean moons system, I'm writing to you to warn that on Monday (probably morning) a relatively important change in IPython will land on master. In particular it does change the notebook structure quite a bit. Cf https://github.com/ipython/ipython/pull/6045 for more information. As nothing is ever perfect, please do back-up your notebooks before upgrading. With this new update, saved notebook of this new of IPython version won't be compatible with older IPython anymore, nor will nbviewer be able to render them. The new notebook format will be back-ported on older (2.x) version of IPython and the new notebook format will be supported on nbviewer, we will just need a few days to port all the changes. There is of course be a way to manually downgrade the notebook from v4 to v3 using nbconvert. Eventually the more adventurous can test the branch this week-end. As usual, be prepared for data loss and bugs, so update your git branches, refresh submodules, clear your browser caches, roll up your sleeves, and send bug reports. Cheers, -- Matthias -------------- next part -------------- An HTML attachment was scrubbed... URL: From ellisonbg at gmail.com Sat Nov 1 14:01:55 2014 From: ellisonbg at gmail.com (Brian Granger) Date: Sat, 1 Nov 2014 11:01:55 -0700 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> Message-ID: Matthias, Thanks for posting here about these changes! And thanks to the whole team (esp Min) for working on all of this! Cheers, Brian On Sat, Nov 1, 2014 at 3:12 AM, Matthias Bussonnier wrote: > Hello Jovyans and other beings from Jupyter galilean moons system, > > > I'm writing to you to warn that on Monday (probably morning) a relatively > important change > in IPython will land on master. In particular it does change the notebook > structure quite a bit. > > Cf https://github.com/ipython/ipython/pull/6045 for more information. > > As nothing is ever perfect, please do back-up your notebooks before > upgrading. With this new update, > saved notebook of this new of IPython version won't be compatible with > older IPython anymore, nor will nbviewer be able to render them. > > The new notebook format will be back-ported on older (2.x) version of > IPython and the new notebook format > will be supported on nbviewer, we will just need a few days to port all the > changes. > > There is of course be a way to manually downgrade the notebook from v4 to v3 > using nbconvert. > > Eventually the more adventurous can test the branch this week-end. > > As usual, be prepared for data loss and bugs, so update your git branches, > refresh submodules, clear your browser caches, roll up your sleeves, and > send bug reports. > > Cheers, > -- > Matthias > > > > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -- Brian E. Granger Cal Poly State University, San Luis Obispo @ellisonbg on Twitter and GitHub bgranger at calpoly.edu and ellisonbg at gmail.com From ccordoba12 at gmail.com Sun Nov 2 12:49:14 2014 From: ccordoba12 at gmail.com (=?UTF-8?B?Q2FybG9zIEPDs3Jkb2Jh?=) Date: Sun, 02 Nov 2014 12:49:14 -0500 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> Message-ID: <54566E9A.2040002@gmail.com> Hi, Could you explain a bit more what this change is about? Cheers, Carlos El 01/11/14 a las 13:01, Brian Granger escribi?: > Matthias, > > Thanks for posting here about these changes! > > And thanks to the whole team (esp Min) for working on all of this! > > Cheers, > > Brian > > On Sat, Nov 1, 2014 at 3:12 AM, Matthias Bussonnier > wrote: >> Hello Jovyans and other beings from Jupyter galilean moons system, >> >> >> I'm writing to you to warn that on Monday (probably morning) a relatively >> important change >> in IPython will land on master. In particular it does change the notebook >> structure quite a bit. >> >> Cf https://github.com/ipython/ipython/pull/6045 for more information. >> >> As nothing is ever perfect, please do back-up your notebooks before >> upgrading. With this new update, >> saved notebook of this new of IPython version won't be compatible with >> older IPython anymore, nor will nbviewer be able to render them. >> >> The new notebook format will be back-ported on older (2.x) version of >> IPython and the new notebook format >> will be supported on nbviewer, we will just need a few days to port all the >> changes. >> >> There is of course be a way to manually downgrade the notebook from v4 to v3 >> using nbconvert. >> >> Eventually the more adventurous can test the branch this week-end. >> >> As usual, be prepared for data loss and bugs, so update your git branches, >> refresh submodules, clear your browser caches, roll up your sleeves, and >> send bug reports. >> >> Cheers, >> -- >> Matthias >> >> >> >> >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From bussonniermatthias at gmail.com Sun Nov 2 13:37:12 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Sun, 2 Nov 2014 19:37:12 +0100 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: <54566E9A.2040002@gmail.com> References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> <54566E9A.2040002@gmail.com> Message-ID: <47C409E3-C4D0-4C3E-B7CA-44D871902B16@gmail.com> Hi Carlos, TL:DR; at the end. With the notebook format v4, the way notebooks are written on disk will change. After now some time with v3, we know that a few things could have been done better or in a more general way, thus we improved the notebook format. Once 3.0 and 2.4 are released this shouldn't change much for the end user. I'll go over some of the changes in the notebook format and their reasons, but you should read the IPep for all details. Get rid of python specific names. pyin, pyout, pyerr are renamed input input, output, and error (roughly, read IPep for exact names). Uniformise jpg/png/text keys in output to be a mimetype. You can now store application/x-pdf or doc/microsoft-word in a notebook. If the frontend now how to read it, it will do something with it i suppose. It remove a lot of special casing. Get rid of worksheets, almost nobody use it, and there are other smarter way to store them if we wanted to. Text cells had their text under the 'text' key, and code cell under 'source'. Which is a bit silly, as it forces you to do some special casing where not necessary. Heading cell are now gone (from the file format), Technology evolve fast, and we can now detect the #*n in markdown cell and convert properly to LaTeX, and add anchor in nbviewer. More importantly, we know have a jsonschema that describe the format. So we can validate the notebook, and know that for example, a prompt number is either null, or a number. v3 was implicitly allowing '*' (star), which would happened if you were saving while cell are running, which took by surprise both us and external library where the conversion of some notebook made crashes. This include also security risk. Which I won't develop, cause I won't develop, but those who know know I love javascipt injection, and you can do nasty things on known websites. By insuring the type of each field of the notebook you lower the attack area, and protect FooBar corp user from attacks even if FooBar corp does not really respond to you when you do responsible disclosure and finally give up. I won't describe "All the things", but you see the big picture. It's better, faster, stronger. TL;DR: Wat ? You don't know ?! I've heard that IPython notebook format v3 might have been responsible for the death of a huge number of kittens due to developers banging their head on their desk. It is also probable that the need of extra computing power to run more test because of it's complexity is in part responsible for global warming. Also the new format prevent the use of %pylab that kills the endangered species of newbies. The naming convention was also really strange for php coders that don't understand the py prefixes. As we like kittens, newbies, developers, php coders (but not php itself [1]), and dislike global warming we decided to fix that. It also comes as ipynb v4+ that have a 5.5 inches screen, and we plan to jump from v8 to v10 directly. We also removed touch-id from metadata so that police cannot force youth unlock your ipynb. Hope that shine a light on a few of the reason and what will change. Tell us if you have any more questions. Cheers, -- M [1]: But MinRK Love Javascript. Le 2 nov. 2014 ? 18:49, Carlos C?rdoba a ?crit : > Hi, > > Could you explain a bit more what this change is about? > > Cheers, > Carlos > > El 01/11/14 a las 13:01, Brian Granger escribi?: >> Matthias, >> >> Thanks for posting here about these changes! >> >> And thanks to the whole team (esp Min) for working on all of this! >> >> Cheers, >> >> Brian >> >> On Sat, Nov 1, 2014 at 3:12 AM, Matthias Bussonnier >> wrote: >>> Hello Jovyans and other beings from Jupyter galilean moons system, >>> >>> >>> I'm writing to you to warn that on Monday (probably morning) a relatively >>> important change >>> in IPython will land on master. In particular it does change the notebook >>> structure quite a bit. >>> >>> Cf https://github.com/ipython/ipython/pull/6045 for more information. >>> >>> As nothing is ever perfect, please do back-up your notebooks before >>> upgrading. With this new update, >>> saved notebook of this new of IPython version won't be compatible with >>> older IPython anymore, nor will nbviewer be able to render them. >>> >>> The new notebook format will be back-ported on older (2.x) version of >>> IPython and the new notebook format >>> will be supported on nbviewer, we will just need a few days to port all the >>> changes. >>> >>> There is of course be a way to manually downgrade the notebook from v4 to v3 >>> using nbconvert. >>> >>> Eventually the more adventurous can test the branch this week-end. >>> >>> As usual, be prepared for data loss and bugs, so update your git branches, >>> refresh submodules, clear your browser caches, roll up your sleeves, and >>> send bug reports. >>> >>> Cheers, >>> -- >>> Matthias -------------- next part -------------- An HTML attachment was scrubbed... URL: From fperez.net at gmail.com Sun Nov 2 18:23:26 2014 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 2 Nov 2014 15:23:26 -0800 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> Message-ID: On Sat, Nov 1, 2014 at 11:01 AM, Brian Granger wrote: > And thanks to the whole team (esp Min) for working on all of this! > +lots. This has been a huge amount of slow, careful, not-very-sexy work. I am really thankful for the patience the whole team has had working on this, to help us set the notebook format as a solid foundation for long-term sharing and archival of computational work. This kind of effort provides few immediate rewards, but can have very significant long-term value. Thanks a lot for everyone who pitched in on that PR... Cheers, f -- Fernando Perez (@fperez_org; http://fperez.org) fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) fernando.perez-at-berkeley: contact me here for any direct mail -------------- next part -------------- An HTML attachment was scrubbed... URL: From DavidAnthonyPowell+python at gmail.com Sun Nov 2 19:55:11 2014 From: DavidAnthonyPowell+python at gmail.com (David Powell) Date: Mon, 3 Nov 2014 11:55:11 +1100 Subject: [IPython-dev] Conflict between IPython notebook javascript and three.js? In-Reply-To: References: Message-ID: Thankyou Patrick, I see that you have managed to get the trackball controls working correctly in the notebook, so obviously it is possible. I already found that HTML output instead of javascript is necessary, because the "element" DOM object does not seem to be present after nbconverting to html. When I get time to check carefully through your code, I'm sure I'll get enough hints to solve my problem. regards David On 31 October 2014 23:05, Patrick Fuller wrote: > In case it helps, here are two projects that work the way you want with > trackball: imolecule and igraph. The code of interest is here. It?s been a > while since I?ve looked at this, but I believe my trick was to use > _repr_html_ instead of _repr_javascript_ in order to make a new div. > > Hope this helps, > Pat > > > On Fri, Oct 31, 2014 at 2:19 AM, David Powell > wrote: >> >> I am trying to display 3D objects in IPython notebooks, and so far >> three.js looks like the most promising approach. In particular, it is >> important that the 3D plot is still visible after converting the >> notebook to html. >> >> So far I have succeeded in getting a basic example working, by >> adapting >> http://nbviewer.ipython.org/github/payne92/notebooks/blob/master/00%20Javascript%20In%20Notebooks.ipynb >> >> However, I run into problems when I include the trackball controls >> from three.js, and restrict the controls to input only on the cavas >> DOM element (so that clicking outside the canvas does not modify the >> perspective). The trackball controls do not work correctly, as can be >> seen by running the attached notebook example. >> >> When I export the notebook to html, I see that the wegbl output is >> present as I wanted. Furthermore, it seems that in the html, the >> trackball controls work perfectly. This suggests that the problem is >> some kind of conflict with the notebook's javascript. Any hints as to >> the possible cause of this problem would be greatly appreciated. >> >> In case it matters, I am using IPython 2.3 >> >> regards >> David >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > From DavidAnthonyPowell+python at gmail.com Sun Nov 2 20:02:06 2014 From: DavidAnthonyPowell+python at gmail.com (David Powell) Date: Mon, 3 Nov 2014 12:02:06 +1100 Subject: [IPython-dev] Conflict between IPython notebook javascript and three.js? In-Reply-To: <54538A46.2040509@creativetrax.com> References: <54538A46.2040509@creativetrax.com> Message-ID: > This may or may not be an issue on top of this, but I've had problems > with TrackballControls and OrbitControls (which I like better) accepting > key events in the global page. See > https://github.com/mrdoob/three.js/pull/5236. I avoided using OrbitControls for the same reason, but I didn't have any problems with TrackballControls. >> >> When I export the notebook to html, I see that the wegbl output is >> present as I wanted. Furthermore, it seems that in the html, the >> trackball controls work perfectly. This suggests that the problem is >> some kind of conflict with the notebook's javascript. Any hints as to >> the possible cause of this problem would be greatly appreciated. > > Do you see any errors in the javascript console? No, I managed to get rid of all errors, and the problem still persists. But I think Patrick's working example should give me enough guidance to fix my problem. > By the way, you may be interested in the pythreejs library, which is a > wrapping of three.js as IPython widgets: > > https://github.com/jasongrout/pythreejs/tree/ipython > > (the ongoing IPython notebook work is in the ipython branch) > > See the examples folder for a few simple (and possibly out-of-date) > examples. It's still a work in progress, but it does work quite nicely > already. For example, see this example: > > http://nbviewer.ipython.org/github/jasongrout/pythreejs/blob/ipython/examples/Surface.ipynb > Great, this looks like it will be the preferred solution to my problem in the long term. I tried running the examples, but I ran into the following javascript exception, and didn't get any output: Uncaught TypeError: Cannot read property 'obj' of undefined pythreejs.js?_=1414974265086:329 > One disadvantage of using widgets, which may be a show-stopper for you > right now, is that you can't see the 3d plot in nbviewer and with > nbconvert. Jon's recent work on widget persistence should fix that, > hopefully soon. We're reviewing his pull request now. Yes I'm aware of this problem, and I'm glad to hear that a fix is on the way. Is this expected to be incorporated into ipython v3.0? thanks, David From claresloggett at gmail.com Mon Nov 3 02:06:07 2014 From: claresloggett at gmail.com (Clare Sloggett) Date: Mon, 3 Nov 2014 18:06:07 +1100 Subject: [IPython-dev] Error running JupyterHub In-Reply-To: References: Message-ID: Hi Doug, I'm happy to report progress :) I can get jupyterhub to run. I *think* that possibly our issue was another application running on the internal (hub) port. If I start from scratch (with updated jupyterhub) and do something like su root jupyterhub --port 9520 --JupyterHubApp.hub_port=8500 it works as expected. My next step is to try to get a redirect working so I can use http://url/jupyter instead of http://url:9520/, but I'll go back to the thread where we were discussing that already. Thanks for your help! Clare On 28 October 2014 21:54, Doug Blank wrote: > On Tue, Oct 28, 2014 at 3:08 AM, Clare Sloggett > wrote: > >> Hi Doug, >> >> > I can confirm that attempting to run the configurable-http-proxy twice > gives the error you describe. There are probably other reasons to get that > error too (eg, maybe something is already listening on that port, > permission error, etc). I would try the items related to node, listed at > the bottom here first: > > >> 1. I'm not installing ipython from source, just with pip, so a bit >> confused by the instructions re ipython-master. Maybe I'm misunderstanding. >> I've run `pip install ipython --upgrade`; do I really need to do ipython >> from source? >> > > I'm not 100% sure if it is necessary. Installing from master is easy, if > that is a concern: > > git clone https://github.com/ipython/ipython.git > cd ipython > python setup.py submodules > pip install . -U > > To update after initial install, just do the last two steps. > > >> 2. I've just tried the jupyterhub README instructions from scratch on a >> new machine (a new cloud instance) from the current jupyterhub master, >> commit db5cf9cf99f68f133b... - same issues unfortunately. >> > > Hmmm.... the error log shows that the error is launching the > http-configurable-proxy... > >> >> 3. I'm not sure how to check for this but it's probably addressed by >> having launched a new machine! >> > > There are two programs that are run by "root": configurable-http-proxy and > jupyterhub, and then for each user there is a "jupyterhub-singleuser" that > is running. You can use something like "ps aux | grep > configurable-http-proxy" or "ps aux | grep jupyterhub" to see the > associated processes. You can use "kill" to kill those, if necessary. But > it sounds like it hasn't made it that far. > > >> >> 4. I'm running as root, no sudo involved. Just >> $ su root >> $ jupyterhub >> > > If this computer is on the internet, then eventually you'll probably want > to move to the more involved "sudo" method for better security. But this > should be fine for testing and getting it working. > > >> And I get all the same errors as previously. >> >> I'm mystified as I can see others posting here saying that they are using >> it, so it must be a problem in my setup, but I really can't see what. Could >> it possibly be a node.js issue? >> > > You should be able to run this line, from your log: > > configurable-http-proxy --ip '' --port 8000 --api-ip localhost --api-port > 8001 --default-target http://localhost:8081 > > Does that work without errors? Control+c to exit, if it does work. I can > confirm that if you try to run this with one already running, you get > exactly the error you describe. > > On the page: > > > https://github.com/jupyter/jupyterhub/wiki/Using-sudo-to-run-the-server-as-non-root > > there is also a note needed for serving on restricted ports: > > sudo setcap 'cap_net_bind_service=+ep' /usr/bin/node > > but read that doc for warnings. You may need that too, even when running > as root. Your issue does appear to be a problem starting the http server > via node. Hope that helps! > > -Doug > > > > >> Thanks again for the help! >> >> Clare >> >> >> On 24 October 2014 23:58, Doug Blank wrote: >> >>> On Fri, Oct 24, 2014 at 8:49 AM, Clare Sloggett >> > wrote: >>> >>>> Hi all, >>>> >>>> I'm getting an error running JupyterHub. Right now all I'm doing is >>>> attempting to run it as root with no command-line options. >>>> >>>> A couple of weeks ago I was trying to set it up and while I was running >>>> into issues, I wasn't getting this particular problem - I was able to at >>>> least run the command. I'm not sure if something has changed in Jupyter. >>>> Entirely possible that I made a mistake in my setup this time! >>>> >>>> Any advice would be appreciated in interpreting the below errors. >>>> >>>> Thanks, >>>> Clare >>>> >>>> Below is the output of `jupyterhub --debug`: >>>> >>>> Traceback (most recent call last): >>>> File "/usr/lib/python2.7/logging/__init__.py", line 846, in emit >>>> msg = self.format(record) >>>> File "/usr/lib/python2.7/logging/__init__.py", line 723, in format >>>> return fmt.format(record) >>>> File >>>> "/usr/local/lib/python2.7/dist-packages/IPython/config/application.py", >>>> line 121, in format >>>> return super(LevelFormatter, self).format(record) >>>> File "/usr/lib/python2.7/logging/__init__.py", line 467, in format >>>> s = self._fmt % record.__dict__ >>>> KeyError: u'color' >>>> Logged from file application.py, line 247 >>>> Traceback (most recent call last): >>>> File "/usr/lib/python2.7/logging/__init__.py", line 846, in emit >>>> msg = self.format(record) >>>> File "/usr/lib/python2.7/logging/__init__.py", line 723, in format >>>> return fmt.format(record) >>>> File >>>> "/usr/local/lib/python2.7/dist-packages/IPython/config/application.py", >>>> line 121, in format >>>> return super(LevelFormatter, self).format(record) >>>> File "/usr/lib/python2.7/logging/__init__.py", line 467, in format >>>> s = self._fmt % record.__dict__ >>>> KeyError: u'color' >>>> Logged from file application.py, line 248 >>>> [D 23:39:03.030 JupyterHubApp] Connecting to db: >>>> sqlite:///jupyterhub.sqlite >>>> [I 23:39:03.062 JupyterHubApp] Not using whitelist. Any authenticated >>>> user will be allowed. >>>> [D 23:39:03.064 JupyterHubApp] Loaded users: >>>> ubuntu admin >>>> [I 23:39:03.072 JupyterHubApp] Starting proxy: >>>> [u'configurable-http-proxy', '--ip', u'', '--port', '8000', '--api-ip', >>>> u'localhost', '--api-port', '8001', '--default-target', ' >>>> http://localhost:8081'] >>>> 23:39:03.178 - info: [ConfigProxy] Proxying http://*:8000 to >>>> http://localhost:8081 >>>> 23:39:03.180 - info: [ConfigProxy] Proxy API at >>>> http://localhost:8001/api/routes >>>> >>>> events.js:72 >>>> throw er; // Unhandled 'error' event >>>> >>> >>> I think that I have had that issue when configurable-http-proxy isn't >>> running, or maybe it is already running and you are trying to start it >>> again. >>> >>> Some things that I do when updating jupyterhub: >>> >>> 1. Updating ipython: make sure you run "python setup.py submodule" in >>> ipython-master to get any new dependencies. Then "git pull" and "sudo pip >>> install . -U" >>> >>> 2. Updating jupyterhub: do the steps in the jupyterhub/README.md >>> again... some dependencies may have changed. Then "sudo pip install . -U" >>> >>> 3. Make sure all jupyterhub-singleuser processes and >>> configurable-http-proxy processes are killed. This is still hard to do >>> sometimes. >>> >>> 4. Are you running with a sudo environment? There is an additional line >>> in the sudoers file that is now needed to kill processes. >>> >>> I think that is most of what I need to do. >>> >>> -Doug >>> >>> >>>> ^ >>>> Error: listen EADDRINUSE >>>> at errnoException (net.js:904:11) >>>> at Server._listen2 (net.js:1042:14) >>>> at listen (net.js:1064:10) >>>> at net.js:1146:9 >>>> at dns.js:72:18 >>>> at process._tickCallback (node.js:419:13) >>>> at Function.Module.runMain (module.js:499:11) >>>> at startup (node.js:119:16) >>>> at node.js:906:3 >>>> [C 23:39:04.087 JupyterHubApp] Failed to start proxy >>>> Traceback (most recent call last): >>>> File "/usr/local/lib/python2.7/dist-packages/jupyterhub/app.py", >>>> line 727, in start >>>> IOLoop().run_sync(self.start_proxy) >>>> File "/usr/local/lib/python2.7/dist-packages/tornado/ioloop.py", >>>> line 389, in run_sync >>>> return future_cell[0].result() >>>> File >>>> "/usr/local/lib/python2.7/dist-packages/tornado/concurrent.py", line 129, >>>> in result >>>> raise_exc_info(self.__exc_info) >>>> File >>>> "/usr/local/lib/python2.7/dist-packages/tornado/stack_context.py", line >>>> 302, in wrapped >>>> ret = fn(*args, **kwargs) >>>> File "/usr/local/lib/python2.7/dist-packages/tornado/gen.py", >>>> line 574, in inner >>>> self.set_result(key, result) >>>> File "/usr/local/lib/python2.7/dist-packages/tornado/gen.py", >>>> line 500, in set_result >>>> self.run() >>>> File "/usr/local/lib/python2.7/dist-packages/tornado/gen.py", >>>> line 529, in run >>>> yielded = self.gen.throw(*exc_info) >>>> File "/usr/local/lib/python2.7/dist-packages/jupyterhub/app.py", >>>> line 556, in start_proxy >>>> _check() >>>> File "/usr/local/lib/python2.7/dist-packages/jupyterhub/app.py", >>>> line 552, in _check >>>> raise e >>>> RuntimeError: Proxy failed to start with exit code 8 >>>> >>>> _______________________________________________ >>>> IPython-dev mailing list >>>> IPython-dev at scipy.org >>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>> >>>> >>> >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>> >>> >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From claresloggett at gmail.com Mon Nov 3 02:13:11 2014 From: claresloggett at gmail.com (Clare Sloggett) Date: Mon, 3 Nov 2014 18:13:11 +1100 Subject: [IPython-dev] Jupyterhub behind NGINX redirect In-Reply-To: References: Message-ID: Hi all, We've managed to troubleshoot our other issues, so I'm now sure that I can run jupyterhub successfully *without* a redirect. So I'm back to trying to get a redirect working. I've cc'd Nuwan who's been working on this with me and probably understands the issues better than I do. Currently, if I run something like $ jupyterhub --port 9520 ----JupyterHubApp.hub_port=8500 then I can access JupyterHub at http://my-url:9520/ . So far, success! This works fine, I can log in as various linux users, edit notebooks, etc. What we'd like is for it to instead be available to users at http://my-url/jupyter/ . We assumed that the command-line option JupyterHubApp.base_url is supposed to be used for this. We are trying to set up an NGINX redirect from /jupyter/ to :9520 and run jupyterhub so that it behaves properly in this situation. However JupyterHubApp.base_url doesn't seem to be doing what we expect it to. Is anyone able to explain how this parameter is supposed to be used? Cheers, Clare On 28 October 2014 18:15, Clare Sloggett wrote: > Hi Min, > > Thanks for your help! > > Actually at the moment I have been pushed back to a more fundamental > problem and can't even reproduce the error I was posting about in this > email. I can't get jupyterhub to run at all. I posted this separate error > in another thread, which I've just replied to a moment ago: "Error running > JupyterHub". > > I suspect the issue is mine as I encountered it a few days ago and am > still encountering it after updating to the latest commit. It seems > unlikely a bug has survived through several commits without someone else > discovering it too. But I'm having trouble working out what the cause is, > and am not sure what to make of the error messages it's throwing (both > python *and* javascript errors are thrown when it crashes). Full error > printout is in that thread! > > Cheers, > Clare > > On 26 October 2014 06:34, MinRK wrote: > >> Clare, >> >> Can you update to the latest master? I think the never-ending redirects >> could be the result of a recently fixed typo. >> >> Thanks, >> -MinRK >> >> On Thu, Oct 23, 2014 at 10:01 PM, Clare Sloggett > > wrote: >> >>> Hi Doug, >>> >>> Thanks for this. I had actually just been thinking about about the NGINX >>> redirect issue, and had assumed everything behind that would be fairly >>> straightforward. But it sounds like you are saying there may be more >>> fundamental issues, and to be honest I haven't tested that the redirect is >>> definitely the source of all my problems. >>> >>> It sounds like I need to do some more direct testing and come back! >>> >>> In the meantime, if anyone has insight into what could be causing a URL >>> rewrite like "/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython...." >>> that could be really helpful. >>> >>> Cheers, >>> Clare >>> >>> On 13 October 2014 00:44, Doug Blank wrote: >>> >>>> On Sun, Oct 12, 2014 at 1:35 AM, Clare Sloggett < >>>> claresloggett at gmail.com> wrote: >>>> > >>>> > Hi all, >>>> > >>>> > Am I right in thinking this list is also the right place for >>>> questions about Jupyterhub? >>>> > >>>> > I'm trying to set up Jupyterhub for multiple users, on the same >>>> server where we are running several other services. Currently there are >>>> NGINX redirects in place to these other services. I'd like to set up http:///ipython/ >>>> to redirect to a port on localhost and run Jupyterhub over that port. Is >>>> this possible? >>>> > >>>> > I've got this working in the past with single-user IPython Notebook. >>>> For that, I set config options c.NotebookApp.base_project_url, >>>> c.NotebookApp.base_kernel_url, and c.NotebookApp.webapp_settings to be >>>> aware of the ipython/ URL prefix. We used an NGINX redirect to forward >>>> requests and handle websockets properly, which looked like >>>> > >>>> > location /ipython/ { >>>> > proxy_pass http://127.0.0.1:9510; >>>> > proxy_set_header Host $host; >>>> > proxy_set_header X-Real-IP $remote_addr; >>>> > proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; >>>> > proxy_set_header Upgrade $http_upgrade; >>>> > proxy_set_header Connection "upgrade"; >>>> > } >>>> > >>>> > This worked for the old single-user notebook. But, I'm not clear on >>>> the model Jupyterhub is using (and I'm not much of a sysadmin). I read >>>> through the command-line options, thought from them that I don't need to >>>> strip the /ipython/ from the incoming requests, and have tried using a >>>> similar redirect to the above and setting >>>> --JupyterHubApp.base_url='ipython/' and also setting --port 9510. This does >>>> seem to see the incoming requests but results in amusing requests like >>>> > >>>> > 500 GET >>>> /ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/...... >>>> > >>>> > ... so clearly I haven't understood what's going on. Does anyone have >>>> any pointers on how this should work? I haven't read any docs other than >>>> the README and the command-line parameter information, so apologies if >>>> there's something obvious that I didn't look at. >>>> > >>>> > I'm also interested in running Jupyterhub as a daemon, so if that's >>>> something that's been done before and there's anything I should know, that >>>> would be great. >>>> >>>> There is one other document, other than the README.md and the reported >>>> issues [1]: >>>> >>>> >>>> https://github.com/jupyter/jupyterhub/wiki/Using-sudo-to-run-the-server-as-non-root >>>> >>>> Despite the warning at the top of that page, you can make jupyterhub >>>> work on some systems (eg, Linux, such as Ubuntu) using sudo, but not as >>>> root. I suspect that this would be the recommended setup when jupyterhub is >>>> complete. One recent change is the ability to save/load state from a >>>> database. The wiki page above hasn't been updated with the note from this >>>> issue: >>>> >>>> https://github.com/jupyter/jupyterhub/issues/57 >>>> >>>> It might be easier to start without NGINX, and then add it after you >>>> have a working jupyterhub system. It would be nice to have a little bash >>>> script to make this a "service" that would support "start", "top", and >>>> "status"... but I haven't had time. Currently, we're just becoming the >>>> non-root sudoer ("rhea" in the docs) and starting the server, something >>>> along the lines: >>>> >>>> jupyterhub --LocalProcessSpawner.set_user=sudo >>>> --JupyterHubApp.ip=165.106.10.83 --JupyterHubApp.port=80 >>>> --db='sqlite:///:memory:' &>> /var/log/jupyterhub/log & >>>> >>>> The next step for us is to get it running under https... looks like >>>> others have blazed that trail, so it looks possible. >>>> >>>> If you have success, it would be great to add to the wiki docs... I >>>> suspect that many of us that aren't sys admins will be wanting to get this >>>> up and running. >>>> >>>> -Doug >>>> >>>> [1] - https://github.com/jupyter/jupyterhub >>>> >>>> > Any help much appreciated! >>>> > >>>> > Clare >>>> > >>>> > >>>> > _______________________________________________ >>>> > IPython-dev mailing list >>>> > IPython-dev at scipy.org >>>> > http://mail.scipy.org/mailman/listinfo/ipython-dev >>>> > >>>> >>>> _______________________________________________ >>>> IPython-dev mailing list >>>> IPython-dev at scipy.org >>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>> >>>> >>> >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>> >>> >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > -------------- next part -------------- An HTML attachment was scrubbed... URL: From wes.turner at gmail.com Mon Nov 3 03:32:59 2014 From: wes.turner at gmail.com (Wes Turner) Date: Mon, 3 Nov 2014 02:32:59 -0600 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> Message-ID: Thanks! 1. Is there a link to the jsonschema? In the documentation? 2. How feasible would it be to write a JSON-LD context? 3. Where/how can/could Dublin Core metadata be added? It would be great to be able to index these documents with a title and authors. https://en.wikipedia.org/wiki/Dublin_Core#DCMI_Metadata_Terms On Sun, Nov 2, 2014 at 5:23 PM, Fernando Perez wrote: > > On Sat, Nov 1, 2014 at 11:01 AM, Brian Granger > wrote: > >> And thanks to the whole team (esp Min) for working on all of this! >> > > +lots. This has been a huge amount of slow, careful, not-very-sexy work. > I am really thankful for the patience the whole team has had working on > this, to help us set the notebook format as a solid foundation for > long-term sharing and archival of computational work. > > This kind of effort provides few immediate rewards, but can have very > significant long-term value. Thanks a lot for everyone who pitched in on > that PR... > > Cheers, > > f > > > -- > Fernando Perez (@fperez_org; http://fperez.org) > fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) > fernando.perez-at-berkeley: contact me here for any direct mail > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -- Wes Turner https://westurner.github.io/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From bussonniermatthias at gmail.com Mon Nov 3 04:05:25 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Mon, 3 Nov 2014 10:05:25 +0100 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> Message-ID: <2BF0E798-197A-4EF1-A3C2-E28AFF4E5FA3@gmail.com> Hi, Le 3 nov. 2014 ? 09:32, Wes Turner a ?crit : > Thanks! > > 1. Is there a link to the jsonschema? In the documentation? Please see the PR. https://github.com/ipython/ipython/pull/6045 > 2. How feasible would it be to write a JSON-LD context? This should have been discussed when we were in the refactoring of the notebook format, not once it's ready to merge. Metadata are free format you can add things like that if you like though. > 3. Where/how can/could Dublin Core metadata be added? It would be great to be able to index these documents with a title and authors. Same as above. The two last issues of extra metadata have extensively discussed, the new format do not change the discussions/problems that have been made. You can also refer to theses. Cheers, -- M > > https://en.wikipedia.org/wiki/Dublin_Core#DCMI_Metadata_Terms > > On Sun, Nov 2, 2014 at 5:23 PM, Fernando Perez wrote: > > On Sat, Nov 1, 2014 at 11:01 AM, Brian Granger wrote: > And thanks to the whole team (esp Min) for working on all of this! > > +lots. This has been a huge amount of slow, careful, not-very-sexy work. I am really thankful for the patience the whole team has had working on this, to help us set the notebook format as a solid foundation for long-term sharing and archival of computational work. > > This kind of effort provides few immediate rewards, but can have very significant long-term value. Thanks a lot for everyone who pitched in on that PR... > > Cheers, > > f > > > -- > Fernando Perez (@fperez_org; http://fperez.org) > fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) > fernando.perez-at-berkeley: contact me here for any direct mail > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > > -- > Wes Turner > https://westurner.github.io/ > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev -------------- next part -------------- An HTML attachment was scrubbed... URL: From wes.turner at gmail.com Mon Nov 3 04:21:35 2014 From: wes.turner at gmail.com (Wes Turner) Date: Mon, 3 Nov 2014 03:21:35 -0600 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: <2BF0E798-197A-4EF1-A3C2-E28AFF4E5FA3@gmail.com> References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> <2BF0E798-197A-4EF1-A3C2-E28AFF4E5FA3@gmail.com> Message-ID: Tardy to the party! IPython notebook nbformat v4 JSONSchema: https://github.com/minrk/ipython/blob/nbformat4/IPython/nbformat/v4/nbformat.v4.schema.json Metadata Documentation: https://github.com/minrk/ipython/blob/nbformat4/docs/source/notebook/nbformat.rst#metadata JSON-LD JSONSchema: https://github.com/json-ld/json-ld.org/blob/master/schemas/jsonld-schema.json On Mon, Nov 3, 2014 at 3:05 AM, Matthias Bussonnier < bussonniermatthias at gmail.com> wrote: > Hi, > Le 3 nov. 2014 ? 09:32, Wes Turner a ?crit : > > Thanks! > > 1. Is there a link to the jsonschema? In the documentation? > > > Please see the PR. https://github.com/ipython/ipython/pull/6045 > > 2. How feasible would it be to write a JSON-LD context? > > > This should have been discussed when we were in the refactoring of the > notebook format, > not once it's ready to merge. Metadata are free format you can add things > like that if you like though. > > 3. Where/how can/could Dublin Core metadata be added? It would be great to > be able to index these documents with a title and authors. > > > Same as above. > > The two last issues of extra metadata have extensively discussed, the new > format do not change > the discussions/problems that have been made. You can also refer to theses. > > Cheers, > -- > M > > > > > > https://en.wikipedia.org/wiki/Dublin_Core#DCMI_Metadata_Terms > > On Sun, Nov 2, 2014 at 5:23 PM, Fernando Perez > wrote: > >> >> On Sat, Nov 1, 2014 at 11:01 AM, Brian Granger >> wrote: >> >>> And thanks to the whole team (esp Min) for working on all of this! >>> >> >> +lots. This has been a huge amount of slow, careful, not-very-sexy work. >> I am really thankful for the patience the whole team has had working on >> this, to help us set the notebook format as a solid foundation for >> long-term sharing and archival of computational work. >> >> This kind of effort provides few immediate rewards, but can have very >> significant long-term value. Thanks a lot for everyone who pitched in on >> that PR... >> >> Cheers, >> >> f >> >> >> -- >> Fernando Perez (@fperez_org; http://fperez.org) >> fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) >> fernando.perez-at-berkeley: contact me here for any direct mail >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > > -- > Wes Turner > https://westurner.github.io/ > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -- Wes Turner https://westurner.github.io/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From jason-sage at creativetrax.com Mon Nov 3 07:50:42 2014 From: jason-sage at creativetrax.com (Jason Grout) Date: Mon, 03 Nov 2014 07:50:42 -0500 Subject: [IPython-dev] Conflict between IPython notebook javascript and three.js? In-Reply-To: References: <54538A46.2040509@creativetrax.com> Message-ID: <54577A22.3080706@creativetrax.com> On 11/2/14, 20:02, David Powell wrote: > Great, this looks like it will be the preferred solution to my problem > in the long term. I tried running the examples, but I ran into the > following javascript exception, and didn't get any output: > > Uncaught TypeError: Cannot read property 'obj' of undefined > pythreejs.js?_=1414974265086:329 It might be an installation issue, which isn't nearly as easy as it should be yet. Right now, (a) you need the ipython branch of the pythree.js repo, not the master branch, (b) you need to make sure three.js and pythree.js is installed and aliased in the right places (as in the instructions at the top of pythreejs.js), (c) you can't be using the latest IPython master, but it also may not work on the most recent released IPython (I've been developing just behind the latest master). It works fine for me on IPython commit fa61a81f879293895127f40004fe6cdbd60f1d0d, pythree.js commit a432be4628e4a0f92ab3651f9aa60ca4329c50ca, and three.js r69. I'll be working more on the installation difficulties in coming weeks, and I'm also working on some IPython commits to make it work in the latest IPython master. Thanks, Jason From jason-sage at creativetrax.com Mon Nov 3 07:54:54 2014 From: jason-sage at creativetrax.com (Jason Grout) Date: Mon, 03 Nov 2014 07:54:54 -0500 Subject: [IPython-dev] Conflict between IPython notebook javascript and three.js? In-Reply-To: References: <54538A46.2040509@creativetrax.com> Message-ID: <54577B1E.4040205@creativetrax.com> On 11/2/14, 20:02, David Powell wrote: > I avoided using OrbitControls for the same reason, but I didn't have > any problems with TrackballControls. I think with OrbitControls, you can use the noKeys option to disable key bindings. My pr mentioned above was about not even registering keybindings, but the noKeys just ignores keystrokes, so it's effectively the same thing for OrbitControls. I'm happy it works with TrackballControls. Maybe they've made some changes since I submitted the PR, which would be great. Thanks, Jason From antgonza at gmail.com Mon Nov 3 13:17:01 2014 From: antgonza at gmail.com (=?UTF-8?Q?Antonio_Gonz=C3=A1lez_Pe=C3=B1a?=) Date: Mon, 3 Nov 2014 11:17:01 -0700 Subject: [IPython-dev] ipcluster and PBS engines Message-ID: Hi, Let's say that I have a system with n nodes and m cores and I will like to start 10 engines using PBS. Note that n*m>10 and all my filesystems are shared so that's not a problem. My current issue is that if do: ipcluster start --n 10 and/or have in my ipcluster_config.py c.IPClusterEngines.n = 10 I always get 1 controller and 1 engine vs. having 1 controller and 10 engines submitted to PBS. Other option is that I'm misunderstanding how this works. Any help will be greatly appreciated. -- Antonio From benjaminrk at gmail.com Mon Nov 3 14:35:55 2014 From: benjaminrk at gmail.com (MinRK) Date: Mon, 3 Nov 2014 11:35:55 -0800 Subject: [IPython-dev] ipcluster and PBS engines In-Reply-To: References: Message-ID: It should create one PBS job array that starts ten engines. Are you only looking at qstat, or are you looking at the actual number of engines that end up connected via a parallel.Client? -MinRK ? On Mon, Nov 3, 2014 at 10:17 AM, Antonio Gonz?lez Pe?a wrote: > Hi, > > Let's say that I have a system with n nodes and m cores and I will > like to start 10 engines using PBS. Note that n*m>10 and all my > filesystems are shared so that's not a problem. My current issue is > that if do: > ipcluster start --n 10 > and/or have in my ipcluster_config.py > c.IPClusterEngines.n = 10 > I always get 1 controller and 1 engine vs. having 1 controller and 10 > engines submitted to PBS. > > Other option is that I'm misunderstanding how this works. > > Any help will be greatly appreciated. > > -- > Antonio > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: From antgonza at gmail.com Mon Nov 3 15:11:04 2014 From: antgonza at gmail.com (=?UTF-8?Q?Antonio_Gonz=C3=A1lez_Pe=C3=B1a?=) Date: Mon, 3 Nov 2014 13:11:04 -0700 Subject: [IPython-dev] ipcluster and PBS engines In-Reply-To: References: Message-ID: Just looking at qstat as I was expecting to have 10 actual jobs running. Expanding on that, what I was expecting is that if I have, let's say, 32 cores (4 nodes/workers) in all my cluster and I specify --n 10 that I have 10 engines and they are distributed between the 4 nodes. Is this possible? On Mon, Nov 3, 2014 at 12:35 PM, MinRK wrote: > It should create one PBS job array that starts ten engines. Are you only > looking at qstat, or are you looking at the actual number of engines that > end up connected via a parallel.Client? > > -MinRK > > > On Mon, Nov 3, 2014 at 10:17 AM, Antonio Gonz?lez Pe?a > wrote: >> >> Hi, >> >> Let's say that I have a system with n nodes and m cores and I will >> like to start 10 engines using PBS. Note that n*m>10 and all my >> filesystems are shared so that's not a problem. My current issue is >> that if do: >> ipcluster start --n 10 >> and/or have in my ipcluster_config.py >> c.IPClusterEngines.n = 10 >> I always get 1 controller and 1 engine vs. having 1 controller and 10 >> engines submitted to PBS. >> >> Other option is that I'm misunderstanding how this works. >> >> Any help will be greatly appreciated. >> >> -- >> Antonio >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -- Antonio From benjaminrk at gmail.com Mon Nov 3 15:22:24 2014 From: benjaminrk at gmail.com (MinRK) Date: Mon, 3 Nov 2014 12:22:24 -0800 Subject: [IPython-dev] ipcluster and PBS engines In-Reply-To: References: Message-ID: It?s up to your PBS system how it assigns the individual jobs in a job array. Under normal circumstance, this will be spread across nodes. If it isn?t you may need to use custom PBS templates to assign jobs how you want, or submit each bunch of engines separately: for i in {1..$NODES}; do ipcluster engines -n $ENGINES_PER_NODE &done The default job file looks like: #!/bin/sh#PBS -t 1-10#PBS -V#PBS -N ipengine /usr/bin/python -m IPython.parallel.engine -MinRK ? On Mon, Nov 3, 2014 at 12:11 PM, Antonio Gonz?lez Pe?a wrote: > Just looking at qstat as I was expecting to have 10 actual jobs > running. Expanding on that, what I was expecting is that if I have, > let's say, 32 cores (4 nodes/workers) in all my cluster and I specify > --n 10 that I have 10 engines and they are distributed between the 4 > nodes. Is this possible? > > On Mon, Nov 3, 2014 at 12:35 PM, MinRK wrote: > > It should create one PBS job array that starts ten engines. Are you only > > looking at qstat, or are you looking at the actual number of engines that > > end up connected via a parallel.Client? > > > > -MinRK > > > > > > On Mon, Nov 3, 2014 at 10:17 AM, Antonio Gonz?lez Pe?a < > antgonza at gmail.com> > > wrote: > >> > >> Hi, > >> > >> Let's say that I have a system with n nodes and m cores and I will > >> like to start 10 engines using PBS. Note that n*m>10 and all my > >> filesystems are shared so that's not a problem. My current issue is > >> that if do: > >> ipcluster start --n 10 > >> and/or have in my ipcluster_config.py > >> c.IPClusterEngines.n = 10 > >> I always get 1 controller and 1 engine vs. having 1 controller and 10 > >> engines submitted to PBS. > >> > >> Other option is that I'm misunderstanding how this works. > >> > >> Any help will be greatly appreciated. > >> > >> -- > >> Antonio > >> _______________________________________________ > >> IPython-dev mailing list > >> IPython-dev at scipy.org > >> http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > > > > > _______________________________________________ > > IPython-dev mailing list > > IPython-dev at scipy.org > > http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > > > -- > Antonio > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: From DavidAnthonyPowell+python at gmail.com Mon Nov 3 17:50:22 2014 From: DavidAnthonyPowell+python at gmail.com (David Powell) Date: Tue, 4 Nov 2014 09:50:22 +1100 Subject: [IPython-dev] Conflict between IPython notebook javascript and three.js? In-Reply-To: <54577B1E.4040205@creativetrax.com> References: <54538A46.2040509@creativetrax.com> <54577B1E.4040205@creativetrax.com> Message-ID: > I'm happy it works with TrackballControls. Maybe they've made some > changes since I submitted the PR, which would be great. I had a closer look. It seems that igraph includes an old copy of TrackballControls (from Feb 2013), and this is why it works without problem in the IPython notebook. To clarify, I found the key-bindings on TrackballControls do not seem to work at all (at least by default). I find this much preferable to the case of OrbitControls, where the key-bindings interfere with normal IPython navigation. From nick.bollweg at gmail.com Mon Nov 3 17:51:51 2014 From: nick.bollweg at gmail.com (Nicholas Bollweg) Date: Mon, 3 Nov 2014 17:51:51 -0500 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> <2BF0E798-197A-4EF1-A3C2-E28AFF4E5FA3@gmail.com> Message-ID: Re: JSON-LD: One of the goals of JSON-LD is to be able to add context to a well-defined corpus of data without changing its native format: in this case, the developers mean what they mean, and its important that it be meant this way... so the challenge is how to extract a useful semantic model of the document out of the existing format. The simplest approach is to use the context of: > { > "@context": { > "@vocab": "http://ipython.org/nbformat/v4/" > } > } > This will map all object keys, and values which are told to be @ids, as URIs in that namespace: example . (the playground doesn't have a way to do out-of-band expansion contexts, but it wouldn't have to be embedded like that) A more thorough-going approach might yield some more interesting things, but this is a good starting point, and just a few additions to the above context would be get really close to reflecting what is being said in a given notebook. IPython notebook nbformat v4 JSONSchema: > https://github.com/minrk/ipython/blob/nbformat4/IPython/nbformat/v4/nbformat.v4.schema.json > The schema is a great start: I haven't tried it, but there are some tools to automatically generate context from schema : some things that might be *interesting* to map to JSON-LD: - patternProperties in mimebundle. - In this case, though, it's referring to large, but agreed-upon enumeration of values (and not, like package.json's dependencies, an infinite number of package names). - with the above context, these would be lumped into the root of the namespace: http://ipython.org/nbformat/v4/text/html - this isn't so bad, probably - all the enums, i.e. cell_type: markdown - with the naive context above, it will map to a string - by setting the @type of cell_type to be @id in the context, markdown would expand to the URI http://ipython.org/nbformat/v4/markdown - this isn't so bad - Another option is to treat enums as more xml-like literals of a specific type, by setting @type of cell_type to be something like CellType - also not so bad - The advantage to having URIs vs. literals (they are both queryable) is that URIs can be the subject of something, and not just the object... not sure what we'd want to say in this case. - the wild west of metadata - JSON-LD can't tell the difference between cell metadata and notebook metadata - this is not so bad, as it is always "isolated" within the context of the s metadata, and wouldn't "pollute" the parent - with the naive context, everything will just fall into the main namespace. - this is bad. i don't see anything that can be done about it - all the lists - in JSON-LD, one can specify @container: @list - these are pretty bad in RDF, as it uses a bizarre lisp-like first and rest to represent them - nothing in the root that can map to an @id or @type - @id: not going there today - @type: nbformat is close, but there's nothing but duck typing to say, "I am a notebook" - loading these up into a graph would be *interesting*, as they would all just be blank nodes knocking around I'll do some more poking around, but think this is worth having! On Mon, Nov 3, 2014 at 4:21 AM, Wes Turner wrote: > Tardy to the party! > > IPython notebook nbformat v4 JSONSchema: > https://github.com/minrk/ipython/blob/nbformat4/IPython/nbformat/v4/nbformat.v4.schema.json > > > Metadata Documentation: > https://github.com/minrk/ipython/blob/nbformat4/docs/source/notebook/nbformat.rst#metadata > > > JSON-LD JSONSchema: > https://github.com/json-ld/json-ld.org/blob/master/schemas/jsonld-schema.json > > On Mon, Nov 3, 2014 at 3:05 AM, Matthias Bussonnier < > bussonniermatthias at gmail.com> wrote: > >> Hi, >> Le 3 nov. 2014 ? 09:32, Wes Turner a ?crit : >> >> Thanks! >> >> 1. Is there a link to the jsonschema? In the documentation? >> >> >> Please see the PR. https://github.com/ipython/ipython/pull/6045 >> >> 2. How feasible would it be to write a JSON-LD context? >> >> >> This should have been discussed when we were in the refactoring of the >> notebook format, >> not once it's ready to merge. Metadata are free format you can add things >> like that if you like though. >> >> 3. Where/how can/could Dublin Core metadata be added? It would be great >> to be able to index these documents with a title and authors. >> >> >> Same as above. >> >> The two last issues of extra metadata have extensively discussed, the new >> format do not change >> the discussions/problems that have been made. You can also refer to >> theses. >> >> Cheers, >> -- >> M >> >> >> >> >> >> https://en.wikipedia.org/wiki/Dublin_Core#DCMI_Metadata_Terms >> >> On Sun, Nov 2, 2014 at 5:23 PM, Fernando Perez >> wrote: >> >>> >>> On Sat, Nov 1, 2014 at 11:01 AM, Brian Granger >>> wrote: >>> >>>> And thanks to the whole team (esp Min) for working on all of this! >>>> >>> >>> +lots. This has been a huge amount of slow, careful, not-very-sexy >>> work. I am really thankful for the patience the whole team has had working >>> on this, to help us set the notebook format as a solid foundation for >>> long-term sharing and archival of computational work. >>> >>> This kind of effort provides few immediate rewards, but can have very >>> significant long-term value. Thanks a lot for everyone who pitched in on >>> that PR... >>> >>> Cheers, >>> >>> f >>> >>> >>> -- >>> Fernando Perez (@fperez_org; http://fperez.org) >>> fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) >>> fernando.perez-at-berkeley: contact me here for any direct mail >>> >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>> >>> >> >> >> -- >> Wes Turner >> https://westurner.github.io/ >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > > -- > Wes Turner > https://westurner.github.io/ > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From nathan12343 at gmail.com Mon Nov 3 17:58:49 2014 From: nathan12343 at gmail.com (Nathan Goldbaum) Date: Mon, 3 Nov 2014 14:58:49 -0800 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: <47C409E3-C4D0-4C3E-B7CA-44D871902B16@gmail.com> References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> <54566E9A.2040002@gmail.com> <47C409E3-C4D0-4C3E-B7CA-44D871902B16@gmail.com> Message-ID: On Sun, Nov 2, 2014 at 10:37 AM, Matthias Bussonnier < bussonniermatthias at gmail.com> wrote: > Hi Carlos, > > TL:DR; at the end. > > With the notebook format v4, the way notebooks are written on disk will > change. > After now some time with v3, we know that a few things could have been > done better > or in a more general way, thus we improved the notebook format. > > Once 3.0 and 2.4 are released this shouldn't change much for the end user. > I'll go over some of the changes in the notebook format and their reasons, > but you > should read the IPep for all details. > > > Get rid of python specific names. > pyin, pyout, pyerr are renamed input input, output, and error (roughly, > read IPep for exact names). > > Uniformise jpg/png/text keys in output to be a mimetype. > You can now store application/x-pdf or doc/microsoft-word in a notebook. > If the frontend now how to read it, it will do something with it i suppose. > It remove a lot of special casing. > > Get rid of worksheets, almost nobody use it, and there are other smarter > way to > store them if we wanted to. > > Text cells had their text under the 'text' key, and code cell under > 'source'. > Which is a bit silly, as it forces you to do some special casing where not > necessary. > > Heading cell are now gone (from the file format), Technology evolve fast, > and we can > now detect the #*n in markdown cell and convert properly to LaTeX, and add > anchor > in nbviewer. > > More importantly, we know have a jsonschema that describe the format. > So we can validate the notebook, and know that for example, a prompt number > is either null, or a number. v3 was implicitly allowing '*' (star), which > would happened > if you were saving while cell are running, which took by surprise both us > and external > library where the conversion of some notebook made crashes. This include > also security risk. > Which I won't develop, cause I won't develop, but those who know know I > love javascipt > injection, and you can do nasty things on known websites. By insuring the > type of each field > of the notebook you lower the attack area, and protect FooBar corp user > from attacks > even if FooBar corp does not really respond to you when you do responsible > disclosure and > finally give up. > > I won't describe "All the things", but you see the big picture. > It's better, faster, stronger. > > TL;DR: > > Wat ? You don't know ?! I've heard that IPython notebook format v3 might > have been responsible > for the death of a huge number of kittens due to developers banging their > head on their desk. > It is also probable that the need of extra computing power to run more > test because of it's complexity > is in part responsible for global warming. Also the new format prevent the > use of %pylab that kills > the endangered species of newbies. The naming convention was also really > strange for php > coders that don't understand the py prefixes. > > As we like kittens, newbies, developers, php coders (but not php itself > [1]), and dislike > global warming we decided to fix that. > > It also comes as ipynb v4+ that have a 5.5 inches screen, and we plan to > jump from v8 to v10 > directly. We also removed touch-id from metadata so that police cannot > force youth unlock your > ipynb. > > Hope that shine a light on a few of the reason and what will change. > Tell us if you have any more questions. > You don't mention it here, but nbformatv4 also includes the autoscroll behavior for a cell as part of the notebook metadata, meaning no more resetting autoscroll and issue 2172 can be fixed! https://github.com/ipython/ipython/issues/2172 > > > Cheers, > -- > M > > [1]: But MinRK Love Javascript. > > > > > > Le 2 nov. 2014 ? 18:49, Carlos C?rdoba a ?crit : > > Hi, > > Could you explain a bit more what this change is about? > > Cheers, > Carlos > > El 01/11/14 a las 13:01, Brian Granger escribi?: > > Matthias, > > Thanks for posting here about these changes! > > And thanks to the whole team (esp Min) for working on all of this! > > Cheers, > > Brian > > On Sat, Nov 1, 2014 at 3:12 AM, Matthias Bussonnier wrote: > > Hello Jovyans and other beings from Jupyter galilean moons system, > > > I'm writing to you to warn that on Monday (probably morning) a relatively > important change > in IPython will land on master. In particular it does change the notebook > structure quite a bit. > > Cf https://github.com/ipython/ipython/pull/6045 for more information. > > As nothing is ever perfect, please do back-up your notebooks before > upgrading. With this new update, > saved notebook of this new of IPython version won't be compatible with > older IPython anymore, nor will nbviewer be able to render them. > > The new notebook format will be back-ported on older (2.x) version of > IPython and the new notebook format > will be supported on nbviewer, we will just need a few days to port all the > changes. > > There is of course be a way to manually downgrade the notebook from v4 to v3 > using nbconvert. > > Eventually the more adventurous can test the branch this week-end. > > As usual, be prepared for data loss and bugs, so update your git branches, > refresh submodules, clear your browser caches, roll up your sleeves, and > send bug reports. > > Cheers, > -- > Matthias > > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jason-sage at creativetrax.com Mon Nov 3 22:12:20 2014 From: jason-sage at creativetrax.com (Jason Grout) Date: Mon, 03 Nov 2014 22:12:20 -0500 Subject: [IPython-dev] Conflict between IPython notebook javascript and three.js? In-Reply-To: References: <54538A46.2040509@creativetrax.com> <54577B1E.4040205@creativetrax.com> Message-ID: <54584414.1070305@creativetrax.com> On 11/3/14, 17:50, David Powell wrote: > I had a closer look. It seems that igraph includes an old copy of > TrackballControls (from Feb 2013), and this is why it works without > problem in the IPython notebook. > > To clarify, I found the key-bindings on TrackballControls do not seem > to work at all (at least by default). I find this much preferable to > the case of OrbitControls, where the key-bindings interfere with > normal IPython navigation. Thanks. In that case, you can use the noKeys option for the OrbitControls to turn off keyboard handling. Thanks, Jason From wes.turner at gmail.com Mon Nov 3 23:00:12 2014 From: wes.turner at gmail.com (Wes Turner) Date: Mon, 3 Nov 2014 22:00:12 -0600 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> <2BF0E798-197A-4EF1-A3C2-E28AFF4E5FA3@gmail.com> Message-ID: > I'll do some more poking around, but think this is worth having! Thank you so much! On Mon, Nov 3, 2014 at 4:51 PM, Nicholas Bollweg wrote: > Re: JSON-LD: > > One of the goals of JSON-LD is to be able to add context to a well-defined > corpus of data without changing its native format: in this case, the > developers mean what they mean, and its important that it be meant this > way... so the challenge is how to extract a useful semantic model of the > document out of the existing format. > > The simplest approach is to use the context of: > >> { >> "@context": { >> "@vocab": "http://ipython.org/nbformat/v4/" >> } >> } >> > > This will map all object keys, and values which are told to be @ids, as > URIs in that namespace: example > > . > > (the playground doesn't have a way to do out-of-band expansion contexts, > but it wouldn't have to be embedded like that) > > > A more thorough-going approach might yield some more interesting things, > but this is a good starting point, and just a few additions to the above > context would be get really close to reflecting what is being said in a > given notebook. > > IPython notebook nbformat v4 JSONSchema: >> https://github.com/minrk/ipython/blob/nbformat4/IPython/nbformat/v4/nbformat.v4.schema.json >> > > The schema is a great start: I haven't tried it, but there are some tools > to automatically generate context from schema > : > > some things that might be *interesting* to map to JSON-LD: > > - patternProperties in mimebundle. > - In this case, though, it's referring to large, but agreed-upon > enumeration of values (and not, like package.json's dependencies, > an infinite number of package names). > - with the above context, these would be lumped into the root of > the namespace: http://ipython.org/nbformat/v4/text/html > - this isn't so bad, probably > - all the enums, i.e. cell_type: markdown > - with the naive context above, it will map to a string > - by setting the @type of cell_type to be @id in the context, markdown > would expand to the URI > http://ipython.org/nbformat/v4/markdown > - this isn't so bad > - Another option is to treat enums as more xml-like literals of > a specific type, by setting @type of cell_type to be something like > CellType > - also not so bad > - The advantage to having URIs vs. literals (they are both > queryable) is that URIs can be the subject of something, and not just the > object... not sure what we'd want to say in this case. > - the wild west of metadata > - JSON-LD can't tell the difference between cell metadata and > notebook metadata > - this is not so bad, as it is always "isolated" within the > context of the s metadata, and wouldn't "pollute" the parent > - with the naive context, everything will just fall into the > main namespace. > - this is bad. i don't see anything that can be done about it > - all the lists > - in JSON-LD, one can specify @container: @list > - these are pretty bad in RDF, as it uses a bizarre lisp-like first > and rest to represent them > - nothing in the root that can map to an @id or @type > - @id: not going there today > - @type: nbformat is close, but there's nothing but duck typing to > say, "I am a notebook" > - loading these up into a graph would be *interesting*, as they > would all just be blank nodes knocking around > > I'll do some more poking around, but think this is worth having! > > On Mon, Nov 3, 2014 at 4:21 AM, Wes Turner wrote: > >> Tardy to the party! >> >> IPython notebook nbformat v4 JSONSchema: >> https://github.com/minrk/ipython/blob/nbformat4/IPython/nbformat/v4/nbformat.v4.schema.json >> >> >> Metadata Documentation: >> https://github.com/minrk/ipython/blob/nbformat4/docs/source/notebook/nbformat.rst#metadata >> >> >> JSON-LD JSONSchema: >> https://github.com/json-ld/json-ld.org/blob/master/schemas/jsonld-schema.json >> >> On Mon, Nov 3, 2014 at 3:05 AM, Matthias Bussonnier < >> bussonniermatthias at gmail.com> wrote: >> >>> Hi, >>> Le 3 nov. 2014 ? 09:32, Wes Turner a ?crit : >>> >>> Thanks! >>> >>> 1. Is there a link to the jsonschema? In the documentation? >>> >>> >>> Please see the PR. https://github.com/ipython/ipython/pull/6045 >>> >>> 2. How feasible would it be to write a JSON-LD context? >>> >>> >>> This should have been discussed when we were in the refactoring of the >>> notebook format, >>> not once it's ready to merge. Metadata are free format you can add >>> things like that if you like though. >>> >>> 3. Where/how can/could Dublin Core metadata be added? It would be great >>> to be able to index these documents with a title and authors. >>> >>> >>> Same as above. >>> >>> The two last issues of extra metadata have extensively discussed, the >>> new format do not change >>> the discussions/problems that have been made. You can also refer to >>> theses. >>> >>> Cheers, >>> -- >>> M >>> >>> >>> >>> >>> >>> https://en.wikipedia.org/wiki/Dublin_Core#DCMI_Metadata_Terms >>> >>> On Sun, Nov 2, 2014 at 5:23 PM, Fernando Perez >>> wrote: >>> >>>> >>>> On Sat, Nov 1, 2014 at 11:01 AM, Brian Granger >>>> wrote: >>>> >>>>> And thanks to the whole team (esp Min) for working on all of this! >>>>> >>>> >>>> +lots. This has been a huge amount of slow, careful, not-very-sexy >>>> work. I am really thankful for the patience the whole team has had working >>>> on this, to help us set the notebook format as a solid foundation for >>>> long-term sharing and archival of computational work. >>>> >>>> This kind of effort provides few immediate rewards, but can have very >>>> significant long-term value. Thanks a lot for everyone who pitched in on >>>> that PR... >>>> >>>> Cheers, >>>> >>>> f >>>> >>>> >>>> -- >>>> Fernando Perez (@fperez_org; http://fperez.org) >>>> fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) >>>> fernando.perez-at-berkeley: contact me here for any direct mail >>>> >>>> _______________________________________________ >>>> IPython-dev mailing list >>>> IPython-dev at scipy.org >>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>> >>>> >>> >>> >>> -- >>> Wes Turner >>> https://westurner.github.io/ >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>> >>> >>> >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>> >>> >> >> >> -- >> Wes Turner >> https://westurner.github.io/ >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -- Wes Turner https://westurner.github.io/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From patrickfuller at gmail.com Mon Nov 3 23:14:16 2014 From: patrickfuller at gmail.com (Patrick Fuller) Date: Mon, 3 Nov 2014 22:14:16 -0600 Subject: [IPython-dev] Conflict between IPython notebook javascript and three.js? In-Reply-To: <54584414.1070305@creativetrax.com> References: <54538A46.2040509@creativetrax.com> <54577B1E.4040205@creativetrax.com> <54584414.1070305@creativetrax.com> Message-ID: David, Thanks for jogging my memory on TrackballControls - I edited that file manually. I submitted a pull request to three.js about a year ago regarding the edits, but just used my modified TrackballControls in the interim and never updated it. Anyway, I don?t think TrackballControls is the issue. I just updated it in the igraph repo, and the only difference I?m noticing is that the rotation speed is much higher. Pat ? On Mon, Nov 3, 2014 at 9:12 PM, Jason Grout wrote: > On 11/3/14, 17:50, David Powell wrote: > > I had a closer look. It seems that igraph includes an old copy of > > TrackballControls (from Feb 2013), and this is why it works without > > problem in the IPython notebook. > > > > To clarify, I found the key-bindings on TrackballControls do not seem > > to work at all (at least by default). I find this much preferable to > > the case of OrbitControls, where the key-bindings interfere with > > normal IPython navigation. > > Thanks. In that case, you can use the noKeys option for the > OrbitControls to turn off keyboard handling. > > Thanks, > > Jason > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: From benjaminrk at gmail.com Mon Nov 3 23:43:10 2014 From: benjaminrk at gmail.com (MinRK) Date: Mon, 3 Nov 2014 20:43:10 -0800 Subject: [IPython-dev] Jupyterhub behind NGINX redirect In-Reply-To: References: Message-ID: I suspect this is just a failure to handle and test the base_url properly. I haven't used it behind a URL prefix yet, so I wouldn't be surprised if I left out the prefix in a few places. Thanks for trying it out! On Sun, Nov 2, 2014 at 11:13 PM, Clare Sloggett wrote: > Hi all, > > We've managed to troubleshoot our other issues, so I'm now sure that I can > run jupyterhub successfully *without* a redirect. So I'm back to trying to > get a redirect working. I've cc'd Nuwan who's been working on this with me > and probably understands the issues better than I do. > > Currently, if I run something like > $ jupyterhub --port 9520 ----JupyterHubApp.hub_port=8500 > > then I can access JupyterHub at http://my-url:9520/ . So far, success! > This works fine, I can log in as various linux users, edit notebooks, etc. > > What we'd like is for it to instead be available to users at > http://my-url/jupyter/ . > > We assumed that the command-line option JupyterHubApp.base_url is supposed > to be used for this. We are trying to set up an NGINX redirect from > /jupyter/ to :9520 and run jupyterhub so that it behaves properly in this > situation. However JupyterHubApp.base_url doesn't seem to be doing what we > expect it to. Is anyone able to explain how this parameter is supposed to > be used? > > Cheers, > Clare > > On 28 October 2014 18:15, Clare Sloggett wrote: > >> Hi Min, >> >> Thanks for your help! >> >> Actually at the moment I have been pushed back to a more fundamental >> problem and can't even reproduce the error I was posting about in this >> email. I can't get jupyterhub to run at all. I posted this separate error >> in another thread, which I've just replied to a moment ago: "Error running >> JupyterHub". >> >> I suspect the issue is mine as I encountered it a few days ago and am >> still encountering it after updating to the latest commit. It seems >> unlikely a bug has survived through several commits without someone else >> discovering it too. But I'm having trouble working out what the cause is, >> and am not sure what to make of the error messages it's throwing (both >> python *and* javascript errors are thrown when it crashes). Full error >> printout is in that thread! >> >> Cheers, >> Clare >> >> On 26 October 2014 06:34, MinRK wrote: >> >>> Clare, >>> >>> Can you update to the latest master? I think the never-ending redirects >>> could be the result of a recently fixed typo. >>> >>> Thanks, >>> -MinRK >>> >>> On Thu, Oct 23, 2014 at 10:01 PM, Clare Sloggett < >>> claresloggett at gmail.com> wrote: >>> >>>> Hi Doug, >>>> >>>> Thanks for this. I had actually just been thinking about about the >>>> NGINX redirect issue, and had assumed everything behind that would be >>>> fairly straightforward. But it sounds like you are saying there may be more >>>> fundamental issues, and to be honest I haven't tested that the redirect is >>>> definitely the source of all my problems. >>>> >>>> It sounds like I need to do some more direct testing and come back! >>>> >>>> In the meantime, if anyone has insight into what could be causing a URL >>>> rewrite like "/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython...." >>>> that could be really helpful. >>>> >>>> Cheers, >>>> Clare >>>> >>>> On 13 October 2014 00:44, Doug Blank wrote: >>>> >>>>> On Sun, Oct 12, 2014 at 1:35 AM, Clare Sloggett < >>>>> claresloggett at gmail.com> wrote: >>>>> > >>>>> > Hi all, >>>>> > >>>>> > Am I right in thinking this list is also the right place for >>>>> questions about Jupyterhub? >>>>> > >>>>> > I'm trying to set up Jupyterhub for multiple users, on the same >>>>> server where we are running several other services. Currently there are >>>>> NGINX redirects in place to these other services. I'd like to set up http:///ipython/ >>>>> to redirect to a port on localhost and run Jupyterhub over that port. Is >>>>> this possible? >>>>> > >>>>> > I've got this working in the past with single-user IPython Notebook. >>>>> For that, I set config options c.NotebookApp.base_project_url, >>>>> c.NotebookApp.base_kernel_url, and c.NotebookApp.webapp_settings to be >>>>> aware of the ipython/ URL prefix. We used an NGINX redirect to forward >>>>> requests and handle websockets properly, which looked like >>>>> > >>>>> > location /ipython/ { >>>>> > proxy_pass http://127.0.0.1:9510; >>>>> > proxy_set_header Host $host; >>>>> > proxy_set_header X-Real-IP $remote_addr; >>>>> > proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; >>>>> > proxy_set_header Upgrade $http_upgrade; >>>>> > proxy_set_header Connection "upgrade"; >>>>> > } >>>>> > >>>>> > This worked for the old single-user notebook. But, I'm not clear on >>>>> the model Jupyterhub is using (and I'm not much of a sysadmin). I read >>>>> through the command-line options, thought from them that I don't need to >>>>> strip the /ipython/ from the incoming requests, and have tried using a >>>>> similar redirect to the above and setting >>>>> --JupyterHubApp.base_url='ipython/' and also setting --port 9510. This does >>>>> seem to see the incoming requests but results in amusing requests like >>>>> > >>>>> > 500 GET >>>>> /ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/...... >>>>> > >>>>> > ... so clearly I haven't understood what's going on. Does anyone >>>>> have any pointers on how this should work? I haven't read any docs other >>>>> than the README and the command-line parameter information, so apologies if >>>>> there's something obvious that I didn't look at. >>>>> > >>>>> > I'm also interested in running Jupyterhub as a daemon, so if that's >>>>> something that's been done before and there's anything I should know, that >>>>> would be great. >>>>> >>>>> There is one other document, other than the README.md and the reported >>>>> issues [1]: >>>>> >>>>> >>>>> https://github.com/jupyter/jupyterhub/wiki/Using-sudo-to-run-the-server-as-non-root >>>>> >>>>> Despite the warning at the top of that page, you can make jupyterhub >>>>> work on some systems (eg, Linux, such as Ubuntu) using sudo, but not as >>>>> root. I suspect that this would be the recommended setup when jupyterhub is >>>>> complete. One recent change is the ability to save/load state from a >>>>> database. The wiki page above hasn't been updated with the note from this >>>>> issue: >>>>> >>>>> https://github.com/jupyter/jupyterhub/issues/57 >>>>> >>>>> It might be easier to start without NGINX, and then add it after you >>>>> have a working jupyterhub system. It would be nice to have a little bash >>>>> script to make this a "service" that would support "start", "top", and >>>>> "status"... but I haven't had time. Currently, we're just becoming the >>>>> non-root sudoer ("rhea" in the docs) and starting the server, something >>>>> along the lines: >>>>> >>>>> jupyterhub --LocalProcessSpawner.set_user=sudo >>>>> --JupyterHubApp.ip=165.106.10.83 --JupyterHubApp.port=80 >>>>> --db='sqlite:///:memory:' &>> /var/log/jupyterhub/log & >>>>> >>>>> The next step for us is to get it running under https... looks like >>>>> others have blazed that trail, so it looks possible. >>>>> >>>>> If you have success, it would be great to add to the wiki docs... I >>>>> suspect that many of us that aren't sys admins will be wanting to get this >>>>> up and running. >>>>> >>>>> -Doug >>>>> >>>>> [1] - https://github.com/jupyter/jupyterhub >>>>> >>>>> > Any help much appreciated! >>>>> > >>>>> > Clare >>>>> > >>>>> > >>>>> > _______________________________________________ >>>>> > IPython-dev mailing list >>>>> > IPython-dev at scipy.org >>>>> > http://mail.scipy.org/mailman/listinfo/ipython-dev >>>>> > >>>>> >>>>> _______________________________________________ >>>>> IPython-dev mailing list >>>>> IPython-dev at scipy.org >>>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>>> >>>>> >>>> >>>> _______________________________________________ >>>> IPython-dev mailing list >>>> IPython-dev at scipy.org >>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>> >>>> >>> >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>> >>> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From zvoros at gmail.com Tue Nov 4 05:23:05 2014 From: zvoros at gmail.com (=?UTF-8?B?Wm9sdMOhbiBWw7Zyw7Zz?=) Date: Tue, 04 Nov 2014 11:23:05 +0100 Subject: [IPython-dev] notebook hangs with latest master Message-ID: <5458A909.8050304@gmail.com> Hi all, I have checked out the latest code from the master branch, and run into two problems. The first is that the notebook frontend hangs, when I try to load older (meaning 2 days old) notebooks. The bullet on the top right hand side stays solid, instead of turning empty. At that point, nothing can be done with the client, it does not respond to anything, and the notebook is not reported in the "Running" tab of the dashboard. The problem persists even, if I make the notebook trusted on the command line. The other problem is a rendering issue: when I create a new notebook (this can be done without problems, and then the results can be loaded later), and I execute, say, 3*3, I get $$9$$ in the output. This happens, when I run the the server with the ipython configuration file that I attached. Is this an issue in the kernel as such, or something in the sympyprinting extension? These issues don't seem to be related to the browser, or the browser history: I have cleared the cache, and tried to run the notebooks in chrome and firefox. I have also run > python setup.py submodule Since that didn't work, I removed the whole git folder, and cloned the content again. That didn't help either. I still get an /> running css />/ />/ Failed to build css sourcemaps: [Errno 2] No such file or directory />/ checking package data/ error, when trying to build. Last time Matthias suggested python setup.py submodule, and that solved the problem, but not now. What I also noticed is that the old notebooks have the header { "metadata": { "name": "", "signature": "sha256:166a240702e584a760e43c90f3d1422154ca5aec932554de10f4922aaa1e517a" }, "nbformat": 3, "nbformat_minor": 0, "worksheets": [ while the new ones are bare; they kick out like this: { "cells": [ { "cell_type": "code", "execution_count": 1, "metadata": { "collapsed": false }, "outputs": [ and the metadata is moved to the end of the file, and the "worksheet" dictionary entry is removed. I wonder, whether some of the problems are related to a change in the notebook format. One more thing: in the toolbar, the dropdown menu for the cell type contains only code, raw, and markdown, but all the headings are gone. Is this on purpose? That would be quite bad. But this might be related to the failed css build. Does someone else have the same difficulties, or something is messed up on my side? Cheers, Zolt?n -------------- next part -------------- A non-text attachment was scrubbed... Name: ipython_config.py Type: text/x-python Size: 977 bytes Desc: not available URL: From lucienboillod at gmail.com Tue Nov 4 06:34:55 2014 From: lucienboillod at gmail.com (Lucien Boillod) Date: Tue, 4 Nov 2014 12:34:55 +0100 Subject: [IPython-dev] conflict between d3 keyCode event and IPython notebook Message-ID: Hello all, I have a problem to use keyCode event in D3.js into a notebook, indeed there is some key that are associated with some notebooks actions, that I can?t use in D3. To show clearly my problem I made a simple example. As you can see delete will show the console log, but L will only execute the associate notebook action. Cheers, Lucien -------------- next part -------------- A non-text attachment was scrubbed... Name: keycode example.ipynb Type: application/octet-stream Size: 3652 bytes Desc: not available URL: From bussonniermatthias at gmail.com Tue Nov 4 06:54:48 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Tue, 4 Nov 2014 12:54:48 +0100 Subject: [IPython-dev] notebook hangs with latest master In-Reply-To: <5458A909.8050304@gmail.com> References: <5458A909.8050304@gmail.com> Message-ID: <687F1821-5449-4509-90BA-15F84EE3F66C@gmail.com> Le 4 nov. 2014 ? 11:23, Zolt?n V?r?s a ?crit : > Hi all, > > I have checked out the latest code from the master branch, and run into two problems. > > The first is that the notebook frontend hangs, when I try to load older (meaning 2 days old) notebooks. The bullet on the top right hand side stays solid, instead of turning empty. At that point, nothing can be done with the client, it does not respond to anything, and the notebook is not reported in the "Running" tab of the dashboard. The problem persists even, if I make the notebook trusted on the command line. Do you get any errors on the Javascript console in the browser ? Are you on tornado 4? > > The other problem is a rendering issue: when I create a new notebook (this can be done without problems, and then the results can be loaded later), and I execute, say, 3*3, I get $$9$$ in the output. This happens, when I run the the server with the ipython configuration file that I attached. Is this an issue in the kernel as such, or something in the sympyprinting extension? Probably a bug in JS (cf above) > > These issues don't seem to be related to the browser, or the browser history: I have cleared the cache, and tried to run the notebooks in chrome and firefox. I have also run > > > python setup.py submodule > > Since that didn't work, I removed the whole git folder, and cloned the content again. That didn't help either. I still get an > > /> running css > />/ />/ Failed to build css sourcemaps: [Errno 2] No such file or directory > />/ checking package data/ You might need less to build the css. But if you clone --recursive, you shouldn't ned to build anything. > error, when trying to build. Last time Matthias suggested python setup.py submodule, and that solved the problem, but not now. > > What I also noticed is that the old notebooks have the header > > { > "metadata": { > "name": "", > "signature": "sha256:166a240702e584a760e43c90f3d1422154ca5aec932554de10f4922aaa1e517a" > }, > "nbformat": 3, > "nbformat_minor": 0, > "worksheets": [ > > > while the new ones are bare; they kick out like this: > > { > "cells": [ > { > "cell_type": "code", > "execution_count": 1, > "metadata": { > "collapsed": false > }, > "outputs": [ > > and the metadata is moved to the end of the file, and the "worksheet" dictionary entry is removed. I wonder, whether some of the problems are related to a change in the notebook format. This is indeed the noes notebook format. I though we were using an ordered dict to write file on disk, but order shouldn't matter. > > One more thing: in the toolbar, the dropdown menu for the cell type contains only code, raw, and markdown, but all the headings are gone. Is this on purpose? That would be quite bad. But this might be related to the failed css build. Markdown cell with # on the beginning will be converted to heading cell, the UI to warn about that is not there yet. -- M > Does someone else have the same difficulties, or something is messed up on my side? > > Cheers, > Zolt?n > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev From zvoros at gmail.com Tue Nov 4 07:34:46 2014 From: zvoros at gmail.com (=?UTF-8?B?Wm9sdMOhbiBWw7Zyw7Zz?=) Date: Tue, 04 Nov 2014 13:34:46 +0100 Subject: [IPython-dev] notebook hangs with latest master In-Reply-To: <687F1821-5449-4509-90BA-15F84EE3F66C@gmail.com> References: <5458A909.8050304@gmail.com> <687F1821-5449-4509-90BA-15F84EE3F66C@gmail.com> Message-ID: <5458C7E6.1090605@gmail.com> Hi Matthias, On 11/04/2014 12:54 PM, Matthias Bussonnier wrote: > Le 4 nov. 2014 ? 11:23, Zolt?n V?r?s a ?crit : > >> Hi all, >> >> I have checked out the latest code from the master branch, and run into two problems. >> >> The first is that the notebook frontend hangs, when I try to load older (meaning 2 days old) notebooks. The bullet on the top right hand side stays solid, instead of turning empty. At that point, nothing can be done with the client, it does not respond to anything, and the notebook is not reported in the "Running" tab of the dashboard. The problem persists even, if I make the notebook trusted on the command line. > Do you get any errors on the Javascript console in the browser ? Here is my error log: Loaded extension: toc :8888/static/base/js/utils.js:26 ? toc.js:147 Default extension for cell metadata editing loaded. default.js:48 Raw Cell Format toolbar preset loaded. rawcell.js:82 Slideshow extension for metadata editing loaded. slideshow.js:43 Ignoring untrusted image/svg+xml output. outputarea.js:518 Uncaught ReferenceError: nb is not defined notebook.js:2198 Failed to load resource: the server responded with a status of 404 (Not Found) http://localhost:8888/static/components/codemirror/lib/codemirror Uncaught Error: Script error for: /static/components/codemirror/lib/codemirror http://requirejs.org/docs/errors.html#scripterror require.js:141 Failed to load resource: the server responded with a status of 404 (Not Found) http://localhost:8888/static/components/codemirror/addon/fold/foldcode Uncaught Error: Script error for: /static/components/codemirror/addon/fold/foldcode http://requirejs.org/docs/errors.html#scripterror require.js:141 GET http://localhost:8888/static/components/underscore/underscore-min.map 404 (Not Found) :8888/static/components/underscore/underscore-min.map:1 GET http://localhost:8888/static/components/backbone/backbone-min.map 404 (Not Found) :8888/static/components/backbone/backbone-min.map:1 The two failed GET requests are there even for new notebooks, so that is probably not the problem. The codemirror errors shouldn't matter either, because I have proper code highlighting in a new notebook. (Although, I have noticed that the contents of subfolders in profile/static/custom are no longer loaded. I have also checked the content of/usr/local/lib/python2.7/dist-packages/IPython/html/static/components/codemirror/addon/fold, and that is not an empty folder, so I suppose that all components are there.) The only major difference that I see in the error log of a new and old notebook is Uncaught ReferenceError: nb is not defined notebook.js:2198 > Are you on tornado 4? No, I was running version 3.1. I have removed that, and now I have 4.0.2. That didn't change the situation. >> The other problem is a rendering issue: when I create a new notebook (this can be done without problems, and then the results can be loaded later), and I execute, say, 3*3, I get $$9$$ in the output. This happens, when I run the the server with the ipython configuration file that I attached. Is this an issue in the kernel as such, or something in the sympyprinting extension? > Probably a bug in JS (cf above) > >> These issues don't seem to be related to the browser, or the browser history: I have cleared the cache, and tried to run the notebooks in chrome and firefox. I have also run >> >>> python setup.py submodule >> Since that didn't work, I removed the whole git folder, and cloned the content again. That didn't help either. I still get an >> >> /> running css >> />/ />/ Failed to build css sourcemaps: [Errno 2] No such file or directory >> />/ checking package data/ > You might need less to build the css. But if you clone --recursive, you shouldn't ned to build anything. I have tried this, too, the same problem. > > >> error, when trying to build. Last time Matthias suggested python setup.py submodule, and that solved the problem, but not now. >> >> What I also noticed is that the old notebooks have the header >> >> { >> "metadata": { >> "name": "", >> "signature": "sha256:166a240702e584a760e43c90f3d1422154ca5aec932554de10f4922aaa1e517a" >> }, >> "nbformat": 3, >> "nbformat_minor": 0, >> "worksheets": [ >> >> >> while the new ones are bare; they kick out like this: >> >> { >> "cells": [ >> { >> "cell_type": "code", >> "execution_count": 1, >> "metadata": { >> "collapsed": false >> }, >> "outputs": [ >> >> and the metadata is moved to the end of the file, and the "worksheet" dictionary entry is removed. I wonder, whether some of the problems are related to a change in the notebook format. > This is indeed the noes notebook format. > I though we were using an ordered dict to write file on disk, but order shouldn't matter. But in the new version, the 'worksheets' entry is completely removed, isn't it? > Markdown cell with # on the beginning will be converted to heading cell, the UI to warn about that is not there yet. Quite neat! Cheers, Zolt?n From doug.blank at gmail.com Tue Nov 4 07:37:45 2014 From: doug.blank at gmail.com (Doug Blank) Date: Tue, 4 Nov 2014 07:37:45 -0500 Subject: [IPython-dev] notebook hangs with latest master In-Reply-To: <5458C7E6.1090605@gmail.com> References: <5458A909.8050304@gmail.com> <687F1821-5449-4509-90BA-15F84EE3F66C@gmail.com> <5458C7E6.1090605@gmail.com> Message-ID: I am seeing this occasionally, too, and I believe I have the same output as Zolt?n. Haven't had time to explore further yet... -Doug On Tue, Nov 4, 2014 at 7:34 AM, Zolt?n V?r?s wrote: > Hi Matthias, > > On 11/04/2014 12:54 PM, Matthias Bussonnier wrote: > > Le 4 nov. 2014 ? 11:23, Zolt?n V?r?s a ?crit : > > > >> Hi all, > >> > >> I have checked out the latest code from the master branch, and run into > two problems. > >> > >> The first is that the notebook frontend hangs, when I try to load older > (meaning 2 days old) notebooks. The bullet on the top right hand side stays > solid, instead of turning empty. At that point, nothing can be done with > the client, it does not respond to anything, and the notebook is not > reported in the "Running" tab of the dashboard. The problem persists even, > if I make the notebook trusted on the command line. > > Do you get any errors on the Javascript console in the browser ? > > Here is my error log: > > Loaded extension: toc :8888/static/base/js/utils.js:26 > toc.css">? toc.js:147 > Default extension for cell metadata editing loaded. default.js:48 > Raw Cell Format toolbar preset loaded. rawcell.js:82 > Slideshow extension for metadata editing loaded. slideshow.js:43 > Ignoring untrusted image/svg+xml output. outputarea.js:518 > Uncaught ReferenceError: nb is not defined notebook.js:2198 > Failed to load resource: the server responded with a status of 404 (Not > Found) http://localhost:8888/static/components/codemirror/lib/codemirror > Uncaught Error: Script error for: > /static/components/codemirror/lib/codemirror > http://requirejs.org/docs/errors.html#scripterror require.js:141 > Failed to load resource: the server responded with a status of 404 (Not > Found) > http://localhost:8888/static/components/codemirror/addon/fold/foldcode > Uncaught Error: Script error for: > /static/components/codemirror/addon/fold/foldcode > http://requirejs.org/docs/errors.html#scripterror require.js:141 > GET > http://localhost:8888/static/components/underscore/underscore-min.map > 404 (Not Found) :8888/static/components/underscore/underscore-min.map:1 > GET http://localhost:8888/static/components/backbone/backbone-min.map > 404 (Not Found) :8888/static/components/backbone/backbone-min.map:1 > > > The two failed GET requests are there even for new notebooks, so that is > probably not the problem. The codemirror errors shouldn't matter either, > because I have proper code highlighting in a new notebook. (Although, I > have noticed that the contents of subfolders in profile/static/custom > are no longer loaded. I have also checked the content > > of/usr/local/lib/python2.7/dist-packages/IPython/html/static/components/codemirror/addon/fold, > and that is not an empty folder, so I suppose that all components are > there.) > > The only major difference that I see in the error log of a new and old > notebook is > > Uncaught ReferenceError: nb is not defined notebook.js:2198 > > > Are you on tornado 4? > > No, I was running version 3.1. I have removed that, and now I have > 4.0.2. That didn't change the situation. > > >> The other problem is a rendering issue: when I create a new notebook > (this can be done without problems, and then the results can be loaded > later), and I execute, say, 3*3, I get $$9$$ in the output. This happens, > when I run the the server with the ipython configuration file that I > attached. Is this an issue in the kernel as such, or something in the > sympyprinting extension? > > Probably a bug in JS (cf above) > > > >> These issues don't seem to be related to the browser, or the browser > history: I have cleared the cache, and tried to run the notebooks in chrome > and firefox. I have also run > >> > >>> python setup.py submodule > >> Since that didn't work, I removed the whole git folder, and cloned the > content again. That didn't help either. I still get an > >> > >> /> running css > >> />/ />/ Failed to build css sourcemaps: [Errno 2] No such file or > directory > >> />/ checking package data/ > > You might need less to build the css. But if you clone --recursive, you > shouldn't ned to build anything. > > I have tried this, too, the same problem. > > > > > > > >> error, when trying to build. Last time Matthias suggested python > setup.py submodule, and that solved the problem, but not now. > >> > >> What I also noticed is that the old notebooks have the header > >> > >> { > >> "metadata": { > >> "name": "", > >> "signature": > "sha256:166a240702e584a760e43c90f3d1422154ca5aec932554de10f4922aaa1e517a" > >> }, > >> "nbformat": 3, > >> "nbformat_minor": 0, > >> "worksheets": [ > >> > >> > >> while the new ones are bare; they kick out like this: > >> > >> { > >> "cells": [ > >> { > >> "cell_type": "code", > >> "execution_count": 1, > >> "metadata": { > >> "collapsed": false > >> }, > >> "outputs": [ > >> > >> and the metadata is moved to the end of the file, and the "worksheet" > dictionary entry is removed. I wonder, whether some of the problems are > related to a change in the notebook format. > > This is indeed the noes notebook format. > > I though we were using an ordered dict to write file on disk, but order > shouldn't matter. > > But in the new version, the 'worksheets' entry is completely removed, > isn't it? > > Markdown cell with # on the beginning will be converted to heading cell, > the UI to warn about that is not there yet. > > Quite neat! > > Cheers, > Zolt?n > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: From bussonniermatthias at gmail.com Tue Nov 4 07:42:31 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Tue, 4 Nov 2014 13:42:31 +0100 Subject: [IPython-dev] notebook hangs with latest master In-Reply-To: <5458C7E6.1090605@gmail.com> References: <5458A909.8050304@gmail.com> <687F1821-5449-4509-90BA-15F84EE3F66C@gmail.com> <5458C7E6.1090605@gmail.com> Message-ID: Le 4 nov. 2014 ? 13:34, Zolt?n V?r?s a ?crit : > Hi Matthias, > > On 11/04/2014 12:54 PM, Matthias Bussonnier wrote: > > Here is my error log: > > Loaded extension: toc :8888/static/base/js/utils.js:26 > ? toc.js:147 > Default extension for cell metadata editing loaded. default.js:48 > Raw Cell Format toolbar preset loaded. rawcell.js:82 > Slideshow extension for metadata editing loaded. slideshow.js:43 > Ignoring untrusted image/svg+xml output. outputarea.js:518 > Uncaught ReferenceError: nb is not defined notebook.js:2198 This one is weird, what's the line of notebook.js you have line 2198 (and context if possible) > Failed to load resource: the server responded with a status of 404 (Not > Found) http://localhost:8888/static/components/codemirror/lib/codemirror > Uncaught Error: Script error for: > /static/components/codemirror/lib/codemirror > http://requirejs.org/docs/errors.html#scripterror require.js:141 > Failed to load resource: the server responded with a status of 404 (Not > Found) Ok, so this is probably one of the problem, try deactivation the fold code add-on in your custom.js, probably an extension that do not work with code mirror 4. be sure that IPython/html/static/components/codemirror/lib/codemirror.js exist, but you di submodule update so should be fine. > http://localhost:8888/static/components/codemirror/addon/fold/foldcode > Uncaught Error: Script error for: > /static/components/codemirror/addon/fold/foldcode > http://requirejs.org/docs/errors.html#scripterror require.js:141 > GET > http://localhost:8888/static/components/underscore/underscore-min.map > 404 (Not Found) :8888/static/components/underscore/underscore-min.map:1 > GET http://localhost:8888/static/components/backbone/backbone-min.map > 404 (Not Found) :8888/static/components/backbone/backbone-min.map:1 (generally the min.map we don't care about) > > > The two failed GET requests are there even for new notebooks, so that is > probably not the problem. The codemirror errors shouldn't matter either, > because I have proper code highlighting in a new notebook. (Although, I > have noticed that the contents of subfolders in profile/static/custom > are no longer loaded. I have also checked the content > of/usr/local/lib/python2.7/dist-packages/IPython/html/static/components/codemirror/addon/fold, > and that is not an empty folder, so I suppose that all components are > there.) > > The only major difference that I see in the error log of a new and old > notebook is > > Uncaught ReferenceError: nb is not defined notebook.js:2198 > >> Are you on tornado 4? > > No, I was running version 3.1. I have removed that, and now I have > 4.0.2. That didn't change the situation. > >>> The other problem is a rendering issue: when I create a new notebook (this can be done without problems, and then the results can be loaded later), and I execute, say, 3*3, I get $$9$$ in the output. This happens, when I run the the server with the ipython configuration file that I attached. Is this an issue in the kernel as such, or something in the sympyprinting extension? >> Probably a bug in JS (cf above) >> >>> These issues don't seem to be related to the browser, or the browser history: I have cleared the cache, and tried to run the notebooks in chrome and firefox. I have also run >>> >>>> python setup.py submodule >>> Since that didn't work, I removed the whole git folder, and cloned the content again. That didn't help either. I still get an >>> >>> /> running css >>> />/ />/ Failed to build css sourcemaps: [Errno 2] No such file or directory >>> />/ checking package data/ >> You might need less to build the css. But if you clone --recursive, you shouldn't ned to build anything. > > I have tried this, too, the same problem. > > >> >> >>> error, when trying to build. Last time Matthias suggested python setup.py submodule, and that solved the problem, but not now. >>> >>> What I also noticed is that the old notebooks have the header >>> >>> { >>> "metadata": { >>> "name": "", >>> "signature": "sha256:166a240702e584a760e43c90f3d1422154ca5aec932554de10f4922aaa1e517a" >>> }, >>> "nbformat": 3, >>> "nbformat_minor": 0, >>> "worksheets": [ >>> >>> >>> while the new ones are bare; they kick out like this: >>> >>> { >>> "cells": [ >>> { >>> "cell_type": "code", >>> "execution_count": 1, >>> "metadata": { >>> "collapsed": false >>> }, >>> "outputs": [ >>> >>> and the metadata is moved to the end of the file, and the "worksheet" dictionary entry is removed. I wonder, whether some of the problems are related to a change in the notebook format. >> This is indeed the noes notebook format. >> I though we were using an ordered dict to write file on disk, but order shouldn't matter. > > But in the new version, the 'worksheets' entry is completely removed, > isn't it? Yes it is. Hope we can figure this out... -- M From zvoros at gmail.com Tue Nov 4 08:35:31 2014 From: zvoros at gmail.com (=?UTF-8?B?Wm9sdMOhbiBWw7Zyw7Zz?=) Date: Tue, 04 Nov 2014 14:35:31 +0100 Subject: [IPython-dev] notebook hangs with latest master In-Reply-To: References: <5458A909.8050304@gmail.com> <687F1821-5449-4509-90BA-15F84EE3F66C@gmail.com> <5458C7E6.1090605@gmail.com> Message-ID: <5458D623.70101@gmail.com> On 11/04/2014 01:42 PM, Matthias Bussonnier wrote: > > Loaded extension: toc :8888/static/base/js/utils.js:26 > ? toc.js:147 > Default extension for cell metadata editing loaded. default.js:48 > Raw Cell Format toolbar preset loaded. rawcell.js:82 > Slideshow extension for metadata editing loaded. slideshow.js:43 > Ignoring untrusted image/svg+xml output. outputarea.js:518 > Uncaught ReferenceError: nb is not defined notebook.js:2198 > This one is weird, what's the line of notebook.js you have line 2198 (and context if possible) Actually, this is the part of the code that checks for the notebook version if (orig_nbformat !== undefined && nbmodel.nbformat !== orig_nbformat) { var src; if (nb.nbformat > nb.orig_nbformat) { src = " an older notebook format "; } else { src = " a newer notebook format "; } I attach a minimal example that fails loading. > >> Failed to load resource: the server responded with a status of 404 (Not >> Found) http://localhost:8888/static/components/codemirror/lib/codemirror >> Uncaught Error: Script error for: >> /static/components/codemirror/lib/codemirror >> http://requirejs.org/docs/errors.html#scripterror require.js:141 >> Failed to load resource: the server responded with a status of 404 (Not >> Found) > Ok, so this is probably one of the problem, try deactivation the fold code add-on in your custom.js, > probably an extension that do not work with code mirror 4. I have removed the code from profile/static/custom/, and the error is not there any more. OK, so you would classify this as a glitch in codemirror. I will try to check this. I'll send an update, once I have more on this. Cheers, Zolt?n -------------- next part -------------- { "metadata": { "name": "", "signature": "sha256:195a4ce45f51f1ab5469174e4f122ed532819c3cbde75b66354f19a01284366e" }, "nbformat": 3, "nbformat_minor": 0, "worksheets": [ { "cells": [ { "cell_type": "code", "collapsed": false, "input": [ "%%javascript\n", "var p = IPython.notebook.ncells()\n", "alert(p)" ], "language": "python", "metadata": {}, "outputs": [ { "javascript": [ "var p = IPython.notebook.ncells()\n", "alert(p)" ], "metadata": {}, "output_type": "display_data", "text": [ "" ] } ], "prompt_number": 8 }, { "cell_type": "code", "collapsed": false, "input": [], "language": "python", "metadata": {}, "outputs": [] } ], "metadata": {} } ] } From bussonniermatthias at gmail.com Tue Nov 4 08:45:28 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Tue, 4 Nov 2014 14:45:28 +0100 Subject: [IPython-dev] notebook hangs with latest master In-Reply-To: <5458D623.70101@gmail.com> References: <5458A909.8050304@gmail.com> <687F1821-5449-4509-90BA-15F84EE3F66C@gmail.com> <5458C7E6.1090605@gmail.com> <5458D623.70101@gmail.com> Message-ID: <9099B02F-2DB2-4CD6-8774-EA3FDC7DC003@gmail.com> Le 4 nov. 2014 ? 14:35, Zolt?n V?r?s a ?crit : > Actually, this is the part of the code that checks for the notebook version > > if (orig_nbformat !== undefined && nbmodel.nbformat !== orig_nbformat) { > var src; > if (nb.nbformat > nb.orig_nbformat) { > src = " an older notebook format "; > } else { > src = " a newer notebook format "; > } Ok, so it shouldn't be nb. but nbmodel. I guess. Would you like to fix it and submit a PR ? You can add "use strict"; at the top of the main function to be sure it does not reproduce anymore. -- M -------------- next part -------------- An HTML attachment was scrubbed... URL: From zvoros at gmail.com Tue Nov 4 09:00:09 2014 From: zvoros at gmail.com (=?windows-1252?Q?Zolt=E1n_V=F6r=F6s?=) Date: Tue, 04 Nov 2014 15:00:09 +0100 Subject: [IPython-dev] notebook hangs with latest master In-Reply-To: <9099B02F-2DB2-4CD6-8774-EA3FDC7DC003@gmail.com> References: <5458A909.8050304@gmail.com> <687F1821-5449-4509-90BA-15F84EE3F66C@gmail.com> <5458C7E6.1090605@gmail.com> <5458D623.70101@gmail.com> <9099B02F-2DB2-4CD6-8774-EA3FDC7DC003@gmail.com> Message-ID: <5458DBE9.9010606@gmail.com> Hi Matthias, Thanks for the help! Here is the pull request: https://github.com/ipython/ipython/pull/6851 Cheers, Zolt?n On 11/04/2014 02:45 PM, Matthias Bussonnier wrote: > > Le 4 nov. 2014 ? 14:35, Zolt?n V?r?s > a ?crit : > >> Actually, this is the part of the code that checks for the notebook >> version >> >> if (orig_nbformat !== undefined && nbmodel.nbformat !== orig_nbformat) { >> var src; >> if (nb.nbformat > nb.orig_nbformat) { >> src = " an older notebook format "; >> } else { >> src = " a newer notebook format "; >> } > > Ok, so it shouldn't be nb. but nbmodel. I guess. > > Would you like to fix it and submit a PR ? > > You can add "use strict"; at the top of the main function to be sure > it does not reproduce anymore. > > -- > M > > > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev From nick.bollweg at gmail.com Tue Nov 4 10:15:26 2014 From: nick.bollweg at gmail.com (Nicholas Bollweg) Date: Tue, 4 Nov 2014 10:15:26 -0500 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> <2BF0E798-197A-4EF1-A3C2-E28AFF4E5FA3@gmail.com> Message-ID: News from the JSON-LD front: Everything looks pretty good, except for the mime-types. Here's the gist . Here's a JSON-LD playground . Here's the cross-post to public-linked-json. - A term can't be mapped to two things, say cells.0.cell_type to both the @type and nb4:cell_type. I've opted for @type, as I think the goal here is to make a context that makes content more machine-understandable, rather than something that can round-trip back to the original format. - I've manually added a few mime types as lists, but this task is almost impossible, as its completely arbitrary... this means you won't be able to ask graph questions about your base64-encoded application/x-pdf, unless it has been captured. - I really wish there was a way to do get an @id. Change signature from foaf:sha1 to @id is so tempting, but then all the URIs would, again, be off the main namespace. However, since the signature contains some secret salt, it may not be worthwhile, as it's capturing notebook+user... though maybe that is useful. I think the most likely way to understanding the meaning of a notebook fully would require some preprocessing, such as an nbconvert exporter :). Cheers! -------------- next part -------------- An HTML attachment was scrubbed... URL: From wes.turner at gmail.com Tue Nov 4 13:47:34 2014 From: wes.turner at gmail.com (Wes Turner) Date: Tue, 4 Nov 2014 12:47:34 -0600 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> <2BF0E798-197A-4EF1-A3C2-E28AFF4E5FA3@gmail.com> Message-ID: So, if I was to try and add structured data RDF properties, I would add those as keys in the notebook-level metadata object? I'm assuming I could just merge the nbformatv4 @context with an @context that would allow me to abbreviate additional URI predicates as regular JSON object strings. OTTOMH: dcterms:title dcterms:creator lang:en schema:author schema:description Yesterday, I sent an email to the public-vocabs at w3.org (schema.org) list asking what subtype of schema:CreativeWork would most accurately describe an IPython notebook, but have not yet received a response. At the cell-level, it would be cool if we could say, "this is the abstract", " these are figures derived from this dataset with thus URI retrieved from this URL" and "thus is the conclusion" e.g. for PLoS. I've looked at PROV for something like recording an additive journal of @datastep transformations ( https://github.com/pydata/pandas/issues/3402 ) but have nothing like an actual implementation. It would be great to see it. For clinical studies (e.g. RCTs), an ontology of study-control URIs would also be helpful. Thanks again for this @context, this is great work. On Nov 4, 2014 9:15 AM, "Nicholas Bollweg" wrote: > News from the JSON-LD front: Everything looks pretty good, except for the > mime-types. > > Here's the gist . > Here's a JSON-LD playground > . > Here's the cross-post > > to public-linked-json. > > > - A term can't be mapped to two things, say cells.0.cell_type to both > the @type and nb4:cell_type. I've opted for @type, as I think the goal > here is to make a context that makes content more machine-understandable, > rather than something that can round-trip back to the original format. > - I've manually added a few mime types as lists, but this task is > almost impossible, as its completely arbitrary... this means you won't be > able to ask graph questions about your base64-encoded application/x-pdf, > unless it has been captured. > - I really wish there was a way to do get an @id. Change signature from > foaf:sha1 to @id is so tempting, but then all the URIs would, again, > be off the main namespace. However, since the signature contains some > secret salt, it may not be worthwhile, as it's capturing notebook+user... > though maybe that is useful. > > > I think the most likely way to understanding the meaning of a notebook > fully would require some preprocessing, such as an nbconvert exporter :). > > Cheers! > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From ocefpaf at gmail.com Wed Nov 5 07:42:25 2014 From: ocefpaf at gmail.com (Filipe Pires Alvarenga Fernandes) Date: Wed, 5 Nov 2014 09:42:25 -0300 Subject: [IPython-dev] Using IFrame with lastest checkout Message-ID: Hi, I used to display local HTML files (like folium maps generated in the notebook itself) as an iframe in the notebook. However, the latest checkout IPython/Jupyter does not display the map, it opens a download dialog instead. I am wondering if this change is related to: https://github.com/ipython/ipython/pull/6120 if so, I am unsure how to set the right headers['X-Frame-Options'] option. I tried something like " http://stackoverflow.com/questions/25327403/run-ipython-notebook-in-iframe-from-another-domain" with no avail. BTW, the HTML is located at he sub-directory. If it I place it in the same directory as the notebook everything works as expected. Thanks for the attention, -Filipe -------------- next part -------------- An HTML attachment was scrubbed... URL: From aradan.d at gmail.com Wed Nov 5 07:56:20 2014 From: aradan.d at gmail.com (david_co) Date: Wed, 5 Nov 2014 04:56:20 -0800 (PST) Subject: [IPython-dev] Nbconvert syntax coloring Message-ID: <1415192180842-5076679.post@n6.nabble.com> Is there a way to change the default style of the code highlighting (the colouring) of the nbconvert to latex? I was trying to modify the file /usr/local/lib/python2.7/dist-packages/IPython/nbconvert/filters/highlight.py and use something like latex = _pygments_highlight(source, LatexFormatter(style='monokai'), language, metadata) in the Highlight2Latex class. When I convert my ipynb file, I don't get errors, but the style is still the same Regards, David -- View this message in context: http://python.6.x6.nabble.com/Nbconvert-syntax-coloring-tp5076679.html Sent from the IPython - Development mailing list archive at Nabble.com. From antgonza at gmail.com Wed Nov 5 09:16:03 2014 From: antgonza at gmail.com (=?UTF-8?Q?Antonio_Gonz=C3=A1lez_Pe=C3=B1a?=) Date: Wed, 5 Nov 2014 07:16:03 -0700 Subject: [IPython-dev] ipcluster and PBS engines In-Reply-To: References: Message-ID: Thank you for your help. On Mon, Nov 3, 2014 at 1:22 PM, MinRK wrote: > It?s up to your PBS system how it assigns the individual jobs in a job > array. Under normal circumstance, this will be spread across nodes. If it > isn?t you may need to use custom PBS templates to assign jobs how you want, > or submit each bunch of engines separately: > > for i in {1..$NODES}; do > ipcluster engines -n $ENGINES_PER_NODE & > done > > The default job file looks like: > > #!/bin/sh > #PBS -t 1-10 > #PBS -V > #PBS -N ipengine > /usr/bin/python -m IPython.parallel.engine > > -MinRK > > > On Mon, Nov 3, 2014 at 12:11 PM, Antonio Gonz?lez Pe?a > wrote: >> >> Just looking at qstat as I was expecting to have 10 actual jobs >> running. Expanding on that, what I was expecting is that if I have, >> let's say, 32 cores (4 nodes/workers) in all my cluster and I specify >> --n 10 that I have 10 engines and they are distributed between the 4 >> nodes. Is this possible? >> >> On Mon, Nov 3, 2014 at 12:35 PM, MinRK wrote: >> > It should create one PBS job array that starts ten engines. Are you only >> > looking at qstat, or are you looking at the actual number of engines >> > that >> > end up connected via a parallel.Client? >> > >> > -MinRK >> > >> > >> > On Mon, Nov 3, 2014 at 10:17 AM, Antonio Gonz?lez Pe?a >> > >> > wrote: >> >> >> >> Hi, >> >> >> >> Let's say that I have a system with n nodes and m cores and I will >> >> like to start 10 engines using PBS. Note that n*m>10 and all my >> >> filesystems are shared so that's not a problem. My current issue is >> >> that if do: >> >> ipcluster start --n 10 >> >> and/or have in my ipcluster_config.py >> >> c.IPClusterEngines.n = 10 >> >> I always get 1 controller and 1 engine vs. having 1 controller and 10 >> >> engines submitted to PBS. >> >> >> >> Other option is that I'm misunderstanding how this works. >> >> >> >> Any help will be greatly appreciated. >> >> >> >> -- >> >> Antonio >> >> _______________________________________________ >> >> IPython-dev mailing list >> >> IPython-dev at scipy.org >> >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> > >> > >> > >> > _______________________________________________ >> > IPython-dev mailing list >> > IPython-dev at scipy.org >> > http://mail.scipy.org/mailman/listinfo/ipython-dev >> > >> >> >> >> -- >> Antonio >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -- Antonio From prabhu at aero.iitb.ac.in Wed Nov 5 10:32:11 2014 From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran) Date: Wed, 05 Nov 2014 21:02:11 +0530 Subject: [IPython-dev] [ANN] CFP: SciPy India 2014: Dec 5 - 7, IIT Bombay Message-ID: <545A42FB.9070307@aero.iitb.ac.in> Hello, [Apologies for the cross-posting] The CFP and registration for SciPy India 2014 (http://scipy.in) is open. SciPy India 2014 will be held in IIT Bombay between December 5th to December 7th, 2014. Please spread the word! SciPy India is an annual conference on using Python for science and engineering research and education. The conference is currently in its sixth year and provides an opportunity to learn and implement Python in education and research. Call for Papers ================ We look forward to your submissions on the use of Python for scientific computing and education. This includes pedagogy, exploration, modeling and analysis from both applied and developmental perspectives. We welcome contributions from academia as well as industry. For details on the paper submission please see here: http://scipy.in/2014/call-for-proposals/ Important Dates ================ - Call for proposals end: 9th November 2014 We look forward to seeing you at SciPy India. Regards, Prabhu Ramachandran and Jarrod Millman From nick.bollweg at gmail.com Wed Nov 5 10:59:37 2014 From: nick.bollweg at gmail.com (Nicholas Bollweg) Date: Wed, 5 Nov 2014 10:59:37 -0500 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> <2BF0E798-197A-4EF1-A3C2-E28AFF4E5FA3@gmail.com> Message-ID: Assuming @context were actually in /@context in the notebook, it could be explicitly amended by using an array of contexts: > "@context": [ > "http://foo/bar/json.ld", > {" dc":"http://..."} > ] > However, i'm not sure this assumption would be appropriate, as, for example, the format doesn't include an explicit $schema... perhaps, again in an nbconvert plugin. For everything you described, i think you could just add an inline @context to the /metadata or /cells/0/metadata. This will override any other meaning of like-named terms in that object and its children. You can't use all keywords, like @base and @vocab, though. Your use cases are great: true to semantic nerd, I was thinking about exploiting the data for learning about how people use the notebook, rather than telling other exploiters what a given notebook means. With the explicit metadata schema you describe, all the properties about the / metadata or /cells/0/metadata, and not the notebook or cell, but then that was the point of metadata: to keep the "wild west" out of the notebook and cell roots. As to the particular ontologies used: that's a whole other kettle of fish! schema.org, doapl, foaf, dc, skos?!? I see the exposure of this to users being a "Linked Data" nbextension, which in turn can offer different notebook/cell-level metadata annotations. Once the metadata tag UI lands, we'll have a better baseline for what this should be like. -------------- next part -------------- An HTML attachment was scrubbed... URL: From satra at mit.edu Wed Nov 5 11:11:12 2014 From: satra at mit.edu (Satrajit Ghosh) Date: Wed, 5 Nov 2014 11:11:12 -0500 Subject: [IPython-dev] Notebook format "incompatible" changes In-Reply-To: References: <9C255513-1603-4111-8B99-72A93E9BF1B4@gmail.com> <2BF0E798-197A-4EF1-A3C2-E28AFF4E5FA3@gmail.com> Message-ID: hi nicholas and wes, i had echo-ed wes' comments over on the pull-request. i think we should move this to an IPEP and continue the discussion there. i would personally like to see this integrated in the next iteration of the notebook format and will make the document a first-class citizen in the linked data web. as a start it could be something as simple as adding @context at a global level and allowing modification of the context pointer within the notebook ui. cheers, satra On Wed, Nov 5, 2014 at 10:59 AM, Nicholas Bollweg wrote: > Assuming @context were actually in /@context in the notebook, it could be > explicitly amended by using an array of contexts: > >> "@context": [ >> "http://foo/bar/json.ld", >> {" dc":"http://..."} >> ] >> > However, i'm not sure this assumption would be appropriate, as, for > example, the format doesn't include an explicit $schema... perhaps, again > in an nbconvert plugin. > > For everything you described, i think you could just add an inline > @context to the /metadata or /cells/0/metadata. This will override any > other meaning of like-named terms in that object and its children. You > can't use all keywords, like @base and @vocab, though. > > Your use cases are great: true to semantic nerd, I was thinking about > exploiting the data for learning about how people use the notebook, rather > than telling other exploiters what a given notebook means. With the > explicit metadata schema you describe, all the properties about the / > metadata or /cells/0/metadata, and not the notebook or cell, but then > that was the point of metadata: to keep the "wild west" out of the notebook > and cell roots. > > As to the particular ontologies used: that's a whole other kettle of fish! > schema.org, doapl, foaf, dc, skos?!? I see the exposure of this to users > being a "Linked Data" nbextension, which in turn can offer different > notebook/cell-level metadata annotations. Once the metadata tag UI lands, > we'll have a better baseline for what this should be like. > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From dcarrera at gmail.com Thu Nov 6 03:50:02 2014 From: dcarrera at gmail.com (Daniel Carrera) Date: Thu, 6 Nov 2014 09:50:02 +0100 Subject: [IPython-dev] Authentication / Access Control Message-ID: Hello, Last Tuesday I had my first experience using the IPython notebook to in the classroom, and it was great. I teach a brief "Matlab for Astronomers" course and this year I used an IPython notebook with the Octave backend. I think that the next step may be to run a public IPython server where my students make their own notebooks. The problem is that, AFAICT, IPython only offers a single password authentication. So I cannot see an obvious way for each student to have his or her own notebook. I am aware of Sage Math Cloud, and I commend the authors for it, but I think that it is too slow. Does anyone have a suggestion? Has anyone else encountered a similar problem? If so, what did you do? How did you get the notebooks to your students? This isn't a critical feature for me, but I would be interested to know if there are any solutions. Cheers, Daniel. -- When an engineer says that something can't be done, it's a code phrase that means it's not fun to do. -------------- next part -------------- An HTML attachment was scrubbed... URL: From vasco+python at tenner.nl Thu Nov 6 04:36:18 2014 From: vasco+python at tenner.nl (Vasco) Date: Thu, 06 Nov 2014 10:36:18 +0100 Subject: [IPython-dev] Authentication / Access Control In-Reply-To: References: Message-ID: <545B4112.4020403@tenner.nl> Hi Daniel, On 06-11-14 09:50, Daniel Carrera wrote: > Does anyone have a suggestion? Has anyone else encountered a similar > problem? If so, what did you do? How did you get the notebooks to your > students? This isn't a critical feature for me, but I would be > interested to know if there are any solutions. You are indeed not the first one that wants this feature. The last months there is a large buzz around the jupyter hub (https://github.com/jupyter/jupyterhub) project, that allows multiple user to login to a central server and work on their own (ipython)-notebooks. There is also posted a lot about this topic on this mailinglist. The archives you can find here: http://mail.scipy.org/pipermail/ipython-dev/ Probably you can use a search engine to search through the archives. Vasco From dcarrera at gmail.com Thu Nov 6 04:58:12 2014 From: dcarrera at gmail.com (Daniel Carrera) Date: Thu, 6 Nov 2014 10:58:12 +0100 Subject: [IPython-dev] Authentication / Access Control In-Reply-To: <545B4112.4020403@tenner.nl> References: <545B4112.4020403@tenner.nl> Message-ID: On 6 November 2014 10:36, Vasco wrote: > Hi Daniel, > > You are indeed not the first one that wants this feature. The last > months there is a large buzz around the jupyter hub > (https://github.com/jupyter/jupyterhub) project, that allows multiple > user to login to a central server and work on their own > (ipython)-notebooks. > > There is also posted a lot about this topic on this mailinglist. The > archives you can find here: http://mail.scipy.org/pipermail/ipython-dev/ > > Probably you can use a search engine to search through the archives. > > Great! Thanks. I didn't know about Jupyter hub. I found a thread that seems to answer all my questions. Cheers, Daniel -- When an engineer says that something can't be done, it's a code phrase that means it's not fun to do. -------------- next part -------------- An HTML attachment was scrubbed... URL: From wstein at gmail.com Thu Nov 6 09:55:52 2014 From: wstein at gmail.com (William Stein) Date: Thu, 6 Nov 2014 06:55:52 -0800 Subject: [IPython-dev] Authentication / Access Control In-Reply-To: References: Message-ID: On Thu, Nov 6, 2014 at 12:50 AM, Daniel Carrera wrote: > Hello, > > Last Tuesday I had my first experience using the IPython notebook to in the > classroom, and it was great. I teach a brief "Matlab for Astronomers" course > and this year I used an IPython notebook with the Octave backend. > > I think that the next step may be to run a public IPython server where my > students make their own notebooks. The problem is that, AFAICT, IPython only > offers a single password authentication. So I cannot see an obvious way for > each student to have his or her own notebook. I am aware of Sage Math Cloud, > and I commend the authors for it, but I think that it is too slow. I know you're not going to use SageMathCloud, but for the benefit of other people, is there any chance you can tell me which aspects of SageMathCloud you found to be too slow, so that I can prioritize fixing it? There's so many possible things you're referring to, that I don't even know where to begin. E.g., do you mean processor speed, RAM, speed when actually using IPython, time to open an IPython notebook, time to open a project, time to upload a file, time to download a file, etc.? For example, I spent quite a bit of time this week on speeding up time to open an IPython notebook. The design of SageMathCloud involves multiple levels of redundancy and high availability, where code can potentially run in any of several different data centers. I've generally opted for safety and availability over speed in the implementation -- first making sure people always have access to their data, even when servers fail, and only then making that access faster. -- William > > Does anyone have a suggestion? Has anyone else encountered a similar > problem? If so, what did you do? How did you get the notebooks to your > students? This isn't a critical feature for me, but I would be interested to > know if there are any solutions. > > Cheers, > Daniel. > -- > When an engineer says that something can't be done, it's a code phrase that > means it's not fun to do. > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -- William Stein Professor of Mathematics University of Washington http://wstein.org From dcarrera at gmail.com Thu Nov 6 10:16:01 2014 From: dcarrera at gmail.com (Daniel Carrera) Date: Thu, 6 Nov 2014 16:16:01 +0100 Subject: [IPython-dev] Authentication / Access Control In-Reply-To: References: Message-ID: On 6 November 2014 15:55, William Stein wrote: > I know you're not going to use SageMathCloud, but for the benefit of > other people, is there any chance you can tell me which aspects of > SageMathCloud you found to be too slow, so that I can prioritize > fixing it? There's so many possible things you're referring to, that > I don't even know where to begin. E.g., do you mean processor speed, > RAM, speed when actually using IPython, time to open an IPython > notebook, time to open a project, time to upload a file, time to > download a file, etc.? > First of all, thank you for making SageMathCloud. I do think it is a great service. Let's see... I just went to my account and clicked on my project. Then I waited 15s for the project to load. After that, navigating through the directories was fast. Then I clicked on a notebook and that took 19s to load. The behaviour of the notebook itself varies a lot. Right now it is fast and it feels nice. But at other times when I press "Shift-Enter" I would just wait and nothing would happen. Some times it would eventually work, and some times I would have to go back to the cell and press "Shift-Enter" again. A couple of times I decided to restart the kernel. > The design of SageMathCloud involves multiple levels of redundancy and > high availability, where code can potentially run in any of several > different data centers. I've generally opted for safety and > availability over speed in the implementation -- first making sure > people always have access to their data, even when servers fail, and > only then making that access faster. > > -- William > That is quite reasonable. Slow and working is always better than fast and broken. Cheers, Daniel. -------------- next part -------------- An HTML attachment was scrubbed... URL: From lucienboillod at gmail.com Thu Nov 6 11:12:41 2014 From: lucienboillod at gmail.com (Lucien Boillod) Date: Thu, 6 Nov 2014 17:12:41 +0100 Subject: [IPython-dev] conflict between d3 keyCode event and IPython notebook In-Reply-To: References: Message-ID: <1EF9D4AC-08D3-405F-816B-F3D764EB5290@gmail.com> Hi all, a little ping, if someone know how to get around this keycode conflict. Thank you, > Le 4 nov. 2014 ? 12:34, Lucien Boillod a ?crit : > > Hello all, > > I have a problem to use keyCode event in D3.js into a notebook, indeed there is some key that are associated with some notebooks actions, that I can?t use in D3. > To show clearly my problem I made a simple example. As you can see delete will show the console log, but L will only execute the associate notebook action. > > Cheers, > Lucien > > From ellisonbg at gmail.com Thu Nov 6 11:36:45 2014 From: ellisonbg at gmail.com (Brian Granger) Date: Thu, 6 Nov 2014 08:36:45 -0800 Subject: [IPython-dev] conflict between d3 keyCode event and IPython notebook In-Reply-To: <1EF9D4AC-08D3-405F-816B-F3D764EB5290@gmail.com> References: <1EF9D4AC-08D3-405F-816B-F3D764EB5290@gmail.com> Message-ID: You need to call KeyboardManager.register_events(your_element) to disable our keyboard handling when you element is focused. Have a look in our source code for how this works... On Thu, Nov 6, 2014 at 8:12 AM, Lucien Boillod wrote: > Hi all, > > a little ping, if someone know how to get around this keycode conflict. > > Thank you, >> Le 4 nov. 2014 ? 12:34, Lucien Boillod a ?crit : >> >> Hello all, >> >> I have a problem to use keyCode event in D3.js into a notebook, indeed there is some key that are associated with some notebooks actions, that I can?t use in D3. >> To show clearly my problem I made a simple example. As you can see delete will show the console log, but L will only execute the associate notebook action. >> >> Cheers, >> Lucien >> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev -- Brian E. Granger Cal Poly State University, San Luis Obispo @ellisonbg on Twitter and GitHub bgranger at calpoly.edu and ellisonbg at gmail.com From roalexan at microsoft.com Thu Nov 6 12:26:00 2014 From: roalexan at microsoft.com (Robert Alexander) Date: Thu, 6 Nov 2014 17:26:00 +0000 Subject: [IPython-dev] Using IPython API to create notebook Message-ID: <1415294753076.82878@microsoft.com> I'm trying to use the IPython API (see: http://ipython.org/ipython-doc/2/api/generated/IPython.html.services.notebooks.nbmanager.html#module-IPython.html.services.notebooks.nbmanager) to programmatically create a notebook in IPython 2.3.0, but hitting some problems. Here are the steps that I tried: ipython --version 2.3.0 python Python 2.7.6 (default, Mar 22 2014, 22:59:56) [GCC 4.8.2] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> import IPython >>> IPython.html.services.notebooks.nbmanager.create_notebook('test') Traceback (most recent call last): File "", line 1, in AttributeError: 'module' object has no attribute 'html' It looks like the module is not being properly imported - any idea what I'm doing wrong here? BTW, it works fine for another module I tested, via: >>> import IPython >>> print IPython.lib.passwd('test') sha1:ac5afaa5ecd8:f8c76f014f3ea016b947c86af6ce176da249688d -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Thu Nov 6 12:34:03 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Thu, 6 Nov 2014 09:34:03 -0800 Subject: [IPython-dev] Using IPython API to create notebook In-Reply-To: <1415294753076.82878@microsoft.com> References: <1415294753076.82878@microsoft.com> Message-ID: Hi Robert On 6 November 2014 09:26, Robert Alexander wrote: > I'm trying to use the IPython API (see: > > > http://ipython.org/ipython-doc/2/api/generated/IPython.html.services.notebooks.nbmanager.html#module-IPython.html.services.notebooks.nbmanager) > to programmatically create a notebook in IPython 2.3.0, but hitting some > problems. Here are the steps that I tried: > > The immediate problem is that importing a Python package does not automatically import modules within that. To import that module, you would do something like: from IPython.html.services.notebooks import nbmanager However, creating notebooks programmatically is easier with nbformat. Create a notebook using new_notebook(): http://ipython.org/ipython-doc/2/api/generated/IPython.nbformat.v3.nbbase.html#module-IPython.nbformat.v3.nbbase And save it using write(): http://ipython.org/ipython-doc/2/api/generated/IPython.nbformat.v3.nbbase.html#module-IPython.nbformat.v3.nbbase Best wishes, Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From roalexan at microsoft.com Thu Nov 6 13:17:25 2014 From: roalexan at microsoft.com (Robert Alexander) Date: Thu, 6 Nov 2014 18:17:25 +0000 Subject: [IPython-dev] Using IPython API to create notebook In-Reply-To: References: <1415294753076.82878@microsoft.com>, Message-ID: <1415297844533.97924@microsoft.com> Thanks (and apologies for my newness to Python) - the import problems are resolved. BTW, for others, I also stumbled upon this link, http://nbviewer.ipython.org/gist/fperez/9716279, which has some good sample code to create and save a new notebook (using the nbformat module that you mentioned). ________________________________ From: ipython-dev-bounces at scipy.org on behalf of Thomas Kluyver Sent: Thursday, November 6, 2014 12:34 PM To: IPython developers list Subject: Re: [IPython-dev] Using IPython API to create notebook Hi Robert On 6 November 2014 09:26, Robert Alexander > wrote: I'm trying to use the IPython API (see: http://ipython.org/ipython-doc/2/api/generated/IPython.html.services.notebooks.nbmanager.html#module-IPython.html.services.notebooks.nbmanager) to programmatically create a notebook in IPython 2.3.0, but hitting some problems. Here are the steps that I tried: The immediate problem is that importing a Python package does not automatically import modules within that. To import that module, you would do something like: from IPython.html.services.notebooks import nbmanager However, creating notebooks programmatically is easier with nbformat. Create a notebook using new_notebook(): http://ipython.org/ipython-doc/2/api/generated/IPython.nbformat.v3.nbbase.html#module-IPython.nbformat.v3.nbbase And save it using write(): http://ipython.org/ipython-doc/2/api/generated/IPython.nbformat.v3.nbbase.html#module-IPython.nbformat.v3.nbbase Best wishes, Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From doug.blank at gmail.com Thu Nov 6 13:57:25 2014 From: doug.blank at gmail.com (Doug Blank) Date: Thu, 6 Nov 2014 13:57:25 -0500 Subject: [IPython-dev] IPython notebook, tmpnb.org in Nature Blog Message-ID: Congrats IPython team on the write-up and working temporary notebook mentioned in yesterday's Nature Blog: http://www.nature.com/news/interactive-notebooks-sharing-the-code-1.16261 A few people must have worked hard to get that put together at Rackspace. Well done! -Doug -------------- next part -------------- An HTML attachment was scrubbed... URL: From rgbkrk at gmail.com Thu Nov 6 15:00:36 2014 From: rgbkrk at gmail.com (Kyle Kelley) Date: Thu, 6 Nov 2014 13:00:36 -0700 Subject: [IPython-dev] IPython notebook, tmpnb.org in Nature Blog In-Reply-To: References: Message-ID: Thanks Doug! Definitely a lot of help from Min RK and Brian Granger too. The other helpful Rackspace parties are Ash Wilson, Dana Bauer, and Jesse Noller. Major kudos. On Thursday, November 6, 2014, Doug Blank wrote: > Congrats IPython team on the write-up and working temporary notebook > mentioned in yesterday's Nature Blog: > > http://www.nature.com/news/interactive-notebooks-sharing-the-code-1.16261 > > A few people must have worked hard to get that put together at Rackspace. > Well done! > > -Doug > > -- Kyle Kelley (@rgbkrk ; http://lambdaops.com) -------------- next part -------------- An HTML attachment was scrubbed... URL: From rgbkrk at gmail.com Thu Nov 6 15:08:54 2014 From: rgbkrk at gmail.com (Kyle Kelley) Date: Thu, 6 Nov 2014 13:08:54 -0700 Subject: [IPython-dev] IPython notebook, tmpnb.org in Nature Blog In-Reply-To: References: Message-ID: How could I forget Aron Ahmadia and Stefan Van der Walt?! At least everyone is noted in the notebook ;) Big thanks to my wife who puts up with me working on IPython/Jupyter at all hours. :) "just one more PR honey" "someone needs help!" Luckily we took some vacation last week. -- Kyle On Thursday, November 6, 2014, Kyle Kelley wrote: > Thanks Doug! Definitely a lot of help from Min RK and Brian Granger too. > > The other helpful Rackspace parties are Ash Wilson, Dana Bauer, and Jesse > Noller. Major kudos. > > On Thursday, November 6, 2014, Doug Blank > wrote: > >> Congrats IPython team on the write-up and working temporary notebook >> mentioned in yesterday's Nature Blog: >> >> http://www.nature.com/news/interactive-notebooks-sharing-the-code-1.16261 >> >> A few people must have worked hard to get that put together at Rackspace. >> Well done! >> >> -Doug >> >> > > -- > Kyle Kelley (@rgbkrk ; http://lambdaops.com) > > -- Kyle Kelley (@rgbkrk ; http://lambdaops.com) -------------- next part -------------- An HTML attachment was scrubbed... URL: From ASchneiderman at asha.org Thu Nov 6 16:57:24 2014 From: ASchneiderman at asha.org (Anders Schneiderman) Date: Thu, 6 Nov 2014 16:57:24 -0500 Subject: [IPython-dev] Help Figuring out Why Horizontal Three Column Widget Container Layout Isn't Working Message-ID: <93268EE5694C0A4BBB888C2EEFEEEC22B3EFC8AEE6@EXCH2008.hq.asha.org> I'm trying to create a horizontal three column widget container layout and it's not working. The alignment examples from the widget documentation work just fine. But when I tried to use the hbox class on my widget containers, it didn't seem to do anything; they are stacked on top of each other. Here's a simplified version of what I'm trying to do: from IPython.html import widgets from IPython.display import display input_columns = widgets.ContainerWidget() display(input_columns) input_columns.add_class("hbox") subnav = widgets.ContainerWidget(children=[widgets.HTMLWidget(value ="Subnav")]) input_box = widgets.ContainerWidget(children=[widgets.HTMLWidget(value = "This is where the input will go")]) tutorial = widgets.ContainerWidget(children=[widgets.HTMLWidget(value ="Tutorial")]) input_columns.children =(subnav,input_box,tutorial) What am I doing wrong? I've tried this both in Google Chrome and Firefox and it's not working either. Also, should twitter bootstrap's grid system work in IPython Notebook? I tried using that, and it didn't seem to do anything. Thanks! Anders -------------- next part -------------- An HTML attachment was scrubbed... URL: From yw5aj at virginia.edu Wed Nov 12 11:54:15 2014 From: yw5aj at virginia.edu (Yuxiang Wang) Date: Wed, 12 Nov 2014 11:54:15 -0500 Subject: [IPython-dev] Some sighs on using ipynb to make slides Message-ID: First of all, sorry if this is not the right place to post this kind of threads - I heard that ipython-user is being gradually deprecated and moved to ipython-dev. (Warning: random whining and sighs below) First, I use Python in my research and I really love this language. A million thanks to IPython, that life is much much easier with python, compared to the idle console. I am learning R recently and noticed the very nice R markdown feature for reproducible research. Then I wondered what is the equivalent in Python, which is actually the IPython Notebook (I never paid attention on the "reproducible" part of ipynb before!) Then, I realized the philosophy seems to be different: while R Studio seems to be making the wheels for R presentation themselves, Python works more as a glue and connected with reveal.js. I really love this philosophy, however the drawback is that the users would have to take care of the dependency. Conda was born for this purpose (non-Python dependencies), but still, it has not supported pandoc + reveal.js + miktex very well yet. On the other hand, R is really friendly even in Windows, that if you use install.packages(), you can solve most of the issue... They have a installr package available in Windows, to install pandoc + miktex for making slides and PDF etc. I happened to have tex + pandoc on my computer but just needed to install reveal.js and (optionally) slideshow support in ipynb, so it is all good; however, this procedure really seems a non-trivial friction to prevent beginners from using this nice feature... PS: a side question, are there any plans to merge ipynb slideshow feature to the major release? -Shawn From takowl at gmail.com Wed Nov 12 19:12:26 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Wed, 12 Nov 2014 16:12:26 -0800 Subject: [IPython-dev] Some sighs on using ipynb to make slides In-Reply-To: References: Message-ID: Hi Shawn, I think it's OK to post this kind of thing here. In the next release, we'll be using a Python package (mistune) to convert Markdown to HTML, so pandoc won't be a requirement (it will still be required if you want to convert to LaTeX - that's not something we can practically solve). So hopefully at that point you should be able to make slides with just Python and (included) JS stuff. There are actually two slideshow features: nbconvert can export a static copy of the notebook as a reveal.js presentation - this is already part of IPython - and you can use Damian's live_reveal [1] extension to display live, editable notebooks as slides. We're supportive of the latter, but I don't think we'll merge it in - we're trying to slim the project down and make extensions easier to use, not assimilate lots of things. [1] https://github.com/damianavila/live_reveal Best wishes, Thomas On 12 November 2014 08:54, Yuxiang Wang wrote: > First of all, sorry if this is not the right place to post this kind > of threads - I heard that ipython-user is being gradually deprecated > and moved to ipython-dev. > > (Warning: random whining and sighs below) > > First, I use Python in my research and I really love this language. A > million thanks to IPython, that life is much much easier with python, > compared to the idle console. > > I am learning R recently and noticed the very nice R markdown feature > for reproducible research. Then I wondered what is the equivalent in > Python, which is actually the IPython Notebook (I never paid attention > on the "reproducible" part of ipynb before!) > > Then, I realized the philosophy seems to be different: while R Studio > seems to be making the wheels for R presentation themselves, Python > works more as a glue and connected with reveal.js. I really love this > philosophy, however the drawback is that the users would have to take > care of the dependency. Conda was born for this purpose (non-Python > dependencies), but still, it has not supported pandoc + reveal.js + > miktex very well yet. On the other hand, R is really friendly even in > Windows, that if you use install.packages(), you can solve most of the > issue... They have a installr package available in Windows, to install > pandoc + miktex for making slides and PDF etc. > > I happened to have tex + pandoc on my computer but just needed to > install reveal.js and (optionally) slideshow support in ipynb, so it > is all good; however, this procedure really seems a non-trivial > friction to prevent beginners from using this nice feature... > > PS: a side question, are there any plans to merge ipynb slideshow > feature to the major release? > > -Shawn > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: From dongta.hds at gmail.com Sun Nov 9 22:45:42 2014 From: dongta.hds at gmail.com (Dong Ta) Date: Sun, 9 Nov 2014 19:45:42 -0800 Subject: [IPython-dev] Trouble installing IPython3 into dist-packages Message-ID: I followed the installation instruction on https://github.com/ipython/ipython, but used pip3 install -e ".[notebook]" (i.e. without --user). However, it doesn't install IPython into the system dist-packages directory (/usr/lib/python3.4/dist-packages). If I run IPython and try to locate the module file, it points to my user's IPython directory (/home/dongta/ipython/IPython). Why is that and how can I install into the system dist-packages so that other users can use it? Thank you very much. DT -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Wed Nov 12 19:27:49 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Wed, 12 Nov 2014 16:27:49 -0800 Subject: [IPython-dev] Trouble installing IPython3 into dist-packages In-Reply-To: References: Message-ID: On 9 November 2014 19:45, Dong Ta wrote: > However, it doesn't install IPython into the system dist-packages > directory (/usr/lib/python3.4/dist-packages). If I run IPython and try to > locate the module file, it points to my user's IPython directory > (/home/dongta/ipython/IPython). > > Why is that and how can I install into the system dist-packages so that > other users can use it? > I would guess that -e effectively implies --user. Try installing without that. You'll probably also need sudo to install it systemwide: sudo pip3 install ".[notebook]" Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From ijpulidos at riseup.net Wed Nov 12 19:26:16 2014 From: ijpulidos at riseup.net (=?UTF-8?B?SXbDoW4gUHVsaWRv?=) Date: Wed, 12 Nov 2014 19:26:16 -0500 Subject: [IPython-dev] Some sighs on using ipynb to make slides In-Reply-To: References: Message-ID: <5463FAA8.4000101@riseup.net> El 12/11/14 a las 19:12, Thomas Kluyver escibi?: > Hi Shawn, > > I think it's OK to post this kind of thing here. > > In the next release, we'll be using a Python package (mistune) to > convert Markdown to HTML, so pandoc won't be a requirement (it will > still be required if you want to convert to LaTeX - that's not > something we can practically solve). So hopefully at that point you > should be able to make slides with just Python and (included) JS stuff. > > There are actually two slideshow features: nbconvert can export a > static copy of the notebook as a reveal.js presentation - this is > already part of IPython - and you can use Damian's live_reveal [1] > extension to display live, editable notebooks as slides. We're > supportive of the latter, but I don't think we'll merge it in - we're > trying to slim the project down and make extensions easier to use, not > assimilate lots of things. > I agree with Thomas, probably the best way to go for you if you want less configuration needed by the user is the live_reveal extension from Damian. I've used it and it's actually pretty good, IMHO. There's a notebook in the github site that installs the extension automatically, after that you just need to run it with the correct profile and that's it. Hope this works. > [1] https://github.com/damianavila/live_reveal > > Best wishes, > Thomas > > On 12 November 2014 08:54, Yuxiang Wang > wrote: > > First of all, sorry if this is not the right place to post this kind > of threads - I heard that ipython-user is being gradually deprecated > and moved to ipython-dev. > > (Warning: random whining and sighs below) > > First, I use Python in my research and I really love this language. A > million thanks to IPython, that life is much much easier with python, > compared to the idle console. > > I am learning R recently and noticed the very nice R markdown feature > for reproducible research. Then I wondered what is the equivalent in > Python, which is actually the IPython Notebook (I never paid attention > on the "reproducible" part of ipynb before!) > > Then, I realized the philosophy seems to be different: while R Studio > seems to be making the wheels for R presentation themselves, Python > works more as a glue and connected with reveal.js. I really love this > philosophy, however the drawback is that the users would have to take > care of the dependency. Conda was born for this purpose (non-Python > dependencies), but still, it has not supported pandoc + reveal.js + > miktex very well yet. On the other hand, R is really friendly even in > Windows, that if you use install.packages(), you can solve most of the > issue... They have a installr package available in Windows, to install > pandoc + miktex for making slides and PDF etc. > > I happened to have tex + pandoc on my computer but just needed to > install reveal.js and (optionally) slideshow support in ipynb, so it > is all good; however, this procedure really seems a non-trivial > friction to prevent beginners from using this nice feature... > > PS: a side question, are there any plans to merge ipynb slideshow > feature to the major release? > > -Shawn > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev -------------- next part -------------- An HTML attachment was scrubbed... URL: From cyrille.rossant at gmail.com Wed Nov 12 05:53:25 2014 From: cyrille.rossant at gmail.com (Cyrille Rossant) Date: Wed, 12 Nov 2014 11:53:25 +0100 Subject: [IPython-dev] WebGL widget in the IPython notebook Message-ID: Hi, Some progress has been made on WebGL integration in the IPython notebook, in the context of the Vispy project; see < https://groups.google.com/forum/#!topic/vispy/pd9ue9EQHlg>. This might interest those of you who want to visualize data interactively in the notebook. All the best, Cyrille -------------- next part -------------- An HTML attachment was scrubbed... URL: From dongta.hds at gmail.com Wed Nov 12 19:53:23 2014 From: dongta.hds at gmail.com (Dong Ta) Date: Wed, 12 Nov 2014 16:53:23 -0800 Subject: [IPython-dev] Trouble installing IPython3 into dist-packages In-Reply-To: References: Message-ID: Without -e, the command fails. Here's the full stdout/stderr: Downloading/unpacking .[notebook] Could not find any downloads that satisfy the requirement .[notebook] Cleaning up... No distributions at all found for .[notebook] Storing debug log for failure in /root/.pip/pip.log And here's the tail of /root/.pip/pip.log Skipping link https://pypi.python.org/simple/zyzz (from https://pypi.python.org/simple/./); not a file Skipping link https://pypi.python.org/simple/zzhmodule (from https://pypi.python.org/simple/./); not a file Skipping link https://pypi.python.org/simple/zzz (from https://pypi.python.org/simple/./); not a file Skipping link https://pypi.python.org/simple/zzzzzzzzz (from https://pypi.python.org/simple/./); not a file Could not find any downloads that satisfy the requirement .[notebook] Cleaning up... Removing temporary dir /tmp/pip_build_root... No distributions at all found for .[notebook] Exception information: Traceback (most recent call last): File "/usr/lib/python3/dist-packages/pip/basecommand.py", line 122, in main status = self.run(options, args) File "/usr/lib/python3/dist-packages/pip/commands/install.py", line 278, in run requirement_set.prepare_files(finder, force_root_egg_info=self.bundle, bundle=self.bundle) File "/usr/lib/python3/dist-packages/pip/req.py", line 1177, in prepare_files url = finder.find_requirement(req_to_install, upgrade=self.upgrade) File "/usr/lib/python3/dist-packages/pip/index.py", line 277, in find_requirement raise DistributionNotFound('No distributions at all found for %s' % req) pip.exceptions.DistributionNotFound: No distributions at all found for .[notebook] If I do pip3 install . it works; but then the original problem remains :(. DT On Wed, Nov 12, 2014 at 4:27 PM, Thomas Kluyver wrote: > On 9 November 2014 19:45, Dong Ta wrote: > >> However, it doesn't install IPython into the system dist-packages >> directory (/usr/lib/python3.4/dist-packages). If I run IPython and try to >> locate the module file, it points to my user's IPython directory >> (/home/dongta/ipython/IPython). >> >> Why is that and how can I install into the system dist-packages so that >> other users can use it? >> > > I would guess that -e effectively implies --user. Try installing without > that. You'll probably also need sudo to install it systemwide: > > sudo pip3 install ".[notebook]" > > Thomas > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Wed Nov 12 20:06:56 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Wed, 12 Nov 2014 17:06:56 -0800 Subject: [IPython-dev] Trouble installing IPython3 into dist-packages In-Reply-To: References: Message-ID: On 12 November 2014 16:53, Dong Ta wrote: > Without -e, the command fails. Here's the full stdout/stderr: I am told that the official spelling of pip install ".[notebook]" is: pip install file://$PWD#egg=ipython[notebook] *sigh* - Python packaging tools never stop finding new ways to suck. Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From yw5aj at virginia.edu Wed Nov 12 20:19:53 2014 From: yw5aj at virginia.edu (Yuxiang Wang) Date: Wed, 12 Nov 2014 20:19:53 -0500 Subject: [IPython-dev] Some sighs on using ipynb to make slides In-Reply-To: References: Message-ID: 0) Thank you Thomas and Ivan! Those really help. And it's great to know that we are moving to mistune! 1) What I was trying to express is actually now, Python packages generally (more than IPython) rely a lot on non-python dependencies, thus making it a pain for users (especially entry-level) to configure the environment. For example, if I'd like to use IPython Notebook to make a slide, I would need to google the solutions and install pandoc, miktex, reveal.js, and of course git on the way. Once I figured it out it sounds cool to me, but this is a major factor to lower Python's popularity as a scientific tool... I feel that most people using R is because R Studio did a lot of these things for them. 2) But - as that said, it's more like "pure whining" because it takes non-trivial effort to change this situation to such a huge ecosystem. And the situation is changing - I think conda is there to solve these issues (configuring Cython on 64-bit windows used to be such a pain, but now it's much easier!). :) 3) Detailed questions - you mentioned the "(included) JS stuff", so are we going to have an IPython distribution with reveal.js included? Thanks again for all your responses! And thanks for approving me to post stuff like this here. A community is more than a technical Q&A place - there's emotional support to know that I'm not all alone, ha! -Shawn On Wed, Nov 12, 2014 at 7:12 PM, Thomas Kluyver wrote: > Hi Shawn, > > I think it's OK to post this kind of thing here. > > In the next release, we'll be using a Python package (mistune) to convert > Markdown to HTML, so pandoc won't be a requirement (it will still be > required if you want to convert to LaTeX - that's not something we can > practically solve). So hopefully at that point you should be able to make > slides with just Python and (included) JS stuff. > > There are actually two slideshow features: nbconvert can export a static > copy of the notebook as a reveal.js presentation - this is already part of > IPython - and you can use Damian's live_reveal [1] extension to display > live, editable notebooks as slides. We're supportive of the latter, but I > don't think we'll merge it in - we're trying to slim the project down and > make extensions easier to use, not assimilate lots of things. > > [1] https://github.com/damianavila/live_reveal > > Best wishes, > Thomas > > On 12 November 2014 08:54, Yuxiang Wang wrote: >> >> First of all, sorry if this is not the right place to post this kind >> of threads - I heard that ipython-user is being gradually deprecated >> and moved to ipython-dev. >> >> (Warning: random whining and sighs below) >> >> First, I use Python in my research and I really love this language. A >> million thanks to IPython, that life is much much easier with python, >> compared to the idle console. >> >> I am learning R recently and noticed the very nice R markdown feature >> for reproducible research. Then I wondered what is the equivalent in >> Python, which is actually the IPython Notebook (I never paid attention >> on the "reproducible" part of ipynb before!) >> >> Then, I realized the philosophy seems to be different: while R Studio >> seems to be making the wheels for R presentation themselves, Python >> works more as a glue and connected with reveal.js. I really love this >> philosophy, however the drawback is that the users would have to take >> care of the dependency. Conda was born for this purpose (non-Python >> dependencies), but still, it has not supported pandoc + reveal.js + >> miktex very well yet. On the other hand, R is really friendly even in >> Windows, that if you use install.packages(), you can solve most of the >> issue... They have a installr package available in Windows, to install >> pandoc + miktex for making slides and PDF etc. >> >> I happened to have tex + pandoc on my computer but just needed to >> install reveal.js and (optionally) slideshow support in ipynb, so it >> is all good; however, this procedure really seems a non-trivial >> friction to prevent beginners from using this nice feature... >> >> PS: a side question, are there any plans to merge ipynb slideshow >> feature to the major release? >> >> -Shawn >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -- Yuxiang "Shawn" Wang Gerling Research Lab University of Virginia yw5aj at virginia.edu +1 (434) 284-0836 https://sites.google.com/a/virginia.edu/yw5aj/ From benjaminrk at gmail.com Mon Nov 10 23:16:32 2014 From: benjaminrk at gmail.com (MinRK) Date: Mon, 10 Nov 2014 20:16:32 -0800 Subject: [IPython-dev] custom Notebook (Contents) Managers and IPython 3.0 Message-ID: Hello all, As we are getting closer to a release of IPython 3.0, one of the major changes is getting nearer its final API. The NotebookManager object that served notebooks in IPython 2.0 is now a generic ContentsManager object that can serve plain files and directories as well. The changes to the REST API are described in IPEP 27 . The gist for authors of NotebookManagers who want to update to ContentsManagers should be covered at the end of the IPEP . The main points: - the addition of non-notebook models - removal of _notebook from various method names - the removal of separate path and name arguments, in favor of a single path argument containing the full path, including the filename. If you are a user or author of a custom NotebookManager, and would like a hand updating it to a ContentsManager, let us know. -MinRK ? -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Wed Nov 12 20:29:26 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Wed, 12 Nov 2014 17:29:26 -0800 Subject: [IPython-dev] Some sighs on using ipynb to make slides In-Reply-To: References: Message-ID: On 12 November 2014 17:19, Yuxiang Wang wrote: > 1) What I was trying to express is actually now, Python packages > generally (more than IPython) rely a lot on non-python dependencies, > thus making it a pain for users (especially entry-level) to configure > the environment. For example, if I'd like to use IPython Notebook to > make a slide, I would need to google the solutions and install pandoc, > miktex, reveal.js, and of course git on the way. Once I figured it out > it sounds cool to me, but this is a major factor to lower Python's > popularity as a scientific tool... I feel that most people using R is > because R Studio did a lot of these things for them. > > 2) But - as that said, it's more like "pure whining" because it takes > non-trivial effort to change this situation to such a huge ecosystem. > And the situation is changing - I think conda is there to solve these > issues (configuring Cython on 64-bit windows used to be such a pain, > but now it's much easier!). :) > Yep, I would say conda is definitely the best approach here. There's no reason in principle you couldn't make a conda package for something like pandoc (though Min tried it, and doing it properly, with the Haskell platform as a build dependency, gets interesting). Conda itself is open source, and there's a big repository of recipes for it ( https://github.com/conda/conda-recipes/ ), so if you want to help improve the situation, that's where I'd start. > 3) Detailed questions - you mentioned the "(included) JS stuff", so > are we going to have an IPython distribution with reveal.js included? > I think we have considered it, but it's not currently in our plans. Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From yw5aj at virginia.edu Wed Nov 12 20:49:57 2014 From: yw5aj at virginia.edu (Yuxiang Wang) Date: Wed, 12 Nov 2014 20:49:57 -0500 Subject: [IPython-dev] Some sighs on using ipynb to make slides In-Reply-To: References: Message-ID: Thomas, Thank you for pointing me to the conda-recipes! I didn't realize that it was open source (I have no idea why I'd think that way). I'll see what I can do, probably that can also solve the installation of reveal.js. Thanks again! -Shawn On Wed, Nov 12, 2014 at 8:29 PM, Thomas Kluyver wrote: > On 12 November 2014 17:19, Yuxiang Wang wrote: >> >> 1) What I was trying to express is actually now, Python packages >> generally (more than IPython) rely a lot on non-python dependencies, >> thus making it a pain for users (especially entry-level) to configure >> the environment. For example, if I'd like to use IPython Notebook to >> make a slide, I would need to google the solutions and install pandoc, >> miktex, reveal.js, and of course git on the way. Once I figured it out >> it sounds cool to me, but this is a major factor to lower Python's >> popularity as a scientific tool... I feel that most people using R is >> because R Studio did a lot of these things for them. >> >> 2) But - as that said, it's more like "pure whining" because it takes >> non-trivial effort to change this situation to such a huge ecosystem. >> And the situation is changing - I think conda is there to solve these >> issues (configuring Cython on 64-bit windows used to be such a pain, >> but now it's much easier!). :) > > > Yep, I would say conda is definitely the best approach here. There's no > reason in principle you couldn't make a conda package for something like > pandoc (though Min tried it, and doing it properly, with the Haskell > platform as a build dependency, gets interesting). Conda itself is open > source, and there's a big repository of recipes for it > (https://github.com/conda/conda-recipes/ ), so if you want to help improve > the situation, that's where I'd start. > >> >> 3) Detailed questions - you mentioned the "(included) JS stuff", so >> are we going to have an IPython distribution with reveal.js included? > > > I think we have considered it, but it's not currently in our plans. > > Thomas > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -- Yuxiang "Shawn" Wang Gerling Research Lab University of Virginia yw5aj at virginia.edu +1 (434) 284-0836 https://sites.google.com/a/virginia.edu/yw5aj/ From wstein at gmail.com Sun Nov 9 22:55:54 2014 From: wstein at gmail.com (William Stein) Date: Sun, 9 Nov 2014 19:55:54 -0800 Subject: [IPython-dev] Nov 17 Message-ID: Hi, I'll be in Berkeley Monday, Nov 17, and would be interested in meeting or coding with some IPython developers if you guys will be around.... Maybe we could talk about synchronization or embedding IPython notebooks in other pages without using an Iframe. William -- William Stein Professor of Mathematics University of Washington http://wstein.org From dongta.hds at gmail.com Thu Nov 13 01:07:15 2014 From: dongta.hds at gmail.com (Dong Ta) Date: Wed, 12 Nov 2014 22:07:15 -0800 Subject: [IPython-dev] Trouble installing IPython3 into dist-packages In-Reply-To: References: Message-ID: Thank you. I've just tried sudo pip3 install file://$PWD#egg=ipython[notebook] --upgrade --force-reinstall which installed successfully. However, the original problem remains, i.e. it doesn't install IPython package into /usr/lib/python3/dist-packages In [1]: from IPython import core In [2]: ??core Type: module String form: File: /home/dongta/ipython/IPython/core/__init__.py Docstring: although the entry point is installed correctly, i.e. ~$ which ipython /usr/local/bin/ipython On Wed, Nov 12, 2014 at 5:06 PM, Thomas Kluyver wrote: > On 12 November 2014 16:53, Dong Ta wrote: > >> Without -e, the command fails. Here's the full stdout/stderr: > > > I am told that the official spelling of pip install ".[notebook]" is: > > pip install file://$PWD#egg=ipython[notebook] > > *sigh* - Python packaging tools never stop finding new ways to suck. > > Thomas > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From benjaminrk at gmail.com Thu Nov 13 01:12:16 2014 From: benjaminrk at gmail.com (MinRK) Date: Wed, 12 Nov 2014 22:12:16 -0800 Subject: [IPython-dev] Nov 17 In-Reply-To: References: Message-ID: Sorry, our ML server has been down for a few days. I think Fernando may still be out of town Monday, but Thomas and I should be on campus and available. -MinRK On Sun, Nov 9, 2014 at 7:55 PM, William Stein wrote: > Hi, > > I'll be in Berkeley Monday, Nov 17, and would be interested in meeting > or coding with some IPython developers if you guys will be around.... > Maybe we could talk about synchronization or embedding IPython > notebooks in other pages without using an Iframe. > > William > > -- > William Stein > Professor of Mathematics > University of Washington > http://wstein.org > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: From benjaminrk at gmail.com Thu Nov 13 01:17:57 2014 From: benjaminrk at gmail.com (MinRK) Date: Wed, 12 Nov 2014 22:17:57 -0800 Subject: [IPython-dev] Trouble installing IPython3 into dist-packages In-Reply-To: References: Message-ID: If you have done a pip install -e or setup.py develop at some point, there may be an egg-link or entry in easy-install.pth pointing to that directory. Repeated pip uninstall ipython until it finds no ipython to uninstall *may* remove it, or you may need to go hunting for it yourself. You can look for it with: grep -r dongta /usr/local/lib/python* grep -r dongta ~/.local/lib/python* Yay, Python packaging! -MinRK ? On Wed, Nov 12, 2014 at 10:07 PM, Dong Ta wrote: > Thank you. I've just tried > > sudo pip3 install file://$PWD#egg=ipython[notebook] --upgrade > --force-reinstall > > > which installed successfully. However, the original problem remains, i.e. > it doesn't install IPython package into /usr/lib/python3/dist-packages > > In [1]: from IPython import core > > In [2]: ??core > Type: module > String form: '/home/dongta/ipython/IPython/core/__init__.py'> > File: /home/dongta/ipython/IPython/core/__init__.py > Docstring: > > > although the entry point is installed correctly, i.e. > > ~$ which ipython > /usr/local/bin/ipython > > > > On Wed, Nov 12, 2014 at 5:06 PM, Thomas Kluyver wrote: > >> On 12 November 2014 16:53, Dong Ta wrote: >> >>> Without -e, the command fails. Here's the full stdout/stderr: >> >> >> I am told that the official spelling of pip install ".[notebook]" is: >> >> pip install file://$PWD#egg=ipython[notebook] >> >> *sigh* - Python packaging tools never stop finding new ways to suck. >> >> Thomas >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Thu Nov 13 02:00:08 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Wed, 12 Nov 2014 23:00:08 -0800 Subject: [IPython-dev] Nov 17 In-Reply-To: References: Message-ID: Yep, I'll be there. Thomas On 12 Nov 2014 22:12, "MinRK" wrote: > Sorry, our ML server has been down for a few days. > > I think Fernando may still be out of town Monday, but Thomas and I should > be on campus and available. > > -MinRK > > On Sun, Nov 9, 2014 at 7:55 PM, William Stein wrote: > >> Hi, >> >> I'll be in Berkeley Monday, Nov 17, and would be interested in meeting >> or coding with some IPython developers if you guys will be around.... >> Maybe we could talk about synchronization or embedding IPython >> notebooks in other pages without using an Iframe. >> >> William >> >> -- >> William Stein >> Professor of Mathematics >> University of Washington >> http://wstein.org >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From dongta.hds at gmail.com Thu Nov 13 03:38:26 2014 From: dongta.hds at gmail.com (Dong Ta) Date: Thu, 13 Nov 2014 00:38:26 -0800 Subject: [IPython-dev] Trouble installing IPython3 into dist-packages In-Reply-To: References: Message-ID: Thanks. This grep -r dongta /usr/local/lib/python* , in addition to repeated pip uninstall ipython, solved the problem. I'm still scratching my head and reading stuffs about Python packing to understand why. DT On Wed, Nov 12, 2014 at 10:17 PM, MinRK wrote: > If you have done a pip install -e or setup.py develop at some point, > there may be an egg-link or entry in easy-install.pth pointing to that > directory. Repeated pip uninstall ipython until it finds no ipython to > uninstall *may* remove it, or you may need to go hunting for it yourself. > You can look for it with: > > grep -r dongta /usr/local/lib/python* > grep -r dongta ~/.local/lib/python* > > Yay, Python packaging! > > -MinRK > ? > > On Wed, Nov 12, 2014 at 10:07 PM, Dong Ta wrote: > >> Thank you. I've just tried >> >> sudo pip3 install file://$PWD#egg=ipython[notebook] --upgrade >> --force-reinstall >> >> >> which installed successfully. However, the original problem remains, i.e. >> it doesn't install IPython package into /usr/lib/python3/dist-packages >> >> In [1]: from IPython import core >> >> In [2]: ??core >> Type: module >> String form: > '/home/dongta/ipython/IPython/core/__init__.py'> >> File: /home/dongta/ipython/IPython/core/__init__.py >> Docstring: >> >> >> although the entry point is installed correctly, i.e. >> >> ~$ which ipython >> /usr/local/bin/ipython >> >> >> >> On Wed, Nov 12, 2014 at 5:06 PM, Thomas Kluyver wrote: >> >>> On 12 November 2014 16:53, Dong Ta wrote: >>> >>>> Without -e, the command fails. Here's the full stdout/stderr: >>> >>> >>> I am told that the official spelling of pip install ".[notebook]" is: >>> >>> pip install file://$PWD#egg=ipython[notebook] >>> >>> *sigh* - Python packaging tools never stop finding new ways to suck. >>> >>> Thomas >>> >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>> >>> >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jtaylor.debian at googlemail.com Thu Nov 13 14:16:03 2014 From: jtaylor.debian at googlemail.com (Julian Taylor) Date: Thu, 13 Nov 2014 20:16:03 +0100 Subject: [IPython-dev] Trouble installing IPython3 into dist-packages In-Reply-To: References: Message-ID: <54650373.5060900@googlemail.com> dist-packages (is package manager territory, admins should not install non-packaged software in it. The local system installation directory for python packages is site-packages which should also be the folder used by default by pip and setup.py install On 13.11.2014 09:38, Dong Ta wrote: > Thanks. This > > grep -r dongta /usr/local/lib/python* > > , in addition to repeated pip uninstall ipython, solved the problem. > > I'm still scratching my head and reading stuffs about Python packing to > understand why. > > DT > > > On Wed, Nov 12, 2014 at 10:17 PM, MinRK > wrote: > > If you have done a |pip install -e| or |setup.py develop| at some > point, there may be an |egg-link| or entry in |easy-install.pth| > pointing to that directory. Repeated |pip uninstall ipython| until > it finds no ipython to uninstall /may/ remove it, or you may need to > go hunting for it yourself. You can look for it with: > > |grep -r dongta /usr/local/lib/python* > grep -r dongta ~/.local/lib/python* > | > > Yay, Python packaging! > > -MinRK > > ? > > On Wed, Nov 12, 2014 at 10:07 PM, Dong Ta > wrote: > > Thank you. I've just tried > > sudo pip3 install file://$PWD#egg=ipython[notebook] > --upgrade --force-reinstall > > > which installed successfully. However, the original problem > remains, i.e. it doesn't install IPython package into > /usr/lib/python3/dist-packages > > In [1]: from IPython import core > > In [2]: ??core > Type: module > String form: '/home/dongta/ipython/IPython/core/__init__.py'> > File: /home/dongta/ipython/IPython/core/__init__.py > Docstring: > > > although the entry point is installed correctly, i.e. > > ~$ which ipython > /usr/local/bin/ipython > > > > On Wed, Nov 12, 2014 at 5:06 PM, Thomas Kluyver > > wrote: > > On 12 November 2014 16:53, Dong Ta > wrote: > > Without -e, the command fails. Here's the full > stdout/stderr: > > > I am told that the official spelling of pip install > ".[notebook]" is: > > pip install file://$PWD#egg=ipython[notebook] > > *sigh* - Python packaging tools never stop finding new ways > to suck. > > Thomas > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > From kikocorreoso at gmail.com Thu Nov 13 15:12:36 2014 From: kikocorreoso at gmail.com (Kiko) Date: Thu, 13 Nov 2014 21:12:36 +0100 Subject: [IPython-dev] notebooks of my talk at PyConES about how to customise the notebook Message-ID: Hi all, Last weekend I gave a talk at the PyConES (Spain) about how to customise the notebook. All the notebooks are in (far from perfect) english and available in this repo: https://github.com/kikocorreoso/PyConES14_talk-Hacking_the_Notebook I talked about how to create magic functions, rich display, how to use the IPython javascript API, create shortcuts, create JS extensions, create profiles, dummy kernels, use the custom.jss/css,... The information could have some issues so if you find something that it is not correct, just let me know to correct it. I hope it could be useful for some people. You could find a lot of references along the notebooks. Kind regards. P.D.: Some examples are based in previous work by Damian, Min, Cyrille and others so thanks to all the IPython/Jupyter community. -------------- next part -------------- An HTML attachment was scrubbed... URL: From benjaminrk at gmail.com Thu Nov 13 15:27:04 2014 From: benjaminrk at gmail.com (MinRK) Date: Thu, 13 Nov 2014 12:27:04 -0800 Subject: [IPython-dev] notebooks of my talk at PyConES about how to customise the notebook In-Reply-To: References: Message-ID: Very cool, thanks for sharing! -MinRK On Thu, Nov 13, 2014 at 12:12 PM, Kiko wrote: > Hi all, > > Last weekend I gave a talk at the PyConES (Spain) about how to customise > the notebook. > > All the notebooks are in (far from perfect) english and available in this > repo: > https://github.com/kikocorreoso/PyConES14_talk-Hacking_the_Notebook > > I talked about how to create magic functions, rich display, how to use the > IPython javascript API, create shortcuts, create JS extensions, create > profiles, dummy kernels, use the custom.jss/css,... > > The information could have some issues so if you find something that it is > not correct, just let me know to correct it. > > I hope it could be useful for some people. > > You could find a lot of references along the notebooks. > > Kind regards. > > P.D.: Some examples are based in previous work by Damian, Min, Cyrille and > others so thanks to all the IPython/Jupyter community. > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From damianavila at gmail.com Thu Nov 13 16:24:20 2014 From: damianavila at gmail.com (=?UTF-8?Q?Dami=C3=A1n_Avila?=) Date: Thu, 13 Nov 2014 18:24:20 -0300 Subject: [IPython-dev] notebooks of my talk at PyConES about how to customise the notebook In-Reply-To: References: Message-ID: Thanks Kiko! Nice stuff... Damian 2014-11-13 17:27 GMT-03:00 MinRK : > Very cool, thanks for sharing! > > -MinRK > > On Thu, Nov 13, 2014 at 12:12 PM, Kiko wrote: > >> Hi all, >> >> Last weekend I gave a talk at the PyConES (Spain) about how to customise >> the notebook. >> >> All the notebooks are in (far from perfect) english and available in this >> repo: >> https://github.com/kikocorreoso/PyConES14_talk-Hacking_the_Notebook >> >> I talked about how to create magic functions, rich display, how to use >> the IPython javascript API, create shortcuts, create JS extensions, create >> profiles, dummy kernels, use the custom.jss/css,... >> >> The information could have some issues so if you find something that it >> is not correct, just let me know to correct it. >> >> I hope it could be useful for some people. >> >> You could find a lot of references along the notebooks. >> >> Kind regards. >> >> P.D.: Some examples are based in previous work by Damian, Min, Cyrille >> and others so thanks to all the IPython/Jupyter community. >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -- *Dami?n* -------------- next part -------------- An HTML attachment was scrubbed... URL: From wstein at gmail.com Thu Nov 13 21:26:31 2014 From: wstein at gmail.com (William Stein) Date: Thu, 13 Nov 2014 18:26:31 -0800 Subject: [IPython-dev] Nov 17 In-Reply-To: References: Message-ID: On Wed, Nov 12, 2014 at 11:00 PM, Thomas Kluyver wrote: > Yep, I'll be there. > > Thomas Where are you guys located on campus these days? Is it in the new space adjacent to the library? William > > On 12 Nov 2014 22:12, "MinRK" wrote: >> >> Sorry, our ML server has been down for a few days. >> >> I think Fernando may still be out of town Monday, but Thomas and I should >> be on campus and available. >> >> -MinRK >> >> On Sun, Nov 9, 2014 at 7:55 PM, William Stein wrote: >>> >>> Hi, >>> >>> I'll be in Berkeley Monday, Nov 17, and would be interested in meeting >>> or coding with some IPython developers if you guys will be around.... >>> Maybe we could talk about synchronization or embedding IPython >>> notebooks in other pages without using an Iframe. >>> >>> William >>> >>> -- >>> William Stein >>> Professor of Mathematics >>> University of Washington >>> http://wstein.org >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -- William Stein Professor of Mathematics University of Washington http://wstein.org From takowl at gmail.com Thu Nov 13 21:40:55 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Thu, 13 Nov 2014 18:40:55 -0800 Subject: [IPython-dev] Nov 17 In-Reply-To: References: Message-ID: On 13 Nov 2014 18:27, "William Stein" wrote: > Where are you guys located on campus these days? Is it in the new > space adjacent to the library? We're still in Barker hall, though the Bids space in the library is now up and running, and we sometimes use meeting rooms there. Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From roalexan at microsoft.com Fri Nov 14 13:37:24 2014 From: roalexan at microsoft.com (Robert Alexander) Date: Fri, 14 Nov 2014 18:37:24 +0000 Subject: [IPython-dev] sample code using ipython api to get a notebook and edit it Message-ID: <1415990242764.13627@microsoft.com> Hi. I can create an IPython notebook using the IPython API. For example: import sys from IPython.nbformat import current as nbf nb = nbf.new_notebook() with open('test.ipynb', 'w') as f: nbf.write(nb, f, 'ipynb') Does somebody have some sample usage of the IPython API to get and edit a notebook, like adding/deleting cells to/from it? So far, I've been trying to use the documentation at: http://ipython.org/ipython-doc/2/api/generated/IPython.nbformat.v3.nbbase.html, but I'm having a hard time getting it to work. Thanks, Robert -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Fri Nov 14 13:42:59 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Fri, 14 Nov 2014 10:42:59 -0800 Subject: [IPython-dev] sample code using ipython api to get a notebook and edit it In-Reply-To: <1415990242764.13627@microsoft.com> References: <1415990242764.13627@microsoft.com> Message-ID: Hi Robert, On 14 November 2014 10:37, Robert Alexander wrote: > Does somebody have some sample usage of the IPython API to get and edit a > notebook, like adding/deleting cells to/from it? So far, I've been trying > to use the documentation at: > http://ipython.org/ipython-doc/2/api/generated/IPython.nbformat.v3.nbbase.html, > but I'm having a hard time getting it to work. > You can load a notebook with the read function. Then its cells are accessible as a list, either at nb.worksheets[0].cells (nbformat 3), or at nb.cells (nbformat 4 in master). You can modify it just like any Python list. Then save it back to disk with write(). There are some changes to this API in master. We're deprecating nbformat.current: now you will use read and write from the top level of nbformat, and new_notebook, new_cell etc. from the version specific subpackages, like IPython.nbformat.v4. Best wishes, Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From bussonniermatthias at gmail.com Sat Nov 15 05:25:42 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Sat, 15 Nov 2014 11:25:42 +0100 Subject: [IPython-dev] Bloomberg Open source Day London 29-30 Message-ID: <3622012C-DD42-4CE8-AA6F-F2D39E5B999A@gmail.com> Hi all, After organizing an equivalent event last september in NYC, Bloomberg is organizing a hackathon on scientific Python this November in London. If you are nearby and like to contribute to some Scientific Python open source project (or just say hi) please come by, I'll be there among others to mentor peoples to get starting on hacking IPython/Jupyter. The even is RSVP, and you can get more info here: http://go.bloomberg.com/promo/invite/bloomberg-open-source-day-scientific-python/ You can view a recap of last event here: https://drive.google.com/file/d/0By_u4j-fZ875Wk9EaElpVmNzQkUwQUVwZTBESkVJb1IxWF9J Hopping to see you there in 2 weeks, -- Matthias -------------- next part -------------- An HTML attachment was scrubbed... URL: From bussonniermatthias at gmail.com Sat Nov 15 12:24:18 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Sat, 15 Nov 2014 18:24:18 +0100 Subject: [IPython-dev] [Done] Time for a Jupyter mailing list? In-Reply-To: References: <544AF18C.7010808@creativetrax.com> Message-ID: <7B4C57BF-3A23-40A9-8C55-E7245C849870@gmail.com> Hello all, The accretion disk of jupyter have now forme a google group: https://groups.google.com/forum/#!forum/jupyter Nearby bodies that are interested in participating are welcome to joined the gravity field. Please also feel free to distribute the link and reach in nearby systems. We'll try to keep this mailing list for everything that is not IPython/Python specific. Cheers, -- Matthias Le 30 oct. 2014 ? 03:23, Fernando Perez a ?crit : > > On Fri, Oct 24, 2014 at 5:40 PM, Jason Grout wrote: > +1 on GG, with gmane.org mirroring. > > Yup, I think that's a great plan. We'll make that and announce the address shortly so interested folks can join. > > Cheers, > > f > > > -- > Fernando Perez (@fperez_org; http://fperez.org) > fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) > fernando.perez-at-berkeley: contact me here for any direct mail > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev -------------- next part -------------- An HTML attachment was scrubbed... URL: From roalexan at microsoft.com Mon Nov 17 09:26:44 2014 From: roalexan at microsoft.com (Robert Alexander) Date: Mon, 17 Nov 2014 14:26:44 +0000 Subject: [IPython-dev] sample code using ipython api to get a notebook and edit it In-Reply-To: References: <1415990242764.13627@microsoft.com>, Message-ID: <1416234401560.36385@microsoft.com> Thanks. Based on your feedback, I wrote some sample code: from IPython.nbformat import current as nbf # Create new notebook with one cell nb = nbf.new_notebook() with open('test.ipynb', 'w') as f: cells = [ nbf.new_code_cell("cell 1 content") ] nb['worksheets'].append(nbf.new_worksheet(cells=cells)) nbf.write(nb, f, 'ipynb') f.close() # Read the notebook and write a new cell to it with open('test.ipynb', 'r+') as f2: nb = nbf.read(f2, 'ipynb') cells = nb.worksheets[0].cells cells.append(nbf.new_code_cell("cell 2 content")) nb.worksheets[0].cells = cells nbf.write(nb, f2, 'ipynb') f2.close() This will correctly create a new notebook with a single sample cell; however, reading the notebook and appending the new cell to the existing worksheet is not quite right. The resulting file has a second worksheet with the two new cells appended to the original worksheet. Is there a better way to do what I'm trying to do? ________________________________ From: ipython-dev-bounces at scipy.org on behalf of Thomas Kluyver Sent: Friday, November 14, 2014 1:42 PM To: IPython developers list Subject: Re: [IPython-dev] sample code using ipython api to get a notebook and edit it Hi Robert, On 14 November 2014 10:37, Robert Alexander > wrote: Does somebody have some sample usage of the IPython API to get and edit a notebook, like adding/deleting cells to/from it? So far, I've been trying to use the documentation at: http://ipython.org/ipython-doc/2/api/generated/IPython.nbformat.v3.nbbase.html, but I'm having a hard time getting it to work. You can load a notebook with the read function. Then its cells are accessible as a list, either at nb.worksheets[0].cells (nbformat 3), or at nb.cells (nbformat 4 in master). You can modify it just like any Python list. Then save it back to disk with write(). There are some changes to this API in master. We're deprecating nbformat.current: now you will use read and write from the top level of nbformat, and new_notebook, new_cell etc. from the version specific subpackages, like IPython.nbformat.v4. Best wishes, Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Mon Nov 17 15:10:10 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Mon, 17 Nov 2014 12:10:10 -0800 Subject: [IPython-dev] sample code using ipython api to get a notebook and edit it In-Reply-To: <1416234401560.36385@microsoft.com> References: <1415990242764.13627@microsoft.com> <1416234401560.36385@microsoft.com> Message-ID: On 17 November 2014 06:26, Robert Alexander wrote: > with open('test.ipynb', 'r+') as f2: You're opening one file, reading from it, and then writing back to the same file descriptor. That writes the new content after the end of the old content, resulting in an invalid notebook. To modify files, we would normally open, read, close, make the changes we want, open, write, close. Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Mon Nov 17 14:06:58 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Mon, 17 Nov 2014 11:06:58 -0800 Subject: [IPython-dev] cite2c: Insert citations into IPython notebooks Message-ID: I'm pleased to announce cite2c, a notebook extension for adding citations and bibliographies to Markdown cells in your notebooks. https://github.com/takluyver/cite2c When you click the insert citation button, you can search your public Zotero library for citations. The citation data is added to notebook metadata as CSL JSON, while the Markdown cell gets an HTML tag with an ID for that citation (this works like the existing references for export to bibtex - integrating with that is a future goal). Citations are rendered using citeproc-js. Consider this project alpha status: it's ready for you to try out, but you should be prepared for bugs and missing features. Please come and help improve it! Thanks, Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From ellisonbg at gmail.com Mon Nov 17 15:14:05 2014 From: ellisonbg at gmail.com (Brian Granger) Date: Mon, 17 Nov 2014 12:14:05 -0800 Subject: [IPython-dev] cite2c: Insert citations into IPython notebooks In-Reply-To: References: Message-ID: Awesome! Thanks for working on this! On Mon, Nov 17, 2014 at 11:06 AM, Thomas Kluyver wrote: > I'm pleased to announce cite2c, a notebook extension for adding citations > and bibliographies to Markdown cells in your notebooks. > > https://github.com/takluyver/cite2c > > When you click the insert citation button, you can search your public Zotero > library for citations. The citation data is added to notebook metadata as > CSL JSON, while the Markdown cell gets an HTML tag with an ID for that > citation (this works like the existing references for export to bibtex - > integrating with that is a future goal). Citations are rendered using > citeproc-js. > > Consider this project alpha status: it's ready for you to try out, but you > should be prepared for bugs and missing features. Please come and help > improve it! > > Thanks, > Thomas > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -- Brian E. Granger Cal Poly State University, San Luis Obispo @ellisonbg on Twitter and GitHub bgranger at calpoly.edu and ellisonbg at gmail.com From pelson.pub at gmail.com Mon Nov 17 11:47:24 2014 From: pelson.pub at gmail.com (Phil Elson) Date: Mon, 17 Nov 2014 16:47:24 +0000 Subject: [IPython-dev] Logging commands using the %logstart magic Message-ID: I'm trying to replicate functionality found in IDL which generates something called "journal" files. Journal files in IDL are essentially a log of commands alongside the (commented out) output and themselves produce valid code. IPython's %logstart with the "-o" flag is almost exactly what I need, except it doesn't seem to record standard out (only IPython out): > ipython Python 2.7.6 IPython 2.2.0 -- An enhanced Interactive Python. In [1]: %logstart -o testing_logging.py Activating auto-logging. Current session state plus future input saved. Filename : testing_logging.py Mode : backup Output logging : True Raw input log : False Timestamping : False State : active In [2]: print 'Hello logging world non-IPython out!' Hello logging world non-IPython out! In [3]: 'Hello logging world IPython out!' Out[3]: 'Hello logging world IPython out!' In [4]: exit The resulting file looks like: # IPython log file print 'Hello logging world non-IPython out!' 'Hello logging world IPython out!' #[Out]# 'Hello logging world IPython out!' exit() Essentially I'd like the contents of the second input to be recorded as it is for the third. Before I go hacking around with the magic and sys.stdout, does anybody have a quick solution to this? Thanks, Phil -------------- next part -------------- An HTML attachment was scrubbed... URL: From souheil.inati at gmail.com Mon Nov 17 20:46:32 2014 From: souheil.inati at gmail.com (Souheil Inati) Date: Mon, 17 Nov 2014 20:46:32 -0500 Subject: [IPython-dev] interacting with ipython kernels from c++ Message-ID: <69A018D4-AE78-4B39-9809-B0CABAD26E8A@gmail.com> Hi all, I have a c++ application from which I would like to do call Python, R, and Julia, etc. and it looks to me like I could use the ipython kernels as a nice uniform interface. I have seen the docs on writing a new kernel in a language other than python. What I am wondering is if someone has documented the complementary problem, i.e. writing a front end in a different language. Any pointers would be greatly appreciated. Thanks, Souheil ------------------------- Souheil Inati, PhD Staff Scientist FMRI Facility NIMH/NIH/DHHS souheil.inati at nih.gov From takowl at gmail.com Mon Nov 17 20:59:12 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Mon, 17 Nov 2014 17:59:12 -0800 Subject: [IPython-dev] interacting with ipython kernels from c++ In-Reply-To: <69A018D4-AE78-4B39-9809-B0CABAD26E8A@gmail.com> References: <69A018D4-AE78-4B39-9809-B0CABAD26E8A@gmail.com> Message-ID: On 17 November 2014 17:46, Souheil Inati wrote: > What I am wondering is if someone has documented the complementary > problem, i.e. writing a front end in a different language. It should absolutely be possible, but I don't think it's very well explored or documented. Besides the messaging spec ( http://ipython.org/ipython-doc/dev/development/messaging.html ), you will want to look at the code in IPython.kernel.manager, which launches and shuts down kernels, and IPython.kernel.client, which controls communicating with a kernel. I hope that gives you some useful things to get started with. Feel free to ask more questions here. Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From nick.bollweg at gmail.com Mon Nov 17 21:36:23 2014 From: nick.bollweg at gmail.com (Nicholas Bollweg) Date: Mon, 17 Nov 2014 21:36:23 -0500 Subject: [IPython-dev] interacting with ipython kernels from c++ In-Reply-To: References: <69A018D4-AE78-4B39-9809-B0CABAD26E8A@gmail.com> Message-ID: This topic just came up in discussion with one of my colleagues: specifically, the widget framework is actually a very good way to provide an arbitrary, evented API for a run-time instance. Imagine being able to connect to a running kernel _from_ a running kernel, and programmatically interact with widgets: from within a notebook, restore a snapshotted docker instance, of _some kernel_. One could ask, "what widgets do you have?" It says back, "I have a button, two float sliders and an output float." slide, slide, click: when the output is populated, the run is "done". This allows building very "human" APIs for potentially quite complicated environments. On Mon, Nov 17, 2014 at 8:59 PM, Thomas Kluyver wrote: > On 17 November 2014 17:46, Souheil Inati wrote: > >> What I am wondering is if someone has documented the complementary >> problem, i.e. writing a front end in a different language. > > > It should absolutely be possible, but I don't think it's very well > explored or documented. Besides the messaging spec ( > http://ipython.org/ipython-doc/dev/development/messaging.html ), you will > want to look at the code in IPython.kernel.manager, which launches and > shuts down kernels, and IPython.kernel.client, which controls communicating > with a kernel. > > I hope that gives you some useful things to get started with. Feel free to > ask more questions here. > > Thomas > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From wes.turner at gmail.com Tue Nov 18 02:26:52 2014 From: wes.turner at gmail.com (Wes Turner) Date: Tue, 18 Nov 2014 01:26:52 -0600 Subject: [IPython-dev] Logging commands using the %logstart magic In-Reply-To: References: Message-ID: * https://stackoverflow.com/questions/2737966/how-to-change-the-stdin-encoding-on-python * https://github.com/ipython/ipython/blob/master/IPython/terminal/interactiveshell.py#L413 * https://github.com/ipython/ipython/blob/master/IPython/core/history.py#L234 On Mon, Nov 17, 2014 at 10:47 AM, Phil Elson wrote: > I'm trying to replicate functionality found in IDL which generates > something called "journal" files. Journal files in IDL are essentially a > log of commands alongside the (commented out) output and themselves produce > valid code. > > IPython's %logstart with the "-o" flag is almost exactly what I need, > except it doesn't seem to record standard out (only IPython out): > > > ipython > Python 2.7.6 > IPython 2.2.0 -- An enhanced Interactive Python. > > In [1]: %logstart -o testing_logging.py > Activating auto-logging. Current session state plus future input saved. > Filename : testing_logging.py > Mode : backup > Output logging : True > Raw input log : False > Timestamping : False > State : active > > In [2]: print 'Hello logging world non-IPython out!' > Hello logging world non-IPython out! > > In [3]: 'Hello logging world IPython out!' > Out[3]: 'Hello logging world IPython out!' > > In [4]: exit > > The resulting file looks like: > > # IPython log file > > print 'Hello logging world non-IPython out!' > 'Hello logging world IPython out!' > #[Out]# 'Hello logging world IPython out!' > exit() > > Essentially I'd like the contents of the second input to be recorded as it > is for the third. > > Before I go hacking around with the magic and sys.stdout, does anybody > have a quick solution to this? > > Thanks, > > Phil > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -- Wes Turner https://westurner.github.io/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From wes.turner at gmail.com Tue Nov 18 02:28:37 2014 From: wes.turner at gmail.com (Wes Turner) Date: Tue, 18 Nov 2014 01:28:37 -0600 Subject: [IPython-dev] Logging commands using the %logstart magic In-Reply-To: References: Message-ID: IPython notebook format docs: IPython notebook nbformat v4 JSONSchema: > https://github.com/minrk/ipython/blob/nbformat4/IPython/nbformat/v4/nbformat.v4.schema.json > Metadata Documentation: > https://github.com/minrk/ipython/blob/nbformat4/docs/source/notebook/nbformat.rst#metadata > JSON-LD JSONSchema: > https://github.com/json-ld/json-ld.org/blob/master/schemas/jsonld-schema.json ( https://youtrack.jetbrains.com/issue/PY-14326 ) On Mon, Nov 17, 2014 at 10:47 AM, Phil Elson wrote: > I'm trying to replicate functionality found in IDL which generates > something called "journal" files. Journal files in IDL are essentially a > log of commands alongside the (commented out) output and themselves produce > valid code. > > IPython's %logstart with the "-o" flag is almost exactly what I need, > except it doesn't seem to record standard out (only IPython out): > > > ipython > Python 2.7.6 > IPython 2.2.0 -- An enhanced Interactive Python. > > In [1]: %logstart -o testing_logging.py > Activating auto-logging. Current session state plus future input saved. > Filename : testing_logging.py > Mode : backup > Output logging : True > Raw input log : False > Timestamping : False > State : active > > In [2]: print 'Hello logging world non-IPython out!' > Hello logging world non-IPython out! > > In [3]: 'Hello logging world IPython out!' > Out[3]: 'Hello logging world IPython out!' > > In [4]: exit > > The resulting file looks like: > > # IPython log file > > print 'Hello logging world non-IPython out!' > 'Hello logging world IPython out!' > #[Out]# 'Hello logging world IPython out!' > exit() > > Essentially I'd like the contents of the second input to be recorded as it > is for the third. > > Before I go hacking around with the magic and sys.stdout, does anybody > have a quick solution to this? > > Thanks, > > Phil > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -- Wes Turner https://westurner.github.io/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From lucienboillod at gmail.com Tue Nov 18 11:09:45 2014 From: lucienboillod at gmail.com (Lucien Boillod) Date: Tue, 18 Nov 2014 17:09:45 +0100 Subject: [IPython-dev] Back-End implementation of IPython Widget Message-ID: <0CDAD968-5908-4149-A807-DA5E2EBD419D@gmail.com> Hello all, I would like to know more about your back-end implementation of your widget, specially using Javascript. Indeed when one of your widget is used into the notebook, the Javascript source code doesn't appear into ipynb source. But in my implementation I use display(Javascript(myjavascript)) in my back-end file (.py) and consequences are unconvenients because each times my .py file is import or executed, the javascript appear directly in the ipynb source. Thank you, Cheers, Lucien Boillod From cyrille.rossant at gmail.com Tue Nov 18 11:29:42 2014 From: cyrille.rossant at gmail.com (Cyrille Rossant) Date: Tue, 18 Nov 2014 17:29:42 +0100 Subject: [IPython-dev] Back-End implementation of IPython Widget In-Reply-To: <0CDAD968-5908-4149-A807-DA5E2EBD419D@gmail.com> References: <0CDAD968-5908-4149-A807-DA5E2EBD419D@gmail.com> Message-ID: > I would like to know more about your back-end implementation of your widget, specially using Javascript. > Indeed when one of your widget is used into the notebook, the Javascript source code doesn't appear into ipynb source. > But in my implementation I use display(Javascript(myjavascript)) in my back-end file (.py) and consequences are unconvenients because > each times my .py file is import or executed, the javascript appear directly in the ipynb source. That might be because built-in widgets use a dedicated widget architecture, whereas you use the rich display feature which is different as far as I can tell. If you want to use the same functionality than built-in widgets, you'll have to build custom widgets as described here: http://nbviewer.ipython.org/github/ipython/ipython/blob/master/examples/Interactive%20Widgets/Custom%20Widget%20-%20Hello%20World.ipynb This tutorial might also be of interest: https://github.com/rossant/euroscipy2014 Cyrille From pelson.pub at gmail.com Tue Nov 18 12:53:43 2014 From: pelson.pub at gmail.com (Phil Elson) Date: Tue, 18 Nov 2014 17:53:43 +0000 Subject: [IPython-dev] Logging commands using the %logstart magic In-Reply-To: References: Message-ID: Thanks Wes, I explicitly am not asking about the notebook, which I know does this admirably (and is what I would use too). I've been asked for command line behaviour from the (I)Python terminal which records input and output in an almost exact fashion to that of "%logstart", except logstart does not store stdout, only IPython displayed output. I haven't yet dug into the possibility of redirecting stdout to IPython out, or indeed directly out to file with the associated input, but I imagine there will be some buffering issues that I'd like not to have to deal with, if the functionality already exists in IPython. Cheers, On 18 November 2014 07:28, Wes Turner wrote: > IPython notebook format docs: > > > IPython notebook nbformat v4 JSONSchema: >> https://github.com/minrk/ipython/blob/nbformat4/IPython/nbformat/v4/nbformat.v4.schema.json >> Metadata Documentation: >> https://github.com/minrk/ipython/blob/nbformat4/docs/source/notebook/nbformat.rst#metadata >> JSON-LD JSONSchema: >> https://github.com/json-ld/json-ld.org/blob/master/schemas/jsonld-schema.json > > > ( https://youtrack.jetbrains.com/issue/PY-14326 ) > > On Mon, Nov 17, 2014 at 10:47 AM, Phil Elson wrote: > >> I'm trying to replicate functionality found in IDL which generates >> something called "journal" files. Journal files in IDL are essentially a >> log of commands alongside the (commented out) output and themselves produce >> valid code. >> >> IPython's %logstart with the "-o" flag is almost exactly what I need, >> except it doesn't seem to record standard out (only IPython out): >> >> > ipython >> Python 2.7.6 >> IPython 2.2.0 -- An enhanced Interactive Python. >> >> In [1]: %logstart -o testing_logging.py >> Activating auto-logging. Current session state plus future input saved. >> Filename : testing_logging.py >> Mode : backup >> Output logging : True >> Raw input log : False >> Timestamping : False >> State : active >> >> In [2]: print 'Hello logging world non-IPython out!' >> Hello logging world non-IPython out! >> >> In [3]: 'Hello logging world IPython out!' >> Out[3]: 'Hello logging world IPython out!' >> >> In [4]: exit >> >> The resulting file looks like: >> >> # IPython log file >> >> print 'Hello logging world non-IPython out!' >> 'Hello logging world IPython out!' >> #[Out]# 'Hello logging world IPython out!' >> exit() >> >> Essentially I'd like the contents of the second input to be recorded as >> it is for the third. >> >> Before I go hacking around with the magic and sys.stdout, does anybody >> have a quick solution to this? >> >> Thanks, >> >> Phil >> >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > > -- > Wes Turner > https://westurner.github.io/ > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Tue Nov 18 13:02:58 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Tue, 18 Nov 2014 10:02:58 -0800 Subject: [IPython-dev] Logging commands using the %logstart magic In-Reply-To: References: Message-ID: Hi Phil, logstart currently only captures Python return values as passed to sys.displayhook(). We do have the necessary machinery to capture other kinds of output: see IPython.utils.capture.capture_output(). Thomas On 18 November 2014 09:53, Phil Elson wrote: > Thanks Wes, > > I explicitly am not asking about the notebook, which I know does this > admirably (and is what I would use too). > I've been asked for command line behaviour from the (I)Python terminal > which records input and output in an almost exact fashion to that of > "%logstart", except logstart does not store stdout, only IPython displayed > output. > > I haven't yet dug into the possibility of redirecting stdout to IPython > out, or indeed directly out to file with the associated input, but I > imagine there will be some buffering issues that I'd like not to have to > deal with, if the functionality already exists in IPython. > > Cheers, > > On 18 November 2014 07:28, Wes Turner wrote: > >> IPython notebook format docs: >> >> >> IPython notebook nbformat v4 JSONSchema: >>> https://github.com/minrk/ipython/blob/nbformat4/IPython/nbformat/v4/nbformat.v4.schema.json >>> Metadata Documentation: >>> https://github.com/minrk/ipython/blob/nbformat4/docs/source/notebook/nbformat.rst#metadata >>> JSON-LD JSONSchema: >>> https://github.com/json-ld/json-ld.org/blob/master/schemas/jsonld-schema.json >> >> >> ( https://youtrack.jetbrains.com/issue/PY-14326 ) >> >> On Mon, Nov 17, 2014 at 10:47 AM, Phil Elson >> wrote: >> >>> I'm trying to replicate functionality found in IDL which generates >>> something called "journal" files. Journal files in IDL are essentially a >>> log of commands alongside the (commented out) output and themselves produce >>> valid code. >>> >>> IPython's %logstart with the "-o" flag is almost exactly what I need, >>> except it doesn't seem to record standard out (only IPython out): >>> >>> > ipython >>> Python 2.7.6 >>> IPython 2.2.0 -- An enhanced Interactive Python. >>> >>> In [1]: %logstart -o testing_logging.py >>> Activating auto-logging. Current session state plus future input saved. >>> Filename : testing_logging.py >>> Mode : backup >>> Output logging : True >>> Raw input log : False >>> Timestamping : False >>> State : active >>> >>> In [2]: print 'Hello logging world non-IPython out!' >>> Hello logging world non-IPython out! >>> >>> In [3]: 'Hello logging world IPython out!' >>> Out[3]: 'Hello logging world IPython out!' >>> >>> In [4]: exit >>> >>> The resulting file looks like: >>> >>> # IPython log file >>> >>> print 'Hello logging world non-IPython out!' >>> 'Hello logging world IPython out!' >>> #[Out]# 'Hello logging world IPython out!' >>> exit() >>> >>> Essentially I'd like the contents of the second input to be recorded as >>> it is for the third. >>> >>> Before I go hacking around with the magic and sys.stdout, does anybody >>> have a quick solution to this? >>> >>> Thanks, >>> >>> Phil >>> >>> >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>> >>> >> >> >> -- >> Wes Turner >> https://westurner.github.io/ >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From lucienboillod at gmail.com Wed Nov 19 04:53:18 2014 From: lucienboillod at gmail.com (Lucien Boillod) Date: Wed, 19 Nov 2014 10:53:18 +0100 Subject: [IPython-dev] Back-End implementation of IPython Widget In-Reply-To: References: <0CDAD968-5908-4149-A807-DA5E2EBD419D@gmail.com> Message-ID: <9120EB40-7B80-42C5-97D6-6DF025B203DD@gmail.com> > Le 18 nov. 2014 ? 17:29, Cyrille Rossant a ?crit : > >> I would like to know more about your back-end implementation of your widget, specially using Javascript. >> Indeed when one of your widget is used into the notebook, the Javascript source code doesn't appear into ipynb source. >> But in my implementation I use display(Javascript(myjavascript)) in my back-end file (.py) and consequences are unconvenients because >> each times my .py file is import or executed, the javascript appear directly in the ipynb source. > > That might be because built-in widgets use a dedicated widget > architecture, whereas you use the rich display feature which is > different as far as I can tell. > I use the same functionality and structure than built-in widgets, but I don?t do it directly in the notebook but in my separate file .py, and I call my widget with a magic command ( because of the Bi-directionnality between my project and the widget ). > If you want to use the same functionality than built-in widgets, > you'll have to build custom widgets as described here: > http://nbviewer.ipython.org/github/ipython/ipython/blob/master/examples/Interactive%20Widgets/Custom%20Widget%20-%20Hello%20World.ipynb > > This tutorial might also be of interest: > https://github.com/rossant/euroscipy2014 > But my question was really on the execution of the javascript part, as far as I can see in both of your examples the same method as me is used, because %%javascript is just a alias for the rich display feature display(Javascript(js)) And I wonder what is you back-end implementation in ipython notebook ( not in widget that people can create ). What I would like is have a new type of widget that I load into my notebook, but I don?t don?t load the javascript in the notebook, so %%javascript can?t work, that?s why I use display(Javascript(js)) and that?s work very well but I just wonder if I can get around the fact my Javascript code appear on the notebook metadata when we explore the .ipynb file where we have load the widget. Maybe it?s not odd and it?s because of the implementation of .ipynb file, I just was surprised to find the whole Javascript in the metadata. Thank you, Lucien From maidos93 at laposte.net Wed Nov 19 07:10:32 2014 From: maidos93 at laposte.net (thwiouz) Date: Wed, 19 Nov 2014 04:10:32 -0800 (PST) Subject: [IPython-dev] Testing IPython 3.0 Message-ID: <1416399032654-5078073.post@n6.nabble.com> Hi, I wanted to know how to install IPython 3.0 into a new environment using Anaconda. Thanks a lot -- View this message in context: http://python.6.x6.nabble.com/Testing-IPython-3-0-tp5078073.html Sent from the IPython - Development mailing list archive at Nabble.com. From nick.bollweg at gmail.com Wed Nov 19 11:13:37 2014 From: nick.bollweg at gmail.com (Nicholas Bollweg) Date: Wed, 19 Nov 2014 11:13:37 -0500 Subject: [IPython-dev] Testing IPython 3.0 In-Reply-To: <1416399032654-5078073.post@n6.nabble.com> References: <1416399032654-5078073.post@n6.nabble.com> Message-ID: There aren't any binaries, yet, so I don't think you can conda install... The usual isn't too bad: shouldn't be any different than this: > > http://ipython.org/ipython-doc/dev/install/install.html#installing-the-development-version > which is just: $ git clone --recursive https://github.com/ipython/ipython.git$ cd ipython$ python setup.py install If you want all the bells and whistles (notebook, tests, etc), i'd follow it up with: $ pip install -e .[all] On Wed, Nov 19, 2014 at 7:10 AM, thwiouz wrote: > Hi, > > I wanted to know how to install IPython 3.0 into a new environment using > Anaconda. > > Thanks a lot > > > > -- > View this message in context: > http://python.6.x6.nabble.com/Testing-IPython-3-0-tp5078073.html > Sent from the IPython - Development mailing list archive at Nabble.com. > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: From fperez.net at gmail.com Wed Nov 19 11:54:13 2014 From: fperez.net at gmail.com (Fernando Perez) Date: Wed, 19 Nov 2014 11:54:13 -0500 Subject: [IPython-dev] IPython notebook, tmpnb.org in Nature Blog In-Reply-To: References: Message-ID: [ Resending, I think this never made it during the recent ML outage ] On Thu, Nov 6, 2014 at 3:57 PM, Doug Blank wrote: > A few people must have worked hard to get that put together at Rackspace. > Well done! > And I really, really want to thank Kyle, Min, Brian, Stefan, Aron, and the whole teams at Rackspace and Nature who made this possible. The home stretch caught me flying to Brazil and speaking here, so I was able to help exactly at the 0% level, but everyone pulled together wonderfully to make it possible. And in the process, we continue refining tools like tmpnb, that we'll get a ton of mileage out of in the future. I can't emphasize enough the value of this mode of team work, especially in a context like this. That article began as a private phone call from Helen Shen, a Nature journalist months ago for an interview about IPython. I've had enough struggles in the past with journalists who insist on finding the "lead author" to credit on a piece that I was worried that would happen again (it's a terrible kind of incentive, pervasive in academia and media but absolutely toxic to good team dynamics). In this case, however, it proved to be very easy to move the conversation into a much more open and collaborative mode, Brian was interviewed as well, then he suggested the demo idea to Kyle who jumped in, then Min, etc... Everybody did their best, and instead of simply an interview, now we have a great piece showcasing not only the history of the project, but much more importantly, its direct potential with a live demo. Repeating in public what I told everyone in a private thread while we were preparing all this: It's an honor and a pleasure to work with a team like all of you. I couldn't be more grateful. All the best, -- Fernando Perez (@fperez_org; http://fperez.org) fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) fernando.perez-at-berkeley: contact me here for any direct mail -------------- next part -------------- An HTML attachment was scrubbed... URL: From rgbkrk at gmail.com Wed Nov 19 12:08:24 2014 From: rgbkrk at gmail.com (Kyle Kelley) Date: Wed, 19 Nov 2014 11:08:24 -0600 Subject: [IPython-dev] IPython notebook, tmpnb.org in Nature Blog In-Reply-To: References: Message-ID: Thanks Fernando! This makes me wonder if my messages ever went through either... On Wed, Nov 19, 2014 at 10:54 AM, Fernando Perez wrote: > [ Resending, I think this never made it during the recent ML outage ] > > On Thu, Nov 6, 2014 at 3:57 PM, Doug Blank wrote: > >> A few people must have worked hard to get that put together at Rackspace. >> Well done! >> > > And I really, really want to thank Kyle, Min, Brian, Stefan, Aron, and the > whole teams at Rackspace and Nature who made this possible. The home > stretch caught me flying to Brazil and speaking here, so I was able to help > exactly at the 0% level, but everyone pulled together wonderfully to make > it possible. And in the process, we continue refining tools like tmpnb, > that we'll get a ton of mileage out of in the future. > > I can't emphasize enough the value of this mode of team work, especially > in a context like this. That article began as a private phone call from > Helen Shen, a Nature journalist months ago for an interview about IPython. > I've had enough struggles in the past with journalists who insist on > finding the "lead author" to credit on a piece that I was worried that > would happen again (it's a terrible kind of incentive, pervasive in > academia and media but absolutely toxic to good team dynamics). In this > case, however, it proved to be very easy to move the conversation into a > much more open and collaborative mode, Brian was interviewed as well, then > he suggested the demo idea to Kyle who jumped in, then Min, etc... > Everybody did their best, and instead of simply an interview, now we have a > great piece showcasing not only the history of the project, but much more > importantly, its direct potential with a live demo. > > Repeating in public what I told everyone in a private thread while we were > preparing all this: > > It's an honor and a pleasure to work with a team like all of you. I > couldn't be more grateful. > > All the best, > > > -- > Fernando Perez (@fperez_org; http://fperez.org) > fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) > fernando.perez-at-berkeley: contact me here for any direct mail > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -- Kyle Kelley (@rgbkrk ; http://lambdaops.com) -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Wed Nov 19 12:47:01 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Wed, 19 Nov 2014 09:47:01 -0800 Subject: [IPython-dev] Back-End implementation of IPython Widget In-Reply-To: <9120EB40-7B80-42C5-97D6-6DF025B203DD@gmail.com> References: <0CDAD968-5908-4149-A807-DA5E2EBD419D@gmail.com> <9120EB40-7B80-42C5-97D6-6DF025B203DD@gmail.com> Message-ID: On 19 November 2014 01:53, Lucien Boillod wrote: > But my question was really on the execution of the javascript part, as far > as I can see in both of your examples the same method as me is used, > because %%javascript is just a alias for the rich display feature > display(Javascript(js)) > I think you're asking about how to get the JS part of the widget onto the page. Have a look at what I do with the turtle widget in mobilechelonian - when you instantiate it, it copies the JS into nbextensions, and the widget specifies _view_module to tell the frontend where to load its JS component from. https://github.com/takluyver/mobilechelonian Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From asmeurer at gmail.com Wed Nov 19 12:52:09 2014 From: asmeurer at gmail.com (Aaron Meurer) Date: Wed, 19 Nov 2014 11:52:09 -0600 Subject: [IPython-dev] Testing IPython 3.0 In-Reply-To: References: <1416399032654-5078073.post@n6.nabble.com> Message-ID: You'll probably want to do that in a new conda environment, so that you don't break your main environment conda create -n ipython-3.0 ipython-notebook source activate ipython-3.0 I installed ipython-notebook so that you get the dependencies of the notebook in the environment. Also, assumedly another option is python setup.py develop (you'll also need to install setuptools into the environment). Aaron Meurer On Wed, Nov 19, 2014 at 10:13 AM, Nicholas Bollweg wrote: > There aren't any binaries, yet, so I don't think you can conda install... > > The usual isn't too bad: shouldn't be any different than this: > >> >> http://ipython.org/ipython-doc/dev/install/install.html#installing-the-development-version >> > which is just: > > $ git clone --recursive https://github.com/ipython/ipython.git$ cd ipython$ python setup.py install > > If you want all the bells and whistles (notebook, tests, etc), i'd follow > it up with: > > $ pip install -e .[all] > > > On Wed, Nov 19, 2014 at 7:10 AM, thwiouz wrote: > >> Hi, >> >> I wanted to know how to install IPython 3.0 into a new environment using >> Anaconda. >> >> Thanks a lot >> >> >> >> -- >> View this message in context: >> http://python.6.x6.nabble.com/Testing-IPython-3-0-tp5078073.html >> Sent from the IPython - Development mailing list archive at Nabble.com. >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Wed Nov 19 13:13:46 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Wed, 19 Nov 2014 10:13:46 -0800 Subject: [IPython-dev] Testing IPython 3.0 In-Reply-To: References: <1416399032654-5078073.post@n6.nabble.com> Message-ID: On 19 November 2014 09:52, Aaron Meurer wrote: > Also, assumedly another option is python setup.py develop (you'll also > need to install setuptools into the environment). Or 'python setup.py symlink', which is specific to IPython. That doesn't require setuptools, but it won't work on Windows. We wrote it because 'setup.py develop' doesn't work if you want IPython in both Python 3 and 2 at the same time. Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From maidos93 at laposte.net Wed Nov 19 15:20:45 2014 From: maidos93 at laposte.net (thwiouz) Date: Wed, 19 Nov 2014 12:20:45 -0800 (PST) Subject: [IPython-dev] Testing IPython 3.0 In-Reply-To: References: <1416399032654-5078073.post@n6.nabble.com> Message-ID: <1416428445536-5078133.post@n6.nabble.com> Ok guys, thanks a lot! I'll try to do that! -- View this message in context: http://python.6.x6.nabble.com/Testing-IPython-3-0-tp5078073p5078133.html Sent from the IPython - Development mailing list archive at Nabble.com. From doug.blank at gmail.com Thu Nov 20 06:04:35 2014 From: doug.blank at gmail.com (Doug Blank) Date: Thu, 20 Nov 2014 06:04:35 -0500 Subject: [IPython-dev] Jupyterhub-based CMS? In-Reply-To: References: Message-ID: This conversation now continues over at the new Jupyter mailing list. Thanks! https://groups.google.com/forum/#!topic/jupyter/oe52-F0Bnuc -Doug On Fri, Oct 24, 2014 at 3:17 PM, Brian Granger wrote: > +1 on a jupyter list > > On Fri, Oct 24, 2014 at 11:54 AM, Fernando Perez > wrote: > > > > On Fri, Oct 24, 2014 at 9:16 AM, Doug Blank > wrote: > >> > >> I'd also love to discuss such a project with like-minded folks. A > million > >> questions and issues come to mind. I guess we can first see what kind of > >> interest there is, and then have some conversations on a separate > mailing > >> list. > > > > > > I think the time is coming to create a separate mailing list for Jupyter > as > > a project, where these discussions that really go well beyond 'IPython' > can > > take place. Now, something like this may even go one step further, where > a > > sub-community may emerge with its own projects around CMS-like ideas, and > > we'd love to see that. But it sounds like at least having a Jupyter area > for > > these conversations to launch from is a good idea. I've gotten similar > > feedback from the Julia folks. > > > > I'll start a separate thread on this. > > > > Cheers, > > > > f > > > > > > -- > > Fernando Perez (@fperez_org; http://fperez.org) > > fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) > > fernando.perez-at-berkeley: contact me here for any direct mail > > > > _______________________________________________ > > IPython-dev mailing list > > IPython-dev at scipy.org > > http://mail.scipy.org/mailman/listinfo/ipython-dev > > > > > > -- > Brian E. Granger > Cal Poly State University, San Luis Obispo > @ellisonbg on Twitter and GitHub > bgranger at calpoly.edu and ellisonbg at gmail.com > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jdemeyer at cage.ugent.be Thu Nov 20 10:04:41 2014 From: jdemeyer at cage.ugent.be (Jeroen Demeyer) Date: Thu, 20 Nov 2014 16:04:41 +0100 Subject: [IPython-dev] fix_frame_records_filenames() messes up Cython tracebacks Message-ID: <546E0309.9060707@cage.ugent.be> In IPython/core/ultratb.py, the fix_frame_records_filenames() tries to "correct" filenames in tracebacks. However, for tracebacks coming from Cython, this actually replaces the perfectly fine .pyx source filename by the compiled .so filename. So here is a suggestion: in fix_frame_records_filenames(), check that the better_fn() is not a binary file(*) and only then do the replacement. This would be easy to implement, any objections? I am asking this because Sage heavily uses Cython and this is annoying. (*) the standard way of doing this is to check for \x00 bytes. From bussonniermatthias at gmail.com Thu Nov 20 10:55:41 2014 From: bussonniermatthias at gmail.com (Matthias BUSSONNIER) Date: Thu, 20 Nov 2014 16:55:41 +0100 Subject: [IPython-dev] fix_frame_records_filenames() messes up Cython tracebacks In-Reply-To: <546E0309.9060707@cage.ugent.be> References: <546E0309.9060707@cage.ugent.be> Message-ID: <8058490D-BD09-4C75-9B1B-D91D0C317D72@gmail.com> Le 20 nov. 2014 ? 16:04, Jeroen Demeyer a ?crit : > In IPython/core/ultratb.py, the fix_frame_records_filenames() tries to > "correct" filenames in tracebacks. However, for tracebacks coming from > Cython, this actually replaces the perfectly fine .pyx source filename > by the compiled .so filename. > > So here is a suggestion: in fix_frame_records_filenames(), check that > the better_fn() is not a binary file(*) and only then do the > replacement. This would be easy to implement, any objections? > > I am asking this because Sage heavily uses Cython and this is annoying. I don't see any reason not to improve current logic. From your description, I suppose we could also whitelist extensions ? -- M > > > (*) the standard way of doing this is to check for \x00 bytes. > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev From jdemeyer at cage.ugent.be Thu Nov 20 11:18:12 2014 From: jdemeyer at cage.ugent.be (Jeroen Demeyer) Date: Thu, 20 Nov 2014 17:18:12 +0100 Subject: [IPython-dev] fix_frame_records_filenames() messes up Cython tracebacks In-Reply-To: <8058490D-BD09-4C75-9B1B-D91D0C317D72@gmail.com> References: <546E0309.9060707@cage.ugent.be> <8058490D-BD09-4C75-9B1B-D91D0C317D72@gmail.com> Message-ID: <546E1444.7000202@cage.ugent.be> On 2014-11-20 16:55, Matthias BUSSONNIER wrote: > From your description, I suppose we could also whitelist extensions ? Whitelisting/blacklisting is always more fragile. I prefer checking for binary files instead. It would also be good to understand better why the filename substitution is needed. Also interesting is that VerboseTB.structured_traceback() uses the source lines from the original filename [those lines are read in _fixed_getinnerframes()], not the changed filename. The changed filename is only used to "Build the list of names on this line of code where the exception occurred." (no idea what that means...) From jdemeyer at cage.ugent.be Thu Nov 20 11:42:34 2014 From: jdemeyer at cage.ugent.be (Jeroen Demeyer) Date: Thu, 20 Nov 2014 17:42:34 +0100 Subject: [IPython-dev] fix_frame_records_filenames() messes up Cython tracebacks In-Reply-To: <8058490D-BD09-4C75-9B1B-D91D0C317D72@gmail.com> References: <546E0309.9060707@cage.ugent.be> <8058490D-BD09-4C75-9B1B-D91D0C317D72@gmail.com> Message-ID: <546E19FA.8070907@cage.ugent.be> On 2014-11-20 16:55, Matthias BUSSONNIER wrote: > I don't see any reason not to improve current logic. Attached patch fixes the issue for me. -------------- next part -------------- A non-text attachment was scrubbed... Name: check_text_file.patch Type: text/x-patch Size: 1585 bytes Desc: not available URL: From bussonniermatthias at gmail.com Thu Nov 20 11:46:40 2014 From: bussonniermatthias at gmail.com (Matthias BUSSONNIER) Date: Thu, 20 Nov 2014 17:46:40 +0100 Subject: [IPython-dev] fix_frame_records_filenames() messes up Cython tracebacks In-Reply-To: <546E19FA.8070907@cage.ugent.be> References: <546E0309.9060707@cage.ugent.be> <8058490D-BD09-4C75-9B1B-D91D0C317D72@gmail.com> <546E19FA.8070907@cage.ugent.be> Message-ID: <3A6E1696-0C77-4C26-B91D-E52DE5F5BF68@gmail.com> Do you want to make a pull request on github ? Otherwise I can apply the patch myself and do a Pull request for you. -- M Le 20 nov. 2014 ? 17:42, Jeroen Demeyer a ?crit : > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jdemeyer at cage.ugent.be Thu Nov 20 11:51:32 2014 From: jdemeyer at cage.ugent.be (Jeroen Demeyer) Date: Thu, 20 Nov 2014 17:51:32 +0100 Subject: [IPython-dev] fix_frame_records_filenames() messes up Cython tracebacks In-Reply-To: <3A6E1696-0C77-4C26-B91D-E52DE5F5BF68@gmail.com> References: <546E0309.9060707@cage.ugent.be> <8058490D-BD09-4C75-9B1B-D91D0C317D72@gmail.com> <546E19FA.8070907@cage.ugent.be> <3A6E1696-0C77-4C26-B91D-E52DE5F5BF68@gmail.com> Message-ID: <546E1C14.6030707@cage.ugent.be> On 2014-11-20 17:46, Matthias BUSSONNIER wrote: > Do you want to make a pull request on github ? Yes, after some more testing. From maidos93 at laposte.net Thu Nov 20 11:54:12 2014 From: maidos93 at laposte.net (thwiouz) Date: Thu, 20 Nov 2014 08:54:12 -0800 (PST) Subject: [IPython-dev] Testing IPython 3.0 In-Reply-To: <1416428445536-5078133.post@n6.nabble.com> References: <1416399032654-5078073.post@n6.nabble.com> <1416428445536-5078133.post@n6.nabble.com> Message-ID: <1416502452825-5078246.post@n6.nabble.com> Thanks, it worked! I just needed to install jsonschema in the new environment too. Question: I saw in the documentation that it was possible to change the kernel of the notebook; (see http://ipython.org/ipython-doc/dev/whatsnew/development.html); knowing that R is installed (maybe the magic are not in the new env), how do I activate the R kernel to test? Thanks a lot (again) Thwiouz -- View this message in context: http://python.6.x6.nabble.com/Testing-IPython-3-0-tp5078073p5078246.html Sent from the IPython - Development mailing list archive at Nabble.com. From bussonniermatthias at gmail.com Thu Nov 20 12:11:30 2014 From: bussonniermatthias at gmail.com (Matthias BUSSONNIER) Date: Thu, 20 Nov 2014 18:11:30 +0100 Subject: [IPython-dev] fix_frame_records_filenames() messes up Cython tracebacks In-Reply-To: <546E1C14.6030707@cage.ugent.be> References: <546E0309.9060707@cage.ugent.be> <8058490D-BD09-4C75-9B1B-D91D0C317D72@gmail.com> <546E19FA.8070907@cage.ugent.be> <3A6E1696-0C77-4C26-B91D-E52DE5F5BF68@gmail.com> <546E1C14.6030707@cage.ugent.be> Message-ID: <8A449B1A-CC7D-409C-9264-86445CC4A3FE@gmail.com> Le 20 nov. 2014 ? 17:51, Jeroen Demeyer a ?crit : > On 2014-11-20 17:46, Matthias BUSSONNIER wrote: >> Do you want to make a pull request on github ? > Yes, after some more testing. The pull request will be (at least) auto tested by the CI server, and it will be easier for other dev to pitch in :-P, but if github is not in your workflow it's fine. -- M > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev From takowl at gmail.com Thu Nov 20 12:50:41 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Thu, 20 Nov 2014 09:50:41 -0800 Subject: [IPython-dev] Testing IPython 3.0 In-Reply-To: <1416502452825-5078246.post@n6.nabble.com> References: <1416399032654-5078073.post@n6.nabble.com> <1416428445536-5078133.post@n6.nabble.com> <1416502452825-5078246.post@n6.nabble.com> Message-ID: On 20 November 2014 08:54, thwiouz wrote: > knowing that > R is installed (maybe the magic are not in the new env), how do I activate > the R kernel to test? > You'll need IRkernel - instructions in the README here: https://github.com/takluyver/IRkernel/ Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From andrew.gibiansky at gmail.com Mon Nov 24 02:01:10 2014 From: andrew.gibiansky at gmail.com (Andrew Gibiansky) Date: Sun, 23 Nov 2014 23:01:10 -0800 Subject: [IPython-dev] Upgrading Kernels for IPython 3 Message-ID: Hey all, I'd like to update my IHaskell kernel for IPython 3 before IPython 3 is released. Sorry to bug this list, but I've found that I have a few questions that online resources / documentation aren't answering (or I'm not looking at the right resources): 1. When is IPython 3 slated to be out (and on PyPI)? I saw the roadmap on the wiki, which indicated Fall 2014 ? are there any more detailed plans? 2. Is there a list of things kernel authors need to do to update kernels? I've attempted to use the same kernel with IPython 3 and found that nothing works (my guess is there are several things in need of updating). 3. It seems that the new architecture assumes that kernels will install themselves into the ipython kernels directory, after which the user will invoke "ipython notebook" and choose the kernel. While I like this and would like to support this usage, historically I have shipped an "IHaskell" command which just ran IHaskell directly via "ipython --profile haskell" (along with doing a whole suite of other things). I would like to continue to provide this sort of mechanism to start IHaskell specifically, but it looks like "--profile" is deprecated (although it still appears in "--help"). What should I do? (Also, is "--profile-dir" also deprecated? It also appears in help and doesn't say it's deprecated.) Finally, if this information exists online somewhere, please link me to it ? I've googled around a bit, looked on the wiki, etc, and it doesn't seem like there's a document that gives me what I need. Thanks! Andrew -------------- next part -------------- An HTML attachment was scrubbed... URL: From hemmecke at gmail.com Mon Nov 24 06:58:43 2014 From: hemmecke at gmail.com (Ralf Hemmecke) Date: Mon, 24 Nov 2014 12:58:43 +0100 Subject: [IPython-dev] cursor invisible Message-ID: <54731D73.8010001@gmail.com> Hello, I'm using IPython3 (ipython notebook) and have opened a worksheet that is longer than one page (including output). When I start to re-evaluate (Shift-Enter) each cell and hit the bottom of the page, the cursor ends up in the next cell. But unfortunately, this cell does not appear in the visible area of my Firefox window. Any idea how Shift-Enter automatically brings the next cell into the visible area? Ralf From bussonniermatthias at gmail.com Mon Nov 24 11:15:06 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Mon, 24 Nov 2014 17:15:06 +0100 Subject: [IPython-dev] Upgrading Kernels for IPython 3 In-Reply-To: References: Message-ID: Hi Andrew, I'm on my phone so I'll stay short. Your kernel should work with IPython 3, we (should) detect and adapt older kernels. The custom js/css is not ready yet. Some other things are still in motion. Date is not fixed. The sooner the better. I, start to doubt there will be a release in the fall (maybe a beta). Kernels install at min a Json spec in .ipython/kernels/kernel name/kernel.json --profile-dir and notebook-dir should still work. I'll give you some links later. Cheers, -- M Envoy? de mon iPhone > Le 24 nov. 2014 ? 08:01, Andrew Gibiansky a ?crit : > > Hey all, > > I'd like to update my IHaskell kernel for IPython 3 before IPython 3 is released. Sorry to bug this list, but I've found that I have a few questions that online resources / documentation aren't answering (or I'm not looking at the right resources): > > 1. When is IPython 3 slated to be out (and on PyPI)? I saw the roadmap on the wiki, which indicated Fall 2014 ? are there any more detailed plans? > 2. Is there a list of things kernel authors need to do to update kernels? I've attempted to use the same kernel with IPython 3 and found that nothing works (my guess is there are several things in need of updating). > 3. It seems that the new architecture assumes that kernels will install themselves into the ipython kernels directory, after which the user will invoke "ipython notebook" and choose the kernel. While I like this and would like to support this usage, historically I have shipped an "IHaskell" command which just ran IHaskell directly via "ipython --profile haskell" (along with doing a whole suite of other things). I would like to continue to provide this sort of mechanism to start IHaskell specifically, but it looks like "--profile" is deprecated (although it still appears in "--help"). What should I do? (Also, is "--profile-dir" also deprecated? It also appears in help and doesn't say it's deprecated.) > > Finally, if this information exists online somewhere, please link me to it ? I've googled around a bit, looked on the wiki, etc, and it doesn't seem like there's a document that gives me what I need. > > Thanks! > Andrew > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev From takowl at gmail.com Mon Nov 24 13:28:32 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Mon, 24 Nov 2014 10:28:32 -0800 Subject: [IPython-dev] Upgrading Kernels for IPython 3 In-Reply-To: References: Message-ID: On 23 November 2014 at 23:01, Andrew Gibiansky wrote: > 1. When is IPython 3 slated to be out (and on PyPI)? I saw the roadmap on > the wiki, which indicated Fall 2014 ? are there any more detailed plans? > It will probably be early in the new year now. > 2. Is there a list of things kernel authors need to do to update kernels? > I've attempted to use the same kernel with IPython 3 and found that nothing > works (my guess is there are several things in need of updating). > There is a new version (5) of the message spec, but as Matthias mentioned, IPython should detect a kernel using the version 4 message spec and translate messages so it keeps working. Other than that, you should write a kernel spec, as described here: http://ipython.org/ipython-doc/dev/development/kernels.html#kernel-specs If you have custom JS overrides, they will almost certainly need to be updated, but there's no documentation of how to do that; the frontend interfaces are all changing much too fast to be worth documenting at present. If you have specific questions, feel free to ask us. We will also make a way to provide custom JS with a kernelspec for the kernel selection mechanism, but this isn't in place yet. > 3. It seems that the new architecture assumes that kernels will install > themselves into the ipython kernels directory, after which the user will > invoke "ipython notebook" and choose the kernel. While I like this and > would like to support this usage, historically I have shipped an "IHaskell" > command which just ran IHaskell directly via "ipython --profile haskell" > (along with doing a whole suite of other things). I would like to continue > to provide this sort of mechanism to start IHaskell specifically, but it > looks like "--profile" is deprecated (although it still appears in > "--help"). What should I do? (Also, is "--profile-dir" also deprecated? It > also appears in help and doesn't say it's deprecated.) > I don't remember that we've deprecated --profile, though we haven't decided yet what will happen with profiles when we separate IPython and Jupyter. Best wishes, Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From jdemeyer at cage.ugent.be Mon Nov 24 15:23:06 2014 From: jdemeyer at cage.ugent.be (Jeroen Demeyer) Date: Mon, 24 Nov 2014 21:23:06 +0100 Subject: [IPython-dev] Undoing transforms in tracebacks Message-ID: <547393AA.1050306@cage.ugent.be> Hello, For the Sage project, the IPython input transforms are important. The problem is that the use of these transforms can confuse users if an error occurs: the traceback shows the *transformed* line, not the line as inputted by the user. How would you feel about storing both the untransformed and transformed lines and displaying the *untransformed* lines in a traceback? I don't have code to do this yet, I am just asking whether you consider this a good idea in principle. Jeroen. From benjaminrk at gmail.com Mon Nov 24 15:31:08 2014 From: benjaminrk at gmail.com (MinRK) Date: Mon, 24 Nov 2014 12:31:08 -0800 Subject: [IPython-dev] Jupyterhub behind NGINX redirect In-Reply-To: References: Message-ID: Running JupyterHub behind nginx on a prefix should be working now after this PR . -MinRK ? On Mon, Nov 3, 2014 at 8:43 PM, MinRK wrote: > I suspect this is just a failure to handle and test the base_url properly. > I haven't used it behind a URL prefix yet, so I wouldn't be surprised if I > left out the prefix in a few places. > > Thanks for trying it out! > > On Sun, Nov 2, 2014 at 11:13 PM, Clare Sloggett > wrote: > >> Hi all, >> >> We've managed to troubleshoot our other issues, so I'm now sure that I >> can run jupyterhub successfully *without* a redirect. So I'm back to trying >> to get a redirect working. I've cc'd Nuwan who's been working on this with >> me and probably understands the issues better than I do. >> >> Currently, if I run something like >> $ jupyterhub --port 9520 ----JupyterHubApp.hub_port=8500 >> >> then I can access JupyterHub at http://my-url:9520/ . So far, success! >> This works fine, I can log in as various linux users, edit notebooks, etc. >> >> What we'd like is for it to instead be available to users at >> http://my-url/jupyter/ . >> >> We assumed that the command-line option JupyterHubApp.base_url is >> supposed to be used for this. We are trying to set up an NGINX redirect >> from /jupyter/ to :9520 and run jupyterhub so that it behaves properly in >> this situation. However JupyterHubApp.base_url doesn't seem to be doing >> what we expect it to. Is anyone able to explain how this parameter is >> supposed to be used? >> >> Cheers, >> Clare >> >> On 28 October 2014 18:15, Clare Sloggett wrote: >> >>> Hi Min, >>> >>> Thanks for your help! >>> >>> Actually at the moment I have been pushed back to a more fundamental >>> problem and can't even reproduce the error I was posting about in this >>> email. I can't get jupyterhub to run at all. I posted this separate error >>> in another thread, which I've just replied to a moment ago: "Error running >>> JupyterHub". >>> >>> I suspect the issue is mine as I encountered it a few days ago and am >>> still encountering it after updating to the latest commit. It seems >>> unlikely a bug has survived through several commits without someone else >>> discovering it too. But I'm having trouble working out what the cause is, >>> and am not sure what to make of the error messages it's throwing (both >>> python *and* javascript errors are thrown when it crashes). Full error >>> printout is in that thread! >>> >>> Cheers, >>> Clare >>> >>> On 26 October 2014 06:34, MinRK wrote: >>> >>>> Clare, >>>> >>>> Can you update to the latest master? I think the never-ending redirects >>>> could be the result of a recently fixed typo. >>>> >>>> Thanks, >>>> -MinRK >>>> >>>> On Thu, Oct 23, 2014 at 10:01 PM, Clare Sloggett < >>>> claresloggett at gmail.com> wrote: >>>> >>>>> Hi Doug, >>>>> >>>>> Thanks for this. I had actually just been thinking about about the >>>>> NGINX redirect issue, and had assumed everything behind that would be >>>>> fairly straightforward. But it sounds like you are saying there may be more >>>>> fundamental issues, and to be honest I haven't tested that the redirect is >>>>> definitely the source of all my problems. >>>>> >>>>> It sounds like I need to do some more direct testing and come back! >>>>> >>>>> In the meantime, if anyone has insight into what could be causing a >>>>> URL rewrite like "/ipython/ipython/hub/ipython/ >>>>> ipython/hub/ipython/ipython/hub/ipython/ipython...." that could be >>>>> really helpful. >>>>> >>>>> Cheers, >>>>> Clare >>>>> >>>>> On 13 October 2014 00:44, Doug Blank wrote: >>>>> >>>>>> On Sun, Oct 12, 2014 at 1:35 AM, Clare Sloggett < >>>>>> claresloggett at gmail.com> wrote: >>>>>> > >>>>>> > Hi all, >>>>>> > >>>>>> > Am I right in thinking this list is also the right place for >>>>>> questions about Jupyterhub? >>>>>> > >>>>>> > I'm trying to set up Jupyterhub for multiple users, on the same >>>>>> server where we are running several other services. Currently there are >>>>>> NGINX redirects in place to these other services. I'd like to set up http:///ipython/ >>>>>> to redirect to a port on localhost and run Jupyterhub over that port. Is >>>>>> this possible? >>>>>> > >>>>>> > I've got this working in the past with single-user IPython >>>>>> Notebook. For that, I set config options c.NotebookApp.base_project_url, >>>>>> c.NotebookApp.base_kernel_url, and c.NotebookApp.webapp_settings to be >>>>>> aware of the ipython/ URL prefix. We used an NGINX redirect to forward >>>>>> requests and handle websockets properly, which looked like >>>>>> > >>>>>> > location /ipython/ { >>>>>> > proxy_pass http://127.0.0.1:9510; >>>>>> > proxy_set_header Host $host; >>>>>> > proxy_set_header X-Real-IP $remote_addr; >>>>>> > proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; >>>>>> > proxy_set_header Upgrade $http_upgrade; >>>>>> > proxy_set_header Connection "upgrade"; >>>>>> > } >>>>>> > >>>>>> > This worked for the old single-user notebook. But, I'm not clear on >>>>>> the model Jupyterhub is using (and I'm not much of a sysadmin). I read >>>>>> through the command-line options, thought from them that I don't need to >>>>>> strip the /ipython/ from the incoming requests, and have tried using a >>>>>> similar redirect to the above and setting >>>>>> --JupyterHubApp.base_url='ipython/' and also setting --port 9510. This does >>>>>> seem to see the incoming requests but results in amusing requests like >>>>>> > >>>>>> > 500 GET >>>>>> /ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/...... >>>>>> > >>>>>> > ... so clearly I haven't understood what's going on. Does anyone >>>>>> have any pointers on how this should work? I haven't read any docs other >>>>>> than the README and the command-line parameter information, so apologies if >>>>>> there's something obvious that I didn't look at. >>>>>> > >>>>>> > I'm also interested in running Jupyterhub as a daemon, so if that's >>>>>> something that's been done before and there's anything I should know, that >>>>>> would be great. >>>>>> >>>>>> There is one other document, other than the README.md and the >>>>>> reported issues [1]: >>>>>> >>>>>> >>>>>> https://github.com/jupyter/jupyterhub/wiki/Using-sudo-to-run-the-server-as-non-root >>>>>> >>>>>> Despite the warning at the top of that page, you can make jupyterhub >>>>>> work on some systems (eg, Linux, such as Ubuntu) using sudo, but not as >>>>>> root. I suspect that this would be the recommended setup when jupyterhub is >>>>>> complete. One recent change is the ability to save/load state from a >>>>>> database. The wiki page above hasn't been updated with the note from this >>>>>> issue: >>>>>> >>>>>> https://github.com/jupyter/jupyterhub/issues/57 >>>>>> >>>>>> It might be easier to start without NGINX, and then add it after you >>>>>> have a working jupyterhub system. It would be nice to have a little bash >>>>>> script to make this a "service" that would support "start", "top", and >>>>>> "status"... but I haven't had time. Currently, we're just becoming the >>>>>> non-root sudoer ("rhea" in the docs) and starting the server, something >>>>>> along the lines: >>>>>> >>>>>> jupyterhub --LocalProcessSpawner.set_user=sudo >>>>>> --JupyterHubApp.ip=165.106.10.83 --JupyterHubApp.port=80 >>>>>> --db='sqlite:///:memory:' &>> /var/log/jupyterhub/log & >>>>>> >>>>>> The next step for us is to get it running under https... looks like >>>>>> others have blazed that trail, so it looks possible. >>>>>> >>>>>> If you have success, it would be great to add to the wiki docs... I >>>>>> suspect that many of us that aren't sys admins will be wanting to get this >>>>>> up and running. >>>>>> >>>>>> -Doug >>>>>> >>>>>> [1] - https://github.com/jupyter/jupyterhub >>>>>> >>>>>> > Any help much appreciated! >>>>>> > >>>>>> > Clare >>>>>> > >>>>>> > >>>>>> > _______________________________________________ >>>>>> > IPython-dev mailing list >>>>>> > IPython-dev at scipy.org >>>>>> > http://mail.scipy.org/mailman/listinfo/ipython-dev >>>>>> > >>>>>> >>>>>> _______________________________________________ >>>>>> IPython-dev mailing list >>>>>> IPython-dev at scipy.org >>>>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>>>> >>>>>> >>>>> >>>>> _______________________________________________ >>>>> IPython-dev mailing list >>>>> IPython-dev at scipy.org >>>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>>> >>>>> >>>> >>>> _______________________________________________ >>>> IPython-dev mailing list >>>> IPython-dev at scipy.org >>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>> >>>> >>> >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Mon Nov 24 15:49:12 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Mon, 24 Nov 2014 12:49:12 -0800 Subject: [IPython-dev] Undoing transforms in tracebacks In-Reply-To: <547393AA.1050306@cage.ugent.be> References: <547393AA.1050306@cage.ugent.be> Message-ID: On 24 November 2014 at 12:23, Jeroen Demeyer wrote: > How would you feel about storing both the untransformed and transformed > lines and displaying the *untransformed* lines in a traceback? > We already store both the transformed and untransformed code. For IPython's uses, I think it's more helpful to show the transformed code in tracebacks, so that it's clear what an error is actually about. But I wouldn't be against making it possible for Sphinx to override that and show untransformed code, if it can be done without adding much complexity. Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From andrew.gibiansky at gmail.com Mon Nov 24 21:48:03 2014 From: andrew.gibiansky at gmail.com (Andrew Gibiansky) Date: Mon, 24 Nov 2014 18:48:03 -0800 Subject: [IPython-dev] Upgrading Kernels for IPython 3 In-Reply-To: References: Message-ID: Thanks all. You've answered most of my questions, but I'm still a little unsure ? when I run IPython (v3), it starts up a IPython+Python kernel console (as expected). In previous IPythons, you'd ask for a different kernel via "--profile". Now, since the kernels directory and registry exists, the profiles will be less important; however, is there still a way to request a particular kernel at startup? (without using --profile?) In addition, I'd like to request/ask about an addition to the kernel directory. Could kernels ship their own nbconvert templates in addition to the custom.css/custom.js? If they could, I think that would eliminate completely my need to use IPython profiles for IHaskell, which would be a boon. Thanks, Andrew On Mon, Nov 24, 2014 at 10:28 AM, Thomas Kluyver wrote: > On 23 November 2014 at 23:01, Andrew Gibiansky > wrote: > >> 1. When is IPython 3 slated to be out (and on PyPI)? I saw the roadmap >> on the wiki, which indicated Fall 2014 ? are there any more detailed plans? >> > > It will probably be early in the new year now. > > >> 2. Is there a list of things kernel authors need to do to update kernels? >> I've attempted to use the same kernel with IPython 3 and found that nothing >> works (my guess is there are several things in need of updating). >> > > There is a new version (5) of the message spec, but as Matthias mentioned, > IPython should detect a kernel using the version 4 message spec and > translate messages so it keeps working. Other than that, you should write a > kernel spec, as described here: > http://ipython.org/ipython-doc/dev/development/kernels.html#kernel-specs > > If you have custom JS overrides, they will almost certainly need to be > updated, but there's no documentation of how to do that; the frontend > interfaces are all changing much too fast to be worth documenting at > present. If you have specific questions, feel free to ask us. We will also > make a way to provide custom JS with a kernelspec for the kernel selection > mechanism, but this isn't in place yet. > > >> 3. It seems that the new architecture assumes that kernels will install >> themselves into the ipython kernels directory, after which the user will >> invoke "ipython notebook" and choose the kernel. While I like this and >> would like to support this usage, historically I have shipped an "IHaskell" >> command which just ran IHaskell directly via "ipython --profile haskell" >> (along with doing a whole suite of other things). I would like to continue >> to provide this sort of mechanism to start IHaskell specifically, but it >> looks like "--profile" is deprecated (although it still appears in >> "--help"). What should I do? (Also, is "--profile-dir" also deprecated? It >> also appears in help and doesn't say it's deprecated.) >> > > I don't remember that we've deprecated --profile, though we haven't > decided yet what will happen with profiles when we separate IPython and > Jupyter. > > Best wishes, > Thomas > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From claresloggett at gmail.com Mon Nov 24 23:59:47 2014 From: claresloggett at gmail.com (Clare Sloggett) Date: Tue, 25 Nov 2014 15:59:47 +1100 Subject: [IPython-dev] Jupyterhub behind NGINX redirect In-Reply-To: References: Message-ID: Hi Min, Thank you, that's awesome. We will definitely try it out - I suspect I won't get a chance to this week, but probably will next week. I just need to upgrade to Python 3 :) Clare On 25 November 2014 at 07:31, MinRK wrote: > Running JupyterHub behind nginx on a prefix should be working now after this > PR . > > -MinRK > ? > > On Mon, Nov 3, 2014 at 8:43 PM, MinRK wrote: > >> I suspect this is just a failure to handle and test the base_url >> properly. I haven't used it behind a URL prefix yet, so I wouldn't be >> surprised if I left out the prefix in a few places. >> >> Thanks for trying it out! >> >> On Sun, Nov 2, 2014 at 11:13 PM, Clare Sloggett >> wrote: >> >>> Hi all, >>> >>> We've managed to troubleshoot our other issues, so I'm now sure that I >>> can run jupyterhub successfully *without* a redirect. So I'm back to trying >>> to get a redirect working. I've cc'd Nuwan who's been working on this with >>> me and probably understands the issues better than I do. >>> >>> Currently, if I run something like >>> $ jupyterhub --port 9520 ----JupyterHubApp.hub_port=8500 >>> >>> then I can access JupyterHub at http://my-url:9520/ . So far, success! >>> This works fine, I can log in as various linux users, edit notebooks, etc. >>> >>> What we'd like is for it to instead be available to users at >>> http://my-url/jupyter/ . >>> >>> We assumed that the command-line option JupyterHubApp.base_url is >>> supposed to be used for this. We are trying to set up an NGINX redirect >>> from /jupyter/ to :9520 and run jupyterhub so that it behaves properly in >>> this situation. However JupyterHubApp.base_url doesn't seem to be doing >>> what we expect it to. Is anyone able to explain how this parameter is >>> supposed to be used? >>> >>> Cheers, >>> Clare >>> >>> On 28 October 2014 18:15, Clare Sloggett >>> wrote: >>> >>>> Hi Min, >>>> >>>> Thanks for your help! >>>> >>>> Actually at the moment I have been pushed back to a more fundamental >>>> problem and can't even reproduce the error I was posting about in this >>>> email. I can't get jupyterhub to run at all. I posted this separate error >>>> in another thread, which I've just replied to a moment ago: "Error running >>>> JupyterHub". >>>> >>>> I suspect the issue is mine as I encountered it a few days ago and am >>>> still encountering it after updating to the latest commit. It seems >>>> unlikely a bug has survived through several commits without someone else >>>> discovering it too. But I'm having trouble working out what the cause is, >>>> and am not sure what to make of the error messages it's throwing (both >>>> python *and* javascript errors are thrown when it crashes). Full error >>>> printout is in that thread! >>>> >>>> Cheers, >>>> Clare >>>> >>>> On 26 October 2014 06:34, MinRK wrote: >>>> >>>>> Clare, >>>>> >>>>> Can you update to the latest master? I think the never-ending >>>>> redirects could be the result of a recently fixed typo. >>>>> >>>>> Thanks, >>>>> -MinRK >>>>> >>>>> On Thu, Oct 23, 2014 at 10:01 PM, Clare Sloggett < >>>>> claresloggett at gmail.com> wrote: >>>>> >>>>>> Hi Doug, >>>>>> >>>>>> Thanks for this. I had actually just been thinking about about the >>>>>> NGINX redirect issue, and had assumed everything behind that would be >>>>>> fairly straightforward. But it sounds like you are saying there may be more >>>>>> fundamental issues, and to be honest I haven't tested that the redirect is >>>>>> definitely the source of all my problems. >>>>>> >>>>>> It sounds like I need to do some more direct testing and come back! >>>>>> >>>>>> In the meantime, if anyone has insight into what could be causing a >>>>>> URL rewrite like "/ipython/ipython/hub/ipython/ >>>>>> ipython/hub/ipython/ipython/hub/ipython/ipython...." that could be >>>>>> really helpful. >>>>>> >>>>>> Cheers, >>>>>> Clare >>>>>> >>>>>> On 13 October 2014 00:44, Doug Blank wrote: >>>>>> >>>>>>> On Sun, Oct 12, 2014 at 1:35 AM, Clare Sloggett < >>>>>>> claresloggett at gmail.com> wrote: >>>>>>> > >>>>>>> > Hi all, >>>>>>> > >>>>>>> > Am I right in thinking this list is also the right place for >>>>>>> questions about Jupyterhub? >>>>>>> > >>>>>>> > I'm trying to set up Jupyterhub for multiple users, on the same >>>>>>> server where we are running several other services. Currently there are >>>>>>> NGINX redirects in place to these other services. I'd like to set up http:///ipython/ >>>>>>> to redirect to a port on localhost and run Jupyterhub over that port. Is >>>>>>> this possible? >>>>>>> > >>>>>>> > I've got this working in the past with single-user IPython >>>>>>> Notebook. For that, I set config options c.NotebookApp.base_project_url, >>>>>>> c.NotebookApp.base_kernel_url, and c.NotebookApp.webapp_settings to be >>>>>>> aware of the ipython/ URL prefix. We used an NGINX redirect to forward >>>>>>> requests and handle websockets properly, which looked like >>>>>>> > >>>>>>> > location /ipython/ { >>>>>>> > proxy_pass http://127.0.0.1:9510; >>>>>>> > proxy_set_header Host $host; >>>>>>> > proxy_set_header X-Real-IP $remote_addr; >>>>>>> > proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; >>>>>>> > proxy_set_header Upgrade $http_upgrade; >>>>>>> > proxy_set_header Connection "upgrade"; >>>>>>> > } >>>>>>> > >>>>>>> > This worked for the old single-user notebook. But, I'm not clear >>>>>>> on the model Jupyterhub is using (and I'm not much of a sysadmin). I read >>>>>>> through the command-line options, thought from them that I don't need to >>>>>>> strip the /ipython/ from the incoming requests, and have tried using a >>>>>>> similar redirect to the above and setting >>>>>>> --JupyterHubApp.base_url='ipython/' and also setting --port 9510. This does >>>>>>> seem to see the incoming requests but results in amusing requests like >>>>>>> > >>>>>>> > 500 GET >>>>>>> /ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/ipython/ipython/hub/...... >>>>>>> > >>>>>>> > ... so clearly I haven't understood what's going on. Does anyone >>>>>>> have any pointers on how this should work? I haven't read any docs other >>>>>>> than the README and the command-line parameter information, so apologies if >>>>>>> there's something obvious that I didn't look at. >>>>>>> > >>>>>>> > I'm also interested in running Jupyterhub as a daemon, so if >>>>>>> that's something that's been done before and there's anything I should >>>>>>> know, that would be great. >>>>>>> >>>>>>> There is one other document, other than the README.md and the >>>>>>> reported issues [1]: >>>>>>> >>>>>>> >>>>>>> https://github.com/jupyter/jupyterhub/wiki/Using-sudo-to-run-the-server-as-non-root >>>>>>> >>>>>>> Despite the warning at the top of that page, you can make jupyterhub >>>>>>> work on some systems (eg, Linux, such as Ubuntu) using sudo, but not as >>>>>>> root. I suspect that this would be the recommended setup when jupyterhub is >>>>>>> complete. One recent change is the ability to save/load state from a >>>>>>> database. The wiki page above hasn't been updated with the note from this >>>>>>> issue: >>>>>>> >>>>>>> https://github.com/jupyter/jupyterhub/issues/57 >>>>>>> >>>>>>> It might be easier to start without NGINX, and then add it after you >>>>>>> have a working jupyterhub system. It would be nice to have a little bash >>>>>>> script to make this a "service" that would support "start", "top", and >>>>>>> "status"... but I haven't had time. Currently, we're just becoming the >>>>>>> non-root sudoer ("rhea" in the docs) and starting the server, something >>>>>>> along the lines: >>>>>>> >>>>>>> jupyterhub --LocalProcessSpawner.set_user=sudo >>>>>>> --JupyterHubApp.ip=165.106.10.83 --JupyterHubApp.port=80 >>>>>>> --db='sqlite:///:memory:' &>> /var/log/jupyterhub/log & >>>>>>> >>>>>>> The next step for us is to get it running under https... looks like >>>>>>> others have blazed that trail, so it looks possible. >>>>>>> >>>>>>> If you have success, it would be great to add to the wiki docs... I >>>>>>> suspect that many of us that aren't sys admins will be wanting to get this >>>>>>> up and running. >>>>>>> >>>>>>> -Doug >>>>>>> >>>>>>> [1] - https://github.com/jupyter/jupyterhub >>>>>>> >>>>>>> > Any help much appreciated! >>>>>>> > >>>>>>> > Clare >>>>>>> > >>>>>>> > >>>>>>> > _______________________________________________ >>>>>>> > IPython-dev mailing list >>>>>>> > IPython-dev at scipy.org >>>>>>> > http://mail.scipy.org/mailman/listinfo/ipython-dev >>>>>>> > >>>>>>> >>>>>>> _______________________________________________ >>>>>>> IPython-dev mailing list >>>>>>> IPython-dev at scipy.org >>>>>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>>>>> >>>>>>> >>>>>> >>>>>> _______________________________________________ >>>>>> IPython-dev mailing list >>>>>> IPython-dev at scipy.org >>>>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>>>> >>>>>> >>>>> >>>>> _______________________________________________ >>>>> IPython-dev mailing list >>>>> IPython-dev at scipy.org >>>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>>> >>>>> >>>> >>> >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>> >>> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From kikocorreoso at gmail.com Tue Nov 25 03:03:42 2014 From: kikocorreoso at gmail.com (Kiko) Date: Tue, 25 Nov 2014 09:03:42 +0100 Subject: [IPython-dev] Upgrading Kernels for IPython 3 In-Reply-To: References: Message-ID: 2014-11-25 3:48 GMT+01:00 Andrew Gibiansky : > Thanks all. You've answered most of my questions, but I'm still a little > unsure ? when I run IPython (v3), it starts up a IPython+Python kernel > console (as expected). In previous IPythons, you'd ask for a different > kernel via "--profile". Now, since the kernels directory and registry > exists, the profiles will be less important; however, is there still a way > to request a particular kernel at startup? (without using --profile?) > See the "Running the notebook" section here: https://github.com/takluyver/IRkernel#running-the-notebook Or at the end of this notebook: http://nbviewer.ipython.org/github/ipython-books/cookbook-code/blob/master/notebooks/chapter01_basic/06_kernel.ipynb > > In addition, I'd like to request/ask about an addition to the kernel > directory. Could kernels ship their own nbconvert templates in addition to > the custom.css/custom.js? If they could, I think that would eliminate > completely my need to use IPython profiles for IHaskell, which would be a > boon. > > Thanks, > Andrew > > On Mon, Nov 24, 2014 at 10:28 AM, Thomas Kluyver wrote: > >> On 23 November 2014 at 23:01, Andrew Gibiansky < >> andrew.gibiansky at gmail.com> wrote: >> >>> 1. When is IPython 3 slated to be out (and on PyPI)? I saw the roadmap >>> on the wiki, which indicated Fall 2014 ? are there any more detailed plans? >>> >> >> It will probably be early in the new year now. >> >> >>> 2. Is there a list of things kernel authors need to do to update >>> kernels? I've attempted to use the same kernel with IPython 3 and found >>> that nothing works (my guess is there are several things in need of >>> updating). >>> >> >> There is a new version (5) of the message spec, but as Matthias >> mentioned, IPython should detect a kernel using the version 4 message spec >> and translate messages so it keeps working. Other than that, you should >> write a kernel spec, as described here: >> http://ipython.org/ipython-doc/dev/development/kernels.html#kernel-specs >> >> If you have custom JS overrides, they will almost certainly need to be >> updated, but there's no documentation of how to do that; the frontend >> interfaces are all changing much too fast to be worth documenting at >> present. If you have specific questions, feel free to ask us. We will also >> make a way to provide custom JS with a kernelspec for the kernel selection >> mechanism, but this isn't in place yet. >> >> >>> 3. It seems that the new architecture assumes that kernels will install >>> themselves into the ipython kernels directory, after which the user will >>> invoke "ipython notebook" and choose the kernel. While I like this and >>> would like to support this usage, historically I have shipped an "IHaskell" >>> command which just ran IHaskell directly via "ipython --profile haskell" >>> (along with doing a whole suite of other things). I would like to continue >>> to provide this sort of mechanism to start IHaskell specifically, but it >>> looks like "--profile" is deprecated (although it still appears in >>> "--help"). What should I do? (Also, is "--profile-dir" also deprecated? It >>> also appears in help and doesn't say it's deprecated.) >>> >> >> I don't remember that we've deprecated --profile, though we haven't >> decided yet what will happen with profiles when we separate IPython and >> Jupyter. >> >> Best wishes, >> Thomas >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From andrew.gibiansky at gmail.com Tue Nov 25 03:07:38 2014 From: andrew.gibiansky at gmail.com (Andrew Gibiansky) Date: Tue, 25 Nov 2014 00:07:38 -0800 Subject: [IPython-dev] Upgrading Kernels for IPython 3 In-Reply-To: References: Message-ID: Kiko, Just to clarify, I need a command that *does* work on IPython 3. I can already do this for IPython 2, and am looking for the equivalent once 3 is released. In both the links you provided, it says that in IPython 3 you have to select it from a dropdown, which does not work for me. -- Andrew On Tue, Nov 25, 2014 at 12:03 AM, Kiko wrote: > > > 2014-11-25 3:48 GMT+01:00 Andrew Gibiansky : > >> Thanks all. You've answered most of my questions, but I'm still a little >> unsure ? when I run IPython (v3), it starts up a IPython+Python kernel >> console (as expected). In previous IPythons, you'd ask for a different >> kernel via "--profile". Now, since the kernels directory and registry >> exists, the profiles will be less important; however, is there still a way >> to request a particular kernel at startup? (without using --profile?) >> > > See the "Running the notebook" section here: > https://github.com/takluyver/IRkernel#running-the-notebook > Or at the end of this notebook: > http://nbviewer.ipython.org/github/ipython-books/cookbook-code/blob/master/notebooks/chapter01_basic/06_kernel.ipynb > > >> >> In addition, I'd like to request/ask about an addition to the kernel >> directory. Could kernels ship their own nbconvert templates in addition to >> the custom.css/custom.js? If they could, I think that would eliminate >> completely my need to use IPython profiles for IHaskell, which would be a >> boon. >> >> Thanks, >> Andrew >> >> On Mon, Nov 24, 2014 at 10:28 AM, Thomas Kluyver >> wrote: >> >>> On 23 November 2014 at 23:01, Andrew Gibiansky < >>> andrew.gibiansky at gmail.com> wrote: >>> >>>> 1. When is IPython 3 slated to be out (and on PyPI)? I saw the roadmap >>>> on the wiki, which indicated Fall 2014 ? are there any more detailed plans? >>>> >>> >>> It will probably be early in the new year now. >>> >>> >>>> 2. Is there a list of things kernel authors need to do to update >>>> kernels? I've attempted to use the same kernel with IPython 3 and found >>>> that nothing works (my guess is there are several things in need of >>>> updating). >>>> >>> >>> There is a new version (5) of the message spec, but as Matthias >>> mentioned, IPython should detect a kernel using the version 4 message spec >>> and translate messages so it keeps working. Other than that, you should >>> write a kernel spec, as described here: >>> http://ipython.org/ipython-doc/dev/development/kernels.html#kernel-specs >>> >>> If you have custom JS overrides, they will almost certainly need to be >>> updated, but there's no documentation of how to do that; the frontend >>> interfaces are all changing much too fast to be worth documenting at >>> present. If you have specific questions, feel free to ask us. We will also >>> make a way to provide custom JS with a kernelspec for the kernel selection >>> mechanism, but this isn't in place yet. >>> >>> >>>> 3. It seems that the new architecture assumes that kernels will install >>>> themselves into the ipython kernels directory, after which the user will >>>> invoke "ipython notebook" and choose the kernel. While I like this and >>>> would like to support this usage, historically I have shipped an "IHaskell" >>>> command which just ran IHaskell directly via "ipython --profile haskell" >>>> (along with doing a whole suite of other things). I would like to continue >>>> to provide this sort of mechanism to start IHaskell specifically, but it >>>> looks like "--profile" is deprecated (although it still appears in >>>> "--help"). What should I do? (Also, is "--profile-dir" also deprecated? It >>>> also appears in help and doesn't say it's deprecated.) >>>> >>> >>> I don't remember that we've deprecated --profile, though we haven't >>> decided yet what will happen with profiles when we separate IPython and >>> Jupyter. >>> >>> Best wishes, >>> Thomas >>> >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>> >>> >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From kikocorreoso at gmail.com Tue Nov 25 03:18:24 2014 From: kikocorreoso at gmail.com (Kiko) Date: Tue, 25 Nov 2014 09:18:24 +0100 Subject: [IPython-dev] Upgrading Kernels for IPython 3 In-Reply-To: References: Message-ID: 2014-11-25 9:07 GMT+01:00 Andrew Gibiansky : > Kiko, > > Just to clarify, I need a command that *does* work on IPython 3. I can > already do this for IPython 2, and am looking for the equivalent once 3 is > released. In both the links you provided, it says that in IPython 3 you > have to select it from a dropdown, which does not work for me. > I created a dummy kernel and it works for me (I see the kernel in the dropdown menu, http://nbviewer.ipython.org/github/kikocorreoso/PyConES14_talk-Hacking_the_Notebook/blob/master/notebooks/09-Kernels____IPythonv3_only.ipynb#Creating-our-own-kernels) in the IPython 3.0.dev version. But yes, maybe a core developer can answer much better than me as I'm far from being an expert with IPython and there isn't much documentation about this. > > -- Andrew > > On Tue, Nov 25, 2014 at 12:03 AM, Kiko wrote: > >> >> >> 2014-11-25 3:48 GMT+01:00 Andrew Gibiansky : >> >>> Thanks all. You've answered most of my questions, but I'm still a little >>> unsure ? when I run IPython (v3), it starts up a IPython+Python kernel >>> console (as expected). In previous IPythons, you'd ask for a different >>> kernel via "--profile". Now, since the kernels directory and registry >>> exists, the profiles will be less important; however, is there still a way >>> to request a particular kernel at startup? (without using --profile?) >>> >> >> See the "Running the notebook" section here: >> https://github.com/takluyver/IRkernel#running-the-notebook >> Or at the end of this notebook: >> http://nbviewer.ipython.org/github/ipython-books/cookbook-code/blob/master/notebooks/chapter01_basic/06_kernel.ipynb >> >> >>> >>> In addition, I'd like to request/ask about an addition to the kernel >>> directory. Could kernels ship their own nbconvert templates in addition to >>> the custom.css/custom.js? If they could, I think that would eliminate >>> completely my need to use IPython profiles for IHaskell, which would be a >>> boon. >>> >>> Thanks, >>> Andrew >>> >>> On Mon, Nov 24, 2014 at 10:28 AM, Thomas Kluyver >>> wrote: >>> >>>> On 23 November 2014 at 23:01, Andrew Gibiansky < >>>> andrew.gibiansky at gmail.com> wrote: >>>> >>>>> 1. When is IPython 3 slated to be out (and on PyPI)? I saw the >>>>> roadmap on the wiki, which indicated Fall 2014 ? are there any more >>>>> detailed plans? >>>>> >>>> >>>> It will probably be early in the new year now. >>>> >>>> >>>>> 2. Is there a list of things kernel authors need to do to update >>>>> kernels? I've attempted to use the same kernel with IPython 3 and found >>>>> that nothing works (my guess is there are several things in need of >>>>> updating). >>>>> >>>> >>>> There is a new version (5) of the message spec, but as Matthias >>>> mentioned, IPython should detect a kernel using the version 4 message spec >>>> and translate messages so it keeps working. Other than that, you should >>>> write a kernel spec, as described here: >>>> http://ipython.org/ipython-doc/dev/development/kernels.html#kernel-specs >>>> >>>> If you have custom JS overrides, they will almost certainly need to be >>>> updated, but there's no documentation of how to do that; the frontend >>>> interfaces are all changing much too fast to be worth documenting at >>>> present. If you have specific questions, feel free to ask us. We will also >>>> make a way to provide custom JS with a kernelspec for the kernel selection >>>> mechanism, but this isn't in place yet. >>>> >>>> >>>>> 3. It seems that the new architecture assumes that kernels will >>>>> install themselves into the ipython kernels directory, after which the user >>>>> will invoke "ipython notebook" and choose the kernel. While I like this and >>>>> would like to support this usage, historically I have shipped an "IHaskell" >>>>> command which just ran IHaskell directly via "ipython --profile haskell" >>>>> (along with doing a whole suite of other things). I would like to continue >>>>> to provide this sort of mechanism to start IHaskell specifically, but it >>>>> looks like "--profile" is deprecated (although it still appears in >>>>> "--help"). What should I do? (Also, is "--profile-dir" also deprecated? It >>>>> also appears in help and doesn't say it's deprecated.) >>>>> >>>> >>>> I don't remember that we've deprecated --profile, though we haven't >>>> decided yet what will happen with profiles when we separate IPython and >>>> Jupyter. >>>> >>>> Best wishes, >>>> Thomas >>>> >>>> _______________________________________________ >>>> IPython-dev mailing list >>>> IPython-dev at scipy.org >>>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>>> >>>> >>> >>> _______________________________________________ >>> IPython-dev mailing list >>> IPython-dev at scipy.org >>> http://mail.scipy.org/mailman/listinfo/ipython-dev >>> >>> >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From bussonniermatthias at gmail.com Tue Nov 25 04:40:33 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Tue, 25 Nov 2014 10:40:33 +0100 Subject: [IPython-dev] Upgrading Kernels for IPython 3 In-Reply-To: References: Message-ID: Le 25 nov. 2014 ? 03:48, Andrew Gibiansky a ?crit : > Thanks all. You've answered most of my questions, but I'm still a little unsure ? when I run IPython (v3), it starts up a IPython+Python kernel console (as expected). In previous IPythons, you'd ask for a different kernel via "--profile". Now, since the kernels directory and registry exists, the profiles will be less important; however, is there still a way to request a particular kernel at startup? (without using --profile?) Yes, using --kernel > > In addition, I'd like to request/ask about an addition to the kernel directory. Could kernels ship their own nbconvert templates in addition to the custom.css/custom.js? I guess that makes sens to search for specific template/conf when you know the language of the notebook. Though I will suspect that it might need a significant modification of nbconvert. Could you open an issue on github for that, and this may be only a future iteration of IPython. -- M > If they could, I think that would eliminate completely my need to use IPython profiles for IHaskell, which would be a boon. > > Thanks, > Andrew > > On Mon, Nov 24, 2014 at 10:28 AM, Thomas Kluyver wrote: > On 23 November 2014 at 23:01, Andrew Gibiansky wrote: > 1. When is IPython 3 slated to be out (and on PyPI)? I saw the roadmap on the wiki, which indicated Fall 2014 ? are there any more detailed plans? > > It will probably be early in the new year now. > > 2. Is there a list of things kernel authors need to do to update kernels? I've attempted to use the same kernel with IPython 3 and found that nothing works (my guess is there are several things in need of updating). > > There is a new version (5) of the message spec, but as Matthias mentioned, IPython should detect a kernel using the version 4 message spec and translate messages so it keeps working. Other than that, you should write a kernel spec, as described here: http://ipython.org/ipython-doc/dev/development/kernels.html#kernel-specs > > If you have custom JS overrides, they will almost certainly need to be updated, but there's no documentation of how to do that; the frontend interfaces are all changing much too fast to be worth documenting at present. If you have specific questions, feel free to ask us. We will also make a way to provide custom JS with a kernelspec for the kernel selection mechanism, but this isn't in place yet. > > 3. It seems that the new architecture assumes that kernels will install themselves into the ipython kernels directory, after which the user will invoke "ipython notebook" and choose the kernel. While I like this and would like to support this usage, historically I have shipped an "IHaskell" command which just ran IHaskell directly via "ipython --profile haskell" (along with doing a whole suite of other things). I would like to continue to provide this sort of mechanism to start IHaskell specifically, but it looks like "--profile" is deprecated (although it still appears in "--help"). What should I do? (Also, is "--profile-dir" also deprecated? It also appears in help and doesn't say it's deprecated.) > > I don't remember that we've deprecated --profile, though we haven't decided yet what will happen with profiles when we separate IPython and Jupyter. > > Best wishes, > Thomas > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev -------------- next part -------------- An HTML attachment was scrubbed... URL: From one at kentran.net Tue Nov 25 05:44:33 2014 From: one at kentran.net (Kenneth Tran) Date: Tue, 25 Nov 2014 02:44:33 -0800 Subject: [IPython-dev] Can't execute cells if run IPython3 Notebook at port 8888 Message-ID: Here's a strange problem that I'm having. If I start the notebook server with either *ipython notebook* or *ipython notebook --port=8888*, then I cannot execute cells in the notebooks. The cursor jumps to the next cells as usual but the cells don't run and the numbers don't change. If I change the port to anything else, then the problem disappears. What are the possible causes of this issue? My last sync of IPython 3 from Github was a few days ago. -Ken -------------- next part -------------- An HTML attachment was scrubbed... URL: From bussonniermatthias at gmail.com Tue Nov 25 06:16:51 2014 From: bussonniermatthias at gmail.com (Matthias BUSSONNIER) Date: Tue, 25 Nov 2014 12:16:51 +0100 Subject: [IPython-dev] Can't execute cells if run IPython3 Notebook at port 8888 In-Reply-To: References: Message-ID: Le 25 nov. 2014 ? 11:44, Kenneth Tran a ?crit : > Here's a strange problem that I'm having. > > If I start the notebook server with either ipython notebook or ipython notebook --port=8888, then I cannot execute cells in the notebooks. The cursor jumps to the next cells as usual but the cells don't run and the numbers don't change. This happen if the web socket connexion never finish opening and the kernel does not appear to be started for the frontend. Does your browser javascript console say anything ? > > If I change the port to anything else, then the problem disappears. > > What are the possible causes of this issue? > > My last sync of IPython 3 from Github was a few days ago. > > -Ken > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev -------------- next part -------------- An HTML attachment was scrubbed... URL: From ndbecker2 at gmail.com Tue Nov 25 09:07:22 2014 From: ndbecker2 at gmail.com (Neal Becker) Date: Tue, 25 Nov 2014 09:07:22 -0500 Subject: [IPython-dev] print preview doesn't work Message-ID: running ipython-2.1.0 on fedora 20 linux. I wanted to print, and thought print preview would work for that. It does show a preview, but then after I edited the page further, and then try preview again, the preview keeps showing the earlier unedited version. I tried save/chktp, but same result. -- -- Those who don't understand recursion are doomed to repeat it From zvoros at gmail.com Tue Nov 25 09:12:14 2014 From: zvoros at gmail.com (=?windows-1252?Q?Zolt=E1n_V=F6r=F6s?=) Date: Tue, 25 Nov 2014 15:12:14 +0100 Subject: [IPython-dev] print preview doesn't work In-Reply-To: References: Message-ID: <54748E3E.9020503@gmail.com> Hi Neal, You should use nbconvert for turning a notebook into a printable pdf document. Hope this helps, Zolt?n On 11/25/2014 03:07 PM, Neal Becker wrote: > running ipython-2.1.0 on fedora 20 linux. > > I wanted to print, and thought print preview would work for that. It does show > a preview, but then after I edited the page further, and then try preview again, > the preview keeps showing the earlier unedited version. I tried save/chktp, but > same result. > From takowl at gmail.com Tue Nov 25 11:34:21 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Tue, 25 Nov 2014 08:34:21 -0800 Subject: [IPython-dev] Can't execute cells if run IPython3 Notebook at port 8888 In-Reply-To: References: Message-ID: Could be a js caching issue; using a non default port will avoid the files in your cache. Try hitting Ctrl f5 a few times. -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Tue Nov 25 13:19:46 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Tue, 25 Nov 2014 10:19:46 -0800 Subject: [IPython-dev] print preview doesn't work In-Reply-To: References: Message-ID: On 25 November 2014 at 06:07, Neal Becker wrote: > running ipython-2.1.0 on fedora 20 linux. > Can you try upgrading to the latest stable version (2.3.1) to see if the problem is already resolved? Or to master, for that matter. > I wanted to print, and thought print preview would work for that. It does > show > a preview, but then after I edited the page further, and then try preview > again, > the preview keeps showing the earlier unedited version. I tried > save/chktp, but > same result. > Is this print preview from IPython's menus, or the browser's print preview feature? Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Tue Nov 25 13:35:27 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Tue, 25 Nov 2014 10:35:27 -0800 Subject: [IPython-dev] Upgrading Kernels for IPython 3 In-Reply-To: References: Message-ID: On 25 November 2014 at 01:40, Matthias Bussonnier < bussonniermatthias at gmail.com> wrote: > Yes, using --kernel --kernel works for the Qt console and the two-process terminal frontend ('ipython console'), but not currently for the notebook. We're still working out what the notion of a default kernel should look like and where it should be set. See #6950: https://github.com/ipython/ipython/pull/6950 Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From lucienboillod at gmail.com Tue Nov 25 17:10:11 2014 From: lucienboillod at gmail.com (Lucien Boillod) Date: Tue, 25 Nov 2014 23:10:11 +0100 Subject: [IPython-dev] Customize the notebook Message-ID: <6825429B-A706-44E8-8B93-2970F19B3650@gmail.com> Hello all, I wonder if there is a way to really customize a notebook, other than just change the css. For example be able to change the title, put a image on the top etc ... Cheers, Lucien Boillod From fperez.net at gmail.com Tue Nov 25 17:50:30 2014 From: fperez.net at gmail.com (Fernando Perez) Date: Tue, 25 Nov 2014 14:50:30 -0800 Subject: [IPython-dev] Can't execute cells if run IPython3 Notebook at port 8888 In-Reply-To: References: Message-ID: It could also be a firewall issue, are you by any chance on Windows? If so, please check any antivirus/firewall you may have running. Sophos is notorious for causing those exact symptoms you report (it prevents the websocket connection from ever completing). On Tue, Nov 25, 2014 at 8:34 AM, Thomas Kluyver wrote: > Could be a js caching issue; using a non default port will avoid the files > in your cache. Try hitting Ctrl f5 a few times. > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -- Fernando Perez (@fperez_org; http://fperez.org) fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) fernando.perez-at-berkeley: contact me here for any direct mail -------------- next part -------------- An HTML attachment was scrubbed... URL: From takowl at gmail.com Tue Nov 25 20:33:27 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Tue, 25 Nov 2014 17:33:27 -0800 Subject: [IPython-dev] No development meeting this Thursday Message-ID: Thursday this week is the Thanksgiving holiday in North America, so our regular development meeting will not be taking place. Depending on how much is added on the agenda, we'll either have a meeting on Tuesday 2nd December, or leave it until the next regular meeting time, on Thursday 4th. In either case, we'll start at 9.30am in California, which is 1730 UTC. Have a good Thanksgiving if you're celebrating it, and a good Thursday otherwise. Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From one at kentran.net Wed Nov 26 11:22:41 2014 From: one at kentran.net (Kenneth Tran) Date: Wed, 26 Nov 2014 08:22:41 -0800 Subject: [IPython-dev] Can't execute cells if run IPython3 Notebook at port 8888 In-Reply-To: References: Message-ID: Thanks. Refreshing does resolve the issue. Fernando, I was actually running on Ubuntu. -Ken On Tue, Nov 25, 2014 at 2:50 PM, Fernando Perez wrote: > It could also be a firewall issue, are you by any chance on Windows? If > so, please check any antivirus/firewall you may have running. Sophos is > notorious for causing those exact symptoms you report (it prevents the > websocket connection from ever completing). > > On Tue, Nov 25, 2014 at 8:34 AM, Thomas Kluyver wrote: > >> Could be a js caching issue; using a non default port will avoid the >> files in your cache. Try hitting Ctrl f5 a few times. >> >> _______________________________________________ >> IPython-dev mailing list >> IPython-dev at scipy.org >> http://mail.scipy.org/mailman/listinfo/ipython-dev >> >> > > > -- > Fernando Perez (@fperez_org; http://fperez.org) > fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) > fernando.perez-at-berkeley: contact me here for any direct mail > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From roalexan at microsoft.com Wed Nov 26 13:08:06 2014 From: roalexan at microsoft.com (Robert Alexander) Date: Wed, 26 Nov 2014 18:08:06 +0000 Subject: [IPython-dev] method to call to perform a checkpoint/save in IPython API Message-ID: <1417025286849.40548@microsoft.com> Hi all. Is there a function that I can call in the IPython API (I'm using 2.3.0) to do a save/checkpoint? In other words, save the current notebook programmatically - using the same "path" as would happen if you used the GUI (File->Save and Checkpoint) to save. Thanks, Robert -------------- next part -------------- An HTML attachment was scrubbed... URL: From nick.bollweg at gmail.com Wed Nov 26 13:19:21 2014 From: nick.bollweg at gmail.com (Nicholas Bollweg) Date: Wed, 26 Nov 2014 13:19:21 -0500 Subject: [IPython-dev] Customize the notebook In-Reply-To: <6825429B-A706-44E8-8B93-2970F19B3650@gmail.com> References: <6825429B-A706-44E8-8B93-2970F19B3650@gmail.com> Message-ID: Lucien: Don't count out CSS: you can do very tricky things with it. Here's a way to change the logo from your custom.css in profile_default: #ipython_notebook::before{content:"My new Title"} #ipython_notebook img{display:none;} You can also use profiles... these let you configure almost everything: http://ipython.org/ipython-doc/dev/config/intro.html#example-config-file Alternately, you can instantiate the server yourself, and do anything, really: here's a pattern I used for overwriting just a little bit of the templates and changing the starting directory: https://gist.github.com/anonymous/682a65934c31b0998d0d Cheers, Nick On Tue, Nov 25, 2014 at 5:10 PM, Lucien Boillod wrote: > Hello all, > > I wonder if there is a way to really customize a notebook, other than just > change the css. For example be able to change the title, put a image on the > top etc ... > > Cheers, > Lucien Boillod > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: From roalexan at microsoft.com Wed Nov 26 14:12:49 2014 From: roalexan at microsoft.com (Robert Alexander) Date: Wed, 26 Nov 2014 19:12:49 +0000 Subject: [IPython-dev] method to call to perform a checkpoint/save in IPython API In-Reply-To: <1417025286849.40548@microsoft.com> References: <1417025286849.40548@microsoft.com> Message-ID: <1417029169586.56267@microsoft.com> FYI, a simplified version of what I'm trying to do is: from IPython.nbformat import current as nbf from IPython.html.services.notebooks import nbmanager as nbm nb = nbf.new_notebook() nbm.save_notebook() which give me the following error when running: AttributeError: 'module' object has no attribute 'save_notebook' I should also point out that I've configured my IPython notebook server to use a custom NotebookApp.notebook_manager_class which, among other methods, implements def save_notebook(self, model, name, path=''): this method is called by the auto-save functionality of the notebook server. What I'm trying to do is basically call an API that performs at auto-save on the notebook server. ________________________________ From: ipython-dev-bounces at scipy.org on behalf of Robert Alexander Sent: Wednesday, November 26, 2014 1:08 PM To: ipython-dev at scipy.org Subject: [IPython-dev] method to call to perform a checkpoint/save in IPython API Hi all. Is there a function that I can call in the IPython API (I'm using 2.3.0) to do a save/checkpoint? In other words, save the current notebook programmatically - using the same "path" as would happen if you used the GUI (File->Save and Checkpoint) to save. Thanks, Robert -------------- next part -------------- An HTML attachment was scrubbed... URL: From lucienboillod at gmail.com Wed Nov 26 15:17:57 2014 From: lucienboillod at gmail.com (Lucien Boillod) Date: Wed, 26 Nov 2014 21:17:57 +0100 Subject: [IPython-dev] Customize the notebook In-Reply-To: References: <6825429B-A706-44E8-8B93-2970F19B3650@gmail.com> Message-ID: <3A226FCB-66AB-4EB3-9A0F-694518DD1B67@gmail.com> > Lucien: > Don't count out CSS: you can do very tricky things with it. Here's a way to change the logo from your custom.css in profile_default: > > #ipython_notebook::before{content:"My new Title"} > #ipython_notebook img{display:none;} > Indeed it does perfectly the job > You can also use profiles... these let you configure almost everything: > http://ipython.org/ipython-doc/dev/config/intro.html#example-config-file > > Alternately, you can instantiate the server yourself, and do anything, really: here's a pattern I used for overwriting just a little bit of the templates and changing the starting directory: > https://gist.github.com/anonymous/682a65934c31b0998d0d > It?s really impressive I didn?t know there was that power of customization, thank you a lot, Cheers, Lucien Boillod -------------- next part -------------- An HTML attachment was scrubbed... URL: From bussonniermatthias at gmail.com Thu Nov 27 02:55:14 2014 From: bussonniermatthias at gmail.com (Matthias BUSSONNIER) Date: Thu, 27 Nov 2014 08:55:14 +0100 Subject: [IPython-dev] method to call to perform a checkpoint/save in IPython API In-Reply-To: <1417029169586.56267@microsoft.com> References: <1417025286849.40548@microsoft.com> <1417029169586.56267@microsoft.com> Message-ID: <643F6354-1D14-412D-B0DD-1173610F6493@gmail.com> Le 26 nov. 2014 ? 20:12, Robert Alexander a ?crit : > FYI, a simplified version of what I'm trying to do is: > from IPython.nbformat import current as nbf > from IPython.html.services.notebooks import nbmanager as nbm > > nb = nbf.new_notebook() > nbm.save_notebook() > which give me the following error when running: > AttributeError: 'module' object has no attribute 'save_notebook' > I should also point out that I've configured my IPython notebook server to use a custom > NotebookApp.notebook_manager_class which, among other methods, implements > def save_notebook(self, model, name, path=''): > this method is called by the auto-save functionality of the notebook server. What I'm trying to do is basically call an API that performs at auto-save on the notebook server. What you are looking for is create_checkpoint I believe. (on master at least, might have moved a bit) https://github.com/ipython/ipython/blob/b7c85f43feb2eaae3d9be358a4f8f1dac289e0ff/IPython/html/services/contents/manager.py#L160-L177 But be careful not to mistake call in python with call in javascript, the autosave in in javascritp on probably call save notebook here : https://github.com/ipython/ipython/blob/b7c85f43feb2eaae3d9be358a4f8f1dac289e0ff/IPython/html/services/contents/manager.py#L160-L177 Cheers, -- M > > From: ipython-dev-bounces at scipy.org on behalf of Robert Alexander > Sent: Wednesday, November 26, 2014 1:08 PM > To: ipython-dev at scipy.org > Subject: [IPython-dev] method to call to perform a checkpoint/save in IPython API > > Hi all. Is there a function that I can call in the IPython API (I'm using 2.3.0) to do a save/checkpoint? In other words, save the current notebook programmatically - using the same "path" as would happen if you used the GUI (File->Save and Checkpoint) to save. > > Thanks, Robert > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev -------------- next part -------------- An HTML attachment was scrubbed... URL: From sk.spiros at hotmail.com Fri Nov 28 08:59:02 2014 From: sk.spiros at hotmail.com (Hack-The-Paradise) Date: Fri, 28 Nov 2014 05:59:02 -0800 (PST) Subject: [IPython-dev] How-To create persistent, interactive UI elements outside cells Message-ID: <1417183142119-5079080.post@n6.nabble.com> Hello World, Me and my colleagues we are working on a project concerning big data handling, filtering and visualizing. We are trying to use IPython Notebook as our UI. We think that Notebook is a very nice and suitable tool. Our goal is to extend its functionality according to our needs. We have encountered some problems during this procedure and we tried to ask for help in https://gitter.im/ipython/ipython/help# but the answers were very limited, so i took the initiative to ask also here. We are trying to create (let?s say) a custom UI element and want to connect it with the server side (e.g. a help button with dynamic content in the toolbar). *Our main objective* is that we want to share this UI element on all our notebooks and load that on the startup of every notebook automatically. *So It should not be coupled with the cells*. A perfect example of what we are trying to achieve is the following: Let?s assume that we have what is described in this github repository: https://github.com/ipython-contrib/IPython-notebook-extensions/wiki/help_panel When the user clicks a button a list of help is shown on a panel on the right side of the screen. Currently, the data that populate this panel are static and what we want is to replace this by using iPython MVC widgets framework to have a model of these data stored on the notebook server. So that the list is synchronized with the model. *The code for this should not be typed into the cells but should be executed automatically upon loading of a notebook* (i.e. creating or opening an existing notebook). *Is there a functionality already built-in in ipython/notebook that implements something similar or we need to create/extend that?* Given the fact that we have limited time. Some of our ideas were: - Hiding specific cells with the code (also protecting them in order to emulate persistency) - execute JavaScript code from markdown cells (but I think this is not possible anymore or not?) i think that the following conversation is similar to our interest: http://python.6.x6.nabble.com/Feasilibity-of-widgets-layout-for-mimicking-this-td5068197.html The following image is also a nice example of what we want to accomplish but it is not complete. In that case the code is implemented in cells *P.S.:* Of course we already managed to build our own widgets following the given structure and using JavaScript but still we could not avoid using cells (at least for display command) We can provide code snippets if it is necessary. *Our default Env:* Anaconda 2.1.0 (x64) ? iPython 2.3.1 ? Python 2.7.8 -- View this message in context: http://python.6.x6.nabble.com/How-To-create-persistent-interactive-UI-elements-outside-cells-tp5079080.html Sent from the IPython - Development mailing list archive at Nabble.com. From bussonniermatthias at gmail.com Fri Nov 28 09:26:48 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Fri, 28 Nov 2014 15:26:48 +0100 Subject: [IPython-dev] How-To create persistent, interactive UI elements outside cells In-Reply-To: <1417183142119-5079080.post@n6.nabble.com> References: <1417183142119-5079080.post@n6.nabble.com> Message-ID: <3EB0767F-032E-4E0E-8E11-56586EB87A92@gmail.com> Hi, Please excuse the following explanation which is hand waved, I'll come back to you at a later point with more detail once I'm not traveling anymore. Feel free to up the question if we don't come back. What you want to do is to have a notebook extension (see github/IPython-contrib) and add a custom button to the main cellToolBar. See default custom.js for code sample that adds a button that open QTconsole. You want to grab a handle to the kernel (IPython.notebook.kernel) in JavaScript, and call it's execute method with your own callbacks. It seem that the route you took is more bending the current architecture where it's not supposed to be instead of reusing base bricks. The exact response to your question depends on wether you want data from the **server** or from the **kernel**. I believe you meant kernel, but it's not completely clear. -- M Envoy? de mon iPhone > Le 28 nov. 2014 ? 14:59, Hack-The-Paradise a ?crit : > > Hello World, > > Me and my colleagues we are working on a project concerning big data > handling, filtering and visualizing. We are trying to use IPython Notebook > as our UI. We think that Notebook is a very nice and suitable tool. > > Our goal is to extend its functionality according to our needs. > We have encountered some problems during this procedure and we tried to ask > for help in https://gitter.im/ipython/ipython/help# > but the answers were very > limited, so i took the initiative to ask also here. > > We are trying to create (let?s say) a custom UI element and want to connect > it with the server side (e.g. a help button with dynamic content in the > toolbar). > > *Our main objective* is that we want to share this UI element on all our > notebooks and load that on the startup of every notebook automatically. *So > It should not be coupled with the cells*. > > A perfect example of what we are trying to achieve is the following: > Let?s assume that we have what is described in this github repository: > https://github.com/ipython-contrib/IPython-notebook-extensions/wiki/help_panel > > When the user clicks a button a list of help is shown on a panel on the > right side of the screen. > Currently, the data that populate this panel are static and what we want is > to replace this by using iPython MVC widgets framework to have a model of > these data stored on the notebook server. So that the list is synchronized > with the model. *The code for this should not be typed into the cells but > should be executed automatically upon loading of a notebook* (i.e. creating > or opening an existing notebook). > > *Is there a functionality already built-in in ipython/notebook that > implements something similar or we need to create/extend that?* Given the > fact that we have limited time. > > Some of our ideas were: > - Hiding specific cells with the code (also protecting them in order to > emulate persistency) > - execute JavaScript code from markdown cells (but I think this is not > possible anymore or not?) > > i think that the following conversation is similar to our interest: > http://python.6.x6.nabble.com/Feasilibity-of-widgets-layout-for-mimicking-this-td5068197.html > > The following image is also a nice example of what we want to accomplish but > it is not complete. In that case the code is implemented in cells > > > *P.S.:* Of course we already managed to build our own widgets following the > given structure and using JavaScript but still we could not avoid using > cells (at least for display command) > We can provide code snippets if it is necessary. > > *Our default Env:* > Anaconda 2.1.0 (x64) ? iPython 2.3.1 ? Python 2.7.8 > > > > > -- > View this message in context: http://python.6.x6.nabble.com/How-To-create-persistent-interactive-UI-elements-outside-cells-tp5079080.html > Sent from the IPython - Development mailing list archive at Nabble.com. > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev From marcinofulus at gmail.com Fri Nov 28 09:40:36 2014 From: marcinofulus at gmail.com (Marcin Kostur) Date: Fri, 28 Nov 2014 15:40:36 +0100 Subject: [IPython-dev] Remote kernels Message-ID: Dear all, The kernel serving the interactive process connected with the notebook used to be run on the same machine as the notebook itself. We would like to run it remotely, in particular using HPC cluster with interactive queue, or ssh command to some server with computational resources. It would be something between IPython.parallel and local process serving the interactivity. Did anyone you considered such an option? The another question is - why ipython (or other) kernels starts at notebook load. In Sage notebook, for example, it starts when one launches the first computation and I like such setting much better. the best Marcin -- Department of Theoretical Physics Institute of Physics University of Silesia 40-007 Katowice, Poland tel. +48 32 3497 612 http://zft.us.edu.pl/kostur http://icse.us.edu.pl http://twing.us.edu.pl --------------------------- Linux: because rebooting is for adding new hardware. From sk.spiros at hotmail.com Fri Nov 28 16:05:10 2014 From: sk.spiros at hotmail.com (Hack-The-Paradise) Date: Fri, 28 Nov 2014 13:05:10 -0800 (PST) Subject: [IPython-dev] How-To create persistent, interactive UI elements outside cells In-Reply-To: <3EB0767F-032E-4E0E-8E11-56586EB87A92@gmail.com> References: <1417183142119-5079080.post@n6.nabble.com> <3EB0767F-032E-4E0E-8E11-56586EB87A92@gmail.com> Message-ID: <1417208710660-5079115.post@n6.nabble.com> I think that you did not get my point. But lets get into the details when you or someone else manage to have enough time. To be honest i also do not have more time today. For now, i would like to stress the title of the thread/topic and based on that & using the image example for having something more tangible to discuss i would like to ask again if you think that it's possible to achieve something similar (as the image) *without putting any code into the cells* of the Notebook. Goal: every time i open a notebook using a profile i want to see something like that (and interact with that) and yes i would like to use the existing base bricks (that's why i asked if there is something built in to use). I will come back with more details and possible snippets of code. -- View this message in context: http://python.6.x6.nabble.com/How-To-create-persistent-interactive-UI-elements-outside-cells-tp5079080p5079115.html Sent from the IPython - Development mailing list archive at Nabble.com. From jdemeyer at cage.ugent.be Sat Nov 29 04:27:44 2014 From: jdemeyer at cage.ugent.be (Jeroen Demeyer) Date: Sat, 29 Nov 2014 10:27:44 +0100 Subject: [IPython-dev] IPython test suite requires "requests" and "mock" Message-ID: <54799190.5070306@cage.ugent.be> Hello ipython-dev, When running the IPython test suite ("iptest"), I got ImportErrors for the "requests" and "mock" modules. Perhaps the page http://ipython.org/ipython-doc/rel-0.13.1/development/testing.html should mention somewhere in the beginning to execute first easy_install nose requests mock Regards, Jeroen. From bussonniermatthias at gmail.com Sat Nov 29 06:47:38 2014 From: bussonniermatthias at gmail.com (Matthias Bussonnier) Date: Sat, 29 Nov 2014 11:47:38 +0000 Subject: [IPython-dev] How-To create persistent, interactive UI elements outside cells In-Reply-To: <1417208710660-5079115.post@n6.nabble.com> References: <1417183142119-5079080.post@n6.nabble.com> <3EB0767F-032E-4E0E-8E11-56586EB87A92@gmail.com> <1417208710660-5079115.post@n6.nabble.com> Message-ID: <13063C4B-1E4A-409F-8070-5B0012142FE3@gmail.com> Hi, So back, from discussion. Le 28 nov. 2014 ? 21:05, Hack-The-Paradise a ?crit : > I think that you did not get my point. > But lets get into the details when you or someone else manage to have enough > time. To be honest i also do not have more time today. > > For now, i would like to stress the title of the thread/topic and based on > that & using the image example for having something more tangible to discuss > i would like to ask again if you think that it's possible to achieve > something similar (as the image) *without putting any code into the cells* > of the Notebook. Yes, and what you want is a nbextension. I took the help_pannel example you link to and modified it to execute code in the kernel. https://gist.github.com/Carreau/682b52d75dacc5ab4fbb the way js extension work, is you create a js file in ~/.ipython/nbextension/ and specify that IPython need to load it (either in custom.js, rennet IPython master you can make that in a config file) the js file is executed when the notebook load. In the case of help_pannel it adds a button to the toolbar. In this JS file you can have access to the the JS object that communicate to the kernel, as well as the DOM to modifie HTML. In the modified example I link form, I generate a button, when you click, this button as the kernel to execute "hi there", which basically just return a string. the Js handler receive thins truing and show it in the side panel. This does not use code in any cell the code is in the JS extension itself. In the example I only show how to handle execute request but you could hook-up anything like widgets... etc. You could also investigate user-expressions that can trigger code execution before user code. > > Goal: every time i open a notebook using a profile i want to see something > like that (and interact with that) and yes i would like to use the existing > base bricks (that's why i asked if there is something built in to use). Nb extension allow to share these extension in between profile, but yes you have to set a config option in the profile to activate the extension. Does that answer more your questions ? > > I will come back with more details and possible snippets of code. > > > > > -- > View this message in context: http://python.6.x6.nabble.com/How-To-create-persistent-interactive-UI-elements-outside-cells-tp5079080p5079115.html > Sent from the IPython - Development mailing list archive at Nabble.com. > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://mail.scipy.org/mailman/listinfo/ipython-dev -------------- next part -------------- An HTML attachment was scrubbed... URL: From jdemeyer at cage.ugent.be Sat Nov 29 07:33:30 2014 From: jdemeyer at cage.ugent.be (Jeroen Demeyer) Date: Sat, 29 Nov 2014 13:33:30 +0100 Subject: [IPython-dev] IPython test suite requires "requests" and "mock" In-Reply-To: <54799190.5070306@cage.ugent.be> References: <54799190.5070306@cage.ugent.be> Message-ID: <5479BD1A.70305@cage.ugent.be> On 2014-11-29 10:27, Jeroen Demeyer wrote: > Hello ipython-dev, > > When running the IPython test suite ("iptest"), I got ImportErrors for > the "requests" and "mock" modules. Perhaps the page > http://ipython.org/ipython-doc/rel-0.13.1/development/testing.html > should mention somewhere in the beginning to execute first > > easy_install nose requests mock When installing ipython, "requests" and "mock" should probably appear in the OPTIONAL DEPENDENCIES too, just like "nose": $ python setup.py install ============================================================================ BUILDING IPYTHON python: 2.7.8 (default, Nov 23 2014, 10:16:46) [GCC 4.6.4] platform: linux2 OPTIONAL DEPENDENCIES sphinx: 1.2.2 pygments: 1.3.1 nose: Not found (required for running the test suite) pexpect: 2.0 pyzmq: 14.3.0 tornado: 3.1.1 readline: yes jinja2: 2.5.5 From takowl at gmail.com Sat Nov 29 19:25:34 2014 From: takowl at gmail.com (Thomas Kluyver) Date: Sat, 29 Nov 2014 16:25:34 -0800 Subject: [IPython-dev] Tabipy Message-ID: Thanks to assistance from Christian Alis, I've just released a first version of Tabipy, a package to easily construct tables for display in the notebook: http://nbviewer.ipython.org/github/takluyver/tabipy/blob/master/docs/IPyTables.ipynb You can get it now with: pip install tabipy This serves a similar function to Eric Moyer's ipy_table package. However: - I don't especially like the API of ipy_table. Tabipy is designed in a more object oriented way, which I think is simpler and more flexible. Your mileage may vary. - Tabipy produces both HTML and Latex representations of tables, so a notebook converted to PDF via Latex has tables that look right (thanks to Christian for implementing this). - Tabipy doesn't currently have as many formatting options as ipy_table. Thanks, Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From sk.spiros at hotmail.com Sun Nov 30 11:26:53 2014 From: sk.spiros at hotmail.com (Hack-The-Paradise) Date: Sun, 30 Nov 2014 08:26:53 -0800 (PST) Subject: [IPython-dev] How-To create persistent, interactive UI elements outside cells In-Reply-To: <13063C4B-1E4A-409F-8070-5B0012142FE3@gmail.com> References: <1417183142119-5079080.post@n6.nabble.com> <3EB0767F-032E-4E0E-8E11-56586EB87A92@gmail.com> <1417208710660-5079115.post@n6.nabble.com> <13063C4B-1E4A-409F-8070-5B0012142FE3@gmail.com> Message-ID: <1417364813834-5079209.post@n6.nabble.com> OK, now we are getting somewhere but again it is not apparent to me how can i hook up a widget. And what exactly you mean by that. If you think that it is possible its good news whatsoever. But let me propose a more representative example: Assuming that we have the following notebooks: 1) http://nbviewer.ipython.org/gist/anonymous/840b5cec9e19f3e39090 2) http://nbviewer.ipython.org/github/rossant/euroscipy2014/blob/master/02_gui.ipynb 3) http://nbviewer.ipython.org/github/rossant/euroscipy2014/blob/master/01_interact.ipynb The 1) notebook has the above embedded image as an outcome. the other two can be found: https://github.com/rossant/euroscipy2014 Our goal is to transfer all the code that is written in the cells elsewhere and upon loading a notebook the same graphs and lists and buttons should appear somewhere. The point that we reached is relocating everything except the display command (as it can be seen in the image). But this is not the solution that we want. The first approach was to just copy paste the code that we have created in the cells inside the command IPython.notebook.kernel.execute(), as you can imagine the outcome was not displayed upon loading a new notebook. So, in this case do you think that the goal achievable? And if yes how ? If we can solve the problem for notebook 1) then i think we can do that also for 2) and 3). -- View this message in context: http://python.6.x6.nabble.com/How-To-create-persistent-interactive-UI-elements-outside-cells-tp5079080p5079209.html Sent from the IPython - Development mailing list archive at Nabble.com.