From laurent.dufrechou at free.fr Sun Dec 2 17:02:06 2007 From: laurent.dufrechou at free.fr (Laurent Dufrechou) Date: Sun, 2 Dec 2007 23:02:06 +0100 Subject: [IPython-dev] wxIPython Message-ID: <007101c8352e$fab12420$f0136c60$@dufrechou@free.fr> Hi everybody, I?ve played today with the gtk example from Eitan Isaacson. My objective was to do the same for Wxpython. Here is the result.(I have only tested it under linux currently, but should work on other platform!) You will need wxpython 2.8.7 and Ipython 0.8.2 (well I think Ipython can be lower) There still some bugs: ### known bug: # - write '?' + enter in the console make IPython behave strangely.(close stream???) # - UP and DOWN key doesn't behave well on multiline commands. # - currently WxConsoleView is a wx.Panel, can we change it to wx.TextCtrl? # - there is some graphical glitch on autocompletion test. How to hide this?? Disable onPaint?? (to check how much symbol I can put on a line I draw a line with the biggest autocompletion symbol and determine the max symbol on a line. Then I remove the line, here is the glitch.) I think it could be useful to guys who want to integrate IPython into wx apps and could help IPython to be better known J Code can be cleaner, but well it works J (especially the part with prompt length ) If you?re ok, I would like to propose it as a wx example. I haven?t so much time to maintain it so, if anybody would like to help improve it, I?m open. Laurent -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: wxIpython.zip Type: application/x-zip-compressed Size: 7349 bytes Desc: not available URL: From gael.varoquaux at normalesup.org Sun Dec 2 18:04:54 2007 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Mon, 3 Dec 2007 00:04:54 +0100 Subject: [IPython-dev] wxIPython Message-ID: <20071202230454.GA3072@clipper.ens.fr> On Sun, Dec 02, 2007 at 11:02:06PM +0100, Laurent Dufrechou wrote: > Hi everybody, Hi Laurent, Congratulations for that. We have currently started an effort to get a wxPython front end for Ipython1, the next version of Ipython. It lives in the Ipython1 svn: http://projects.scipy.org/ipython/ipython/browser/ipython1/trunk/ipython1/fronte nd The idea is to have a readline replacement, ipreadline, that can act as readline, but have several front ends. We have currently started the wxPython front-end, but to make all this general-enough, it would be a good idea to start other front ends. As none of us are good wxPython coder (unlike you, as it seems), and as we don't have much time, it is moving along very slowly. But we sure could use some help. Cheers, Ga?l From laurent.dufrechou at free.fr Sun Dec 2 18:57:08 2007 From: laurent.dufrechou at free.fr (Laurent Dufrechou) Date: Mon, 3 Dec 2007 00:57:08 +0100 Subject: [IPython-dev] wxIPython In-Reply-To: <20071202230454.GA3072@clipper.ens.fr> References: <20071202230454.GA3072@clipper.ens.fr> Message-ID: <009301c8353f$0c58d820$250a8860$@dufrechou@free.fr> I have just updated the code to support multiline history recall. (pushing up + enter will rerun a multiline call. With that I'm quite happy. I've found that it doesn't support ctrl+c, so I need to check this...I f anybody want to add it :) >Hi Laurent, > >Congratulations for that. > > >We have currently started an effort to get a wxPython front end for >Ipython1, the next version of Ipython. It lives in the Ipython1 svn: >http://projects.scipy.org/ipython/ipython/browser/ipython1/trunk/ipython1/f ronte >nd >The idea is to have a readline replacement, ipreadline, that can act as >readline, but have several front ends. We have currently started the >wxPython front-end, but to make all this general-enough, it would be a >good idea to start other front ends. > >As none of us are good wxPython coder (unlike you, as it seems), and as >we don't have much time, it is moving along very slowly. But we sure >could use some help. I've worked 6 month ago with ipython1, an excellent piece of code :) (for code execution on multiple machine) I'm happy to learn that you are trying to do a gui for it. Will take a look at your svn when possible :) Bye, Laurent -------------- next part -------------- A non-text attachment was scrubbed... Name: wxIpython_0.2.zip Type: application/x-zip-compressed Size: 7415 bytes Desc: not available URL: From vivainio at gmail.com Mon Dec 3 03:00:48 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Mon, 3 Dec 2007 10:00:48 +0200 Subject: [IPython-dev] wxIPython In-Reply-To: <4715758331733126211@unknownmsgid> References: <4715758331733126211@unknownmsgid> Message-ID: <46cb515a0712030000h74c9ecddi53f1fb6d67620397@mail.gmail.com> On Dec 3, 2007 12:02 AM, Laurent Dufrechou wrote: > Here is the result.(I have only tested it under linux currently, but should > work on other platform!) Worked fine on win32. A very cool hack! Can you easily remove (redundant) the standard interactive console from it? -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From vivainio at gmail.com Mon Dec 3 03:32:36 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Mon, 3 Dec 2007 10:32:36 +0200 Subject: [IPython-dev] wxIPython In-Reply-To: <46cb515a0712030000h74c9ecddi53f1fb6d67620397@mail.gmail.com> References: <4715758331733126211@unknownmsgid> <46cb515a0712030000h74c9ecddi53f1fb6d67620397@mail.gmail.com> Message-ID: <46cb515a0712030032n55423321k25bf21f20349b400@mail.gmail.com> On Dec 3, 2007 10:00 AM, Ville M. Vainio wrote: > Worked fine on win32. A very cool hack! In addition to other perks, one good idea would be a menu option to "Terminate current shell command". ctrl+c is a rather unpredictable way to go about it, but direct os.kill() operation can typically get the job done. -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From vivainio at gmail.com Mon Dec 3 05:07:11 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Mon, 3 Dec 2007 12:07:11 +0200 Subject: [IPython-dev] wxIPython In-Reply-To: <4753d190.25675e0a.0b48.ffff8c84@mx.google.com> References: <4715758331733126211@unknownmsgid> <46cb515a0712030000h74c9ecddi53f1fb6d67620397@mail.gmail.com> <46cb515a0712030032n55423321k25bf21f20349b400@mail.gmail.com> <4753d190.25675e0a.0b48.ffff8c84@mx.google.com> Message-ID: <46cb515a0712030207t2d36a273ife7921d8eec0e778@mail.gmail.com> On Dec 3, 2007 11:50 AM, Laurent Dufr?chou wrote: > Yep I can easily remove the old shell. Ok, do that then :-). > I'll take a look at the os.kill command. > I've tried: > "While(1): > A=1" > But in this case all resource are taken by the ipython execution and I can't > launch ctrl+c :). Yes, I wasn't thinking of interrupting python code with os.kill. I was thinking of terminating external (system) processes. To interrupt python execution, you could send SIGTERM to the process, perhaps? (unless wx eats it up...) > Perhaps a good way is to launch ipython in a other thread??? > How have you done this in your command shell? Running the IPython in other thread is more up the alley of IPython1... -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From jorgen.stenarson at bostream.nu Mon Dec 3 13:06:09 2007 From: jorgen.stenarson at bostream.nu (=?ISO-8859-1?Q?J=F6rgen_Stenarson?=) Date: Mon, 03 Dec 2007 19:06:09 +0100 Subject: [IPython-dev] wxIPython In-Reply-To: <20071202230454.GA3072@clipper.ens.fr> References: <20071202230454.GA3072@clipper.ens.fr> Message-ID: <47544591.5050503@bostream.nu> Gael Varoquaux skrev: > On Sun, Dec 02, 2007 at 11:02:06PM +0100, Laurent Dufrechou wrote: >> Hi everybody, > > Hi Laurent, > > Congratulations for that. > > We have currently started an effort to get a wxPython front end for > Ipython1, the next version of Ipython. It lives in the Ipython1 svn: > http://projects.scipy.org/ipython/ipython/browser/ipython1/trunk/ipython1/fronte > nd > Hi Ga?l, have you considered basing your effort on pyreadline? It should be possible to just write a new console class and some functions to handle the clipboard for wxpython. That way you won't have to reinvent all the other stuff. There are currently two console backends, the standard windows console and one for ironpython. However I have not tested the ironpython version in a while so it is probably broken now after the latest unicode changes. /J?rgen From gael.varoquaux at normalesup.org Tue Dec 4 04:10:05 2007 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Tue, 4 Dec 2007 10:10:05 +0100 Subject: [IPython-dev] wxIPython In-Reply-To: <47544591.5050503@bostream.nu> References: <20071202230454.GA3072@clipper.ens.fr> <47544591.5050503@bostream.nu> Message-ID: <20071204091005.GB21476@clipper.ens.fr> On Mon, Dec 03, 2007 at 07:06:09PM +0100, J?rgen Stenarson wrote: > have you considered basing your effort on pyreadline? No I hadn't thought about it, but you are definitly right about that. Next time I work on this I'll have a close look at pyreadline. Thanks for the tip. Ga?l From jorgen.stenarson at bostream.nu Tue Dec 4 13:13:48 2007 From: jorgen.stenarson at bostream.nu (=?ISO-8859-1?Q?J=F6rgen_Stenarson?=) Date: Tue, 04 Dec 2007 19:13:48 +0100 Subject: [IPython-dev] wxIPython In-Reply-To: <20071204091005.GB21476@clipper.ens.fr> References: <20071202230454.GA3072@clipper.ens.fr> <47544591.5050503@bostream.nu> <20071204091005.GB21476@clipper.ens.fr> Message-ID: <475598DC.3090606@bostream.nu> Gael Varoquaux skrev: > On Mon, Dec 03, 2007 at 07:06:09PM +0100, J?rgen Stenarson wrote: >> have you considered basing your effort on pyreadline? > > No I hadn't thought about it, but you are definitly right about that. > Next time I work on this I'll have a close look at pyreadline. Thanks for > the tip. > There is a patch in the pipeline for adding a callback based interface. Which may be necessary for your application. I'm planning to comit it soon. /J?rgen From gael.varoquaux at normalesup.org Tue Dec 4 16:03:15 2007 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Tue, 4 Dec 2007 22:03:15 +0100 Subject: [IPython-dev] wxIPython In-Reply-To: <475598DC.3090606@bostream.nu> References: <20071202230454.GA3072@clipper.ens.fr> <47544591.5050503@bostream.nu> <20071204091005.GB21476@clipper.ens.fr> <475598DC.3090606@bostream.nu> Message-ID: <20071204210315.GB3083@clipper.ens.fr> On Tue, Dec 04, 2007 at 07:13:48PM +0100, J?rgen Stenarson wrote: > There is a patch in the pipeline for adding a callback based interface. > Which may be necessary for your application. I'm planning to comit it soon. Great. I am really happy to have you on board. You have much more insight than me on these problems. Don't expect me to be very fast on this. On top of my commitment to too many day-work/foss activities, I am currently switching jobs and countries. Cheers, Ga?l From jorgen.stenarson at bostream.nu Tue Dec 4 16:39:16 2007 From: jorgen.stenarson at bostream.nu (=?ISO-8859-1?Q?J=F6rgen_Stenarson?=) Date: Tue, 04 Dec 2007 22:39:16 +0100 Subject: [IPython-dev] [Matplotlib-users] Bug in mathtext In-Reply-To: <4755AF7D.4070703@stsci.edu> References: <4755A612.4000505@bostream.nu> <4755AF7D.4070703@stsci.edu> Message-ID: <4755C904.8010508@bostream.nu> Michael Droettboom skrev: > Fernando Perez wrote: >> Hey Jorgen, >> >> On Dec 4, 2007 12:10 PM, J?rgen Stenarson wrote: >>> Hi, >>> >>> I get a complete crash dumping me in the console when there are unknown >>> latex commands in a mathtext expression, this when doing interactive >>> stuff in ipython -pylab. >>> >>> examples: >>> title("$|S_{11}|$") >> that's odd, it's OK for mpl to throw the exception, what shouldn't be >> happening is for ipython to fully crash out. I can't reproduce it >> with SVN mpl on my box, I tried both tkagg and gtkagg as backends and >> in both cases I see the exception traceback (as Michael intended by >> raising the error) but I simply get back the next ipython prompt, as >> usual. I don't understand how this particular exception could crash >> out ipython, since it's being raised inside regular user code... > > You're right -- I misread the original report. It doesn't crash ipython > for me either. I just get another ipython prompt. > > Cheers, > Mike > Thanks Mike. I hope we can figure out what is causing the crashes but I guess we can take this over on the ipython-list. Fernando have you tried it on a windows machine? I'm using tkagg. Do you have any ideas on what I could start testing to isolate the problem? /J?rgen From fperez.net at gmail.com Tue Dec 4 16:43:49 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Tue, 4 Dec 2007 14:43:49 -0700 Subject: [IPython-dev] [Matplotlib-users] Bug in mathtext In-Reply-To: <4755C904.8010508@bostream.nu> References: <4755A612.4000505@bostream.nu> <4755AF7D.4070703@stsci.edu> <4755C904.8010508@bostream.nu> Message-ID: On Dec 4, 2007 2:39 PM, J?rgen Stenarson wrote: > Thanks Mike. I hope we can figure out what is causing the crashes but I > guess we can take this over on the ipython-list. Fernando have you tried > it on a windows machine? I'm using tkagg. Do you have any ideas on what > I could start testing to isolate the problem? I'm taking the mpl list off, this is our problem :) No, I only have linux here in the office. It's even stranger that you can crash it with Tk, since that one doesn't even have any of the black magic threading stuff. Why don't you try starting with a clean config (temporarily rename your _ipython directory to force the creation of a new, clean one). I want to sort out whether it may be something triggered by your personal config. I'll try to test it tonight at home, I have VMWare windows on my desktop. Cheers, f From Laurent.Dufrechou at free.fr Wed Dec 5 08:37:05 2007 From: Laurent.Dufrechou at free.fr (Laurent Dufrechou) Date: Wed, 5 Dec 2007 05:37:05 -0800 (PST) Subject: [IPython-dev] wxIPython In-Reply-To: <475598DC.3090606@bostream.nu> References: <007101c8352e$fab12420$f0136c60$@dufrechou@free.fr> <20071202230454.GA3072@clipper.ens.fr> <47544591.5050503@bostream.nu> <20071204091005.GB21476@clipper.ens.fr> <475598DC.3090606@bostream.nu> Message-ID: <14171722.post@talk.nabble.com> Hi jorgen,gael Can you explain briefly what for pyreadline is used inside ipython? Has I've hacked a little ipython shell, perhaps i've hacked some things that could be done directly by pyreadline... Cheers, Laurent J?rgen Stenarson-2 wrote: > > Gael Varoquaux skrev: >> On Mon, Dec 03, 2007 at 07:06:09PM +0100, J?rgen Stenarson wrote: >>> have you considered basing your effort on pyreadline? >> >> No I hadn't thought about it, but you are definitly right about that. >> Next time I work on this I'll have a close look at pyreadline. Thanks for >> the tip. >> > > There is a patch in the pipeline for adding a callback based interface. > Which may be necessary for your application. I'm planning to comit it > soon. > > /J?rgen > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://lists.ipython.scipy.org/mailman/listinfo/ipython-dev > > -- View this message in context: http://www.nabble.com/wxIPython-tf4933380.html#a14171722 Sent from the IPython - Development mailing list archive at Nabble.com. From Laurent.Dufrechou at free.fr Wed Dec 5 08:51:25 2007 From: Laurent.Dufrechou at free.fr (Laurent Dufrechou) Date: Wed, 5 Dec 2007 05:51:25 -0800 (PST) Subject: [IPython-dev] Need help from Ipython guru... Message-ID: <14171949.post@talk.nabble.com> Hi IPythoners, I?m trying to do a litttle wx widget for ipyton.(I?ve posted some version on the list last day) Currently all is working quite well, but I?m stuck to a bug that I can?t handle. If I write: ?import smtplib? ?smtplib.re?? Then ipython shell call the ? magic function and because the text is too long, it is stuck to a getch inside genutils? >Term.cout.write('---Return to continue, q to quit--- ') > ans = msvcrt.getch() Too bad I can?t do anything. Is there a way to say to ipython: ?Hey guy! Give me the full text I will split it by myself and please don?t ask me to push a key!? Perhaps there is a secret var to set the line before breaking text? Or pehaps I couls define my pager? But then how to say to ipython, that my pager is this one...? I've seen that previous guy set os.environ['TERM'] = 'dumb' if this the reason why dump_pager is called? Other question: When i go back in history, previous '?' key appears like this: ex: smtplib? appears as: "#?smtplib" Is the # ok? if yes, is there any reason why the history return that? (to get history I do: rv = self.IP.user_ns['In'][self.history_level].strip('\n')) Another question for gurus: Currently to embbed the Ipython shell, I?ve reused the code from gtk Ipython widget guy: ? self.term = IPython.genutils.IOTerm(cin=cin, cout=cout, cerr=cerr) os.environ['TERM'] = 'dumb' excepthook = sys.excepthook self.IP = IPython.Shell.make_IPython( argv,user_ns=user_ns, user_global_ns=user_global_ns, embedded=True, shell_class=IPython.Shell.InteractiveShell) self.IP.system = lambda cmd: self.shell(self.IP.var_expand(cmd), header='IPython system call: ', verbose=self.IP.rc.system_verbose) ? Is this the best way to embed ipython nowadays. I mean I?ve seen that there is some classes: IPShellEmbed for example. Currently the design of the widget is limited by the fact that the ipython shell is not in a different process or thread, so if it locks, it locks all the app? Is there a best way to do embed an ipython shell then my current way? Cheers, Laurent -- View this message in context: http://www.nabble.com/Need-help-from-Ipython-guru...-tf4949781.html#a14171949 Sent from the IPython - Development mailing list archive at Nabble.com. From fperez.net at gmail.com Wed Dec 5 12:46:58 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Wed, 5 Dec 2007 10:46:58 -0700 Subject: [IPython-dev] Need help from Ipython guru... In-Reply-To: <14171949.post@talk.nabble.com> References: <14171949.post@talk.nabble.com> Message-ID: Hey Laurent, sorry for a rather curt reply right now. I'm really happy to see you doing this, so thanks a lot! I'm extremely swamped with a python workshop deadline fast approaching, so I won't be able to properly help until next week. Very briefly... On Dec 5, 2007 6:51 AM, Laurent Dufrechou wrote: > > Hi IPythoners, > > I'm trying to do a litttle wx widget for ipyton.(I've posted some version on > the list last day) > Currently all is working quite well, but I'm stuck to a bug that I can't > handle. > If I write: > "import smtplib" > "smtplib.re?" > Then ipython shell call the ? magic function and because the text is too > long, it is stuck to a getch inside genutils? > >Term.cout.write('---Return to continue, q to quit--- ') > > ans = msvcrt.getch() > > Too bad I can't do anything. > Is there a way to say to ipython: > "Hey guy! Give me the full text I will split it by myself and please don't > ask me to push a key!" > Perhaps there is a secret var to set the line before breaking text? > Or pehaps I couls define my pager? But then how to say to ipython, that my > pager is this one...? > I've seen that previous guy set os.environ['TERM'] = 'dumb' if this the > reason why dump_pager is called? Yes, but you can define your own pager via environ['PAGER']. That should allow you to give ipython your own paging routine. Look in genutils for the pager code... > Other question: > When i go back in history, previous '?' key appears like this: > ex: smtplib? > appears as: "#?smtplib" > Is the # ok? if yes, is there any reason why the history return that? > (to get history I do: rv = > self.IP.user_ns['In'][self.history_level].strip('\n')) Look at the 'raw history' instead, which has exactly what was typed without translation: In [7]: print __IP.input_hist_raw ['\n', u'x=1\n', u'x?\n', u'In\n', u'__IP.raw_input\n', u'__IP.*hist*?\n', u'__IP.input_hist_raw\n', u'print __IP.input_hist_raw\n'] > Another question for gurus: > Currently to embbed the Ipython shell, I've reused the code from gtk Ipython > widget guy: > " > self.term = IPython.genutils.IOTerm(cin=cin, cout=cout, cerr=cerr) > os.environ['TERM'] = 'dumb' > excepthook = sys.excepthook > self.IP = IPython.Shell.make_IPython( > argv,user_ns=user_ns, > user_global_ns=user_global_ns, > embedded=True, > shell_class=IPython.Shell.InteractiveShell) > self.IP.system = lambda cmd: self.shell(self.IP.var_expand(cmd), > header='IPython system call: ', > > verbose=self.IP.rc.system_verbose) " > > Is this the best way to embed ipython nowadays. > I mean I've seen that there is some classes: > IPShellEmbed for example. > Currently the design of the widget is limited by the fact that the ipython > shell is not in a different process or thread, so if it locks, it locks all > the app? > Is there a best way to do embed an ipython shell then my current way? Yup, that's the stuff which is currently in raw, in-progress mode in the ipython1 repo, that Gael mentioned. I hope to have time to finish it in a few months, but it won't be until early next year at best. This end-of-year is already booked solid with a py4science workshop and a numpy/scipy sprint. In the meantime, it may be possible to finish up your current approach to the point of usability, while we complete that work. But again, many thanks for pitching in with this, and keep bugging us. We'll reply :) Cheers, f From jorgen.stenarson at bostream.nu Wed Dec 5 14:19:35 2007 From: jorgen.stenarson at bostream.nu (=?ISO-8859-1?Q?J=F6rgen_Stenarson?=) Date: Wed, 05 Dec 2007 20:19:35 +0100 Subject: [IPython-dev] [Matplotlib-users] Bug in mathtext In-Reply-To: <4755C904.8010508@bostream.nu> References: <4755A612.4000505@bostream.nu> <4755AF7D.4070703@stsci.edu> <4755C904.8010508@bostream.nu> Message-ID: <4756F9C7.2000405@bostream.nu> J?rgen Stenarson skrev: >> > > Thanks Mike. I hope we can figure out what is causing the crashes but I > guess we can take this over on the ipython-list. Fernando have you tried > it on a windows machine? I'm using tkagg. Do you have any ideas on what > I could start testing to isolate the problem? > > /J?rgen Fernando, I found the problem. I have a pylab.show() in my generate_prompt hook to get automatic refresh after calling scripts. Anyway it seems exceptions raised here crash ipython. I don't know if this is intentional or not. I just put a try block around it as a workaround. /J?rgen From fperez.net at gmail.com Wed Dec 5 14:34:18 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Wed, 5 Dec 2007 12:34:18 -0700 Subject: [IPython-dev] [Matplotlib-users] Bug in mathtext In-Reply-To: <4756F9C7.2000405@bostream.nu> References: <4755A612.4000505@bostream.nu> <4755AF7D.4070703@stsci.edu> <4755C904.8010508@bostream.nu> <4756F9C7.2000405@bostream.nu> Message-ID: On Dec 5, 2007 12:19 PM, J?rgen Stenarson wrote: > J?rgen Stenarson skrev: > >> > > > > Thanks Mike. I hope we can figure out what is causing the crashes but I > > guess we can take this over on the ipython-list. Fernando have you tried > > it on a windows machine? I'm using tkagg. Do you have any ideas on what > > I could start testing to isolate the problem? > > > > /J?rgen > > > Fernando, I found the problem. I have a pylab.show() in my > generate_prompt hook to get automatic refresh after calling scripts. > Anyway it seems exceptions raised here crash ipython. I don't know if > this is intentional or not. I just put a try block around it as a > workaround. OK, that makes sense now. I think it qualifies as a bug in our part, since hooks should always be protected. They should generate the traceback, but not crash out. Trac-ticket it, baby :) [ aside] BTW, when I first wrote the pylab support, I added an automatic show() at the end. I later removed it to force users (including myself, since I use this all day long) to put a pylab.show() as the last line of my scripts. Doing this ensures that the script actually runs inside or outside of ipython equally well, without funny surprises when you send it to a colleague (who doesn't use ipython) and they see no plots :) So now I've just become used to always finishing all my scripts with the show() call. That way, if there's an exception in there, it still gets handled just like any other exception. [/aside] Cheers, f From jorgen.stenarson at bostream.nu Wed Dec 5 15:00:23 2007 From: jorgen.stenarson at bostream.nu (=?ISO-8859-1?Q?J=F6rgen_Stenarson?=) Date: Wed, 05 Dec 2007 21:00:23 +0100 Subject: [IPython-dev] [Matplotlib-users] Bug in mathtext In-Reply-To: References: <4755A612.4000505@bostream.nu> <4755AF7D.4070703@stsci.edu> <4755C904.8010508@bostream.nu> <4756F9C7.2000405@bostream.nu> Message-ID: <47570357.1020500@bostream.nu> Fernando Perez skrev: > On Dec 5, 2007 12:19 PM, J?rgen Stenarson wrote: >> J?rgen Stenarson skrev: >>> Thanks Mike. I hope we can figure out what is causing the crashes but I >>> guess we can take this over on the ipython-list. Fernando have you tried >>> it on a windows machine? I'm using tkagg. Do you have any ideas on what >>> I could start testing to isolate the problem? >>> >>> /J?rgen >> >> Fernando, I found the problem. I have a pylab.show() in my >> generate_prompt hook to get automatic refresh after calling scripts. >> Anyway it seems exceptions raised here crash ipython. I don't know if >> this is intentional or not. I just put a try block around it as a >> workaround. > > OK, that makes sense now. I think it qualifies as a bug in our part, > since hooks should always be protected. They should generate the > traceback, but not crash out. > > Trac-ticket it, baby :) > Done #202 > [ aside] BTW, when I first wrote the pylab support, I added an > automatic show() at the end. I later removed it to force users > (including myself, since I use this all day long) to put a > > pylab.show() > > as the last line of my scripts. Doing this ensures that the script > actually runs inside or outside of ipython equally well, without funny > surprises when you send it to a colleague (who doesn't use ipython) > and they see no plots :) So now I've just become used to always > finishing all my scripts with the show() call. That way, if there's > an exception in there, it still gets handled just like any other > exception. > [/aside] > That is certainly a good point. But I usually have a script that saves my figure to a file as well as showing it. And I don't want it to block when running as a standalone script but I want to see them when running from ipython. /J?rgen From vivainio at gmail.com Thu Dec 6 07:33:03 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Thu, 6 Dec 2007 14:33:03 +0200 Subject: [IPython-dev] Need help from Ipython guru... In-Reply-To: <14171949.post@talk.nabble.com> References: <14171949.post@talk.nabble.com> Message-ID: <46cb515a0712060433g107267ddp5cad42ce5ea4aa29@mail.gmail.com> On Dec 5, 2007 3:51 PM, Laurent Dufrechou wrote: > Currently the design of the widget is limited by the fact that the ipython > shell is not in a different process or thread, so if it locks, it locks all > the app? So, have you tried launching IPython in its own thread (separate from apps UI thread) yet? -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From Laurent.Dufrechou at free.fr Thu Dec 6 08:13:22 2007 From: Laurent.Dufrechou at free.fr (Laurent Dufrechou) Date: Thu, 6 Dec 2007 05:13:22 -0800 (PST) Subject: [IPython-dev] Need help from Ipython guru... In-Reply-To: <46cb515a0712060433g107267ddp5cad42ce5ea4aa29@mail.gmail.com> References: <14171949.post@talk.nabble.com> <46cb515a0712060433g107267ddp5cad42ce5ea4aa29@mail.gmail.com> Message-ID: <14191929.post@talk.nabble.com> Not yet, I'm playing with the shell currently, to see how it react and how it interact with Ipytohn instance. I'll try it after. I prefer to stabilize a little the whole thing and get some sort of working thing before blowing it with the traeds :) Thanks to ferando, i've solved the history bug. unicode character seem to be solved too (when i wrote '?' i got a problem with ipython) Here is my current bug list : ### known bug: # - currently WxConsoleView is a wx.Panel, can we change it to wx.TextCtrl? # - autoindent doesn't work # - I must find a way to handle crtl+c with a os.fork() in case the ipython do an infinite loop # while (1): # a=1 # - 'pass' key when call make the interpreter to stop current multiline command # class a: # def b(): # pass # def c(): # --> syntax error '):' # - some color glitch if userr click in a prompt color zone and start to write text. # - 'ls' command don't work, all the result is sent on external console... # - '?' can lock up the system if the default pager is used: # import smtplib # smtplib.re? # Need to find a way to call my own pager. Perhaps ville you know how os.environ declaration are used by IPython? I don't understand this line: os.environ['TERM'] = 'dumb' Fernando says me to do a os.environ['PAGER'] but i don't understand how that interact wiht ipython shell What do i need to write to make Ipython instance call mypager let say : self.pager(...)? I've tied os.environ['PAGER'] = 'self.pager' but it seems to try to make some sort of os.popen('self.pager') :( Too bad because under windows i prefer not to use command line shell.(it is really bad :) ) Perhaps if I do the ugly thing IPython.genutils.dumb_pager = self.pager :) ? hum... prefer to avoid it... but... Any idea welcome :) Once this bug resolved I'll send you my curent version. Laurent Ville M. Vainio wrote: > > On Dec 5, 2007 3:51 PM, Laurent Dufrechou > wrote: > >> Currently the design of the widget is limited by the fact that the >> ipython >> shell is not in a different process or thread, so if it locks, it locks >> all >> the app? > > So, have you tried launching IPython in its own thread (separate from > apps UI thread) yet? > > > -- > Ville M. Vainio - vivainio.googlepages.com > blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://lists.ipython.scipy.org/mailman/listinfo/ipython-dev > > -- View this message in context: http://www.nabble.com/Need-help-from-Ipython-guru...-tf4949781.html#a14191929 Sent from the IPython - Development mailing list archive at Nabble.com. From laurent.dufrechou at free.fr Thu Dec 6 18:05:31 2007 From: laurent.dufrechou at free.fr (Laurent Dufrechou) Date: Fri, 7 Dec 2007 00:05:31 +0100 Subject: [IPython-dev] wxIPython In-Reply-To: <20071205153301.GB8314@mentat.za.net> References: <20071202230454.GA3072@clipper.ens.fr> <009301c8353f$0c58d820$250a8860$@dufrechou@free.fr> <20071205153301.GB8314@mentat.za.net> Message-ID: <002601c8385c$80a4f4a0$81eedde0$@dufrechou@free.fr> Hi guys, Here is my last version. I've removed the bug with '?' and probably also autocompletion. Feel free to report any other bugs. It doesn't support all the functions of true ipython shell but it's I think it's a good start. I need now to do some cleaning, comments etc... Current boring unsolved issue: 'Ls' command result is written in console not into the widget ?_? If you really need ls just do... __IP.system("dir") Or whatever your os need :) How and crlc+c isn't supported... so don't write an infinite loop ! Any comment, other bug report welcome. Cheers, Laurent ### known bug: # - currently WxConsoleView is a wx.Panel, can we change it to wx.TextCtrl? # - autoindent doesn't work # - I must find a way to handle crtl+c with a os.fork() in case the ipython do an infinite loop # while (1): # a=1 # - 'pass' key when call make the interpreter to stop current multiline command # class a: # def b(): # pass # def c(): # --> syntax error '):' # - some color glitch if userr click in a prompt color zone and start to write text. # - 'ls' command don't work (and probably all external commands) -----Message d'origine----- De?: Stefan van der Walt [mailto:stefan at sun.ac.za] Envoy??: mercredi 5 d?cembre 2007 16:33 ??: Laurent Dufrechou Objet?: Re: [IPython-dev] wxIPython Hi Laurent I see TAB completion still freezes everything up. Maybe you can take a look at that next? Cheers St?fan On Mon, Dec 03, 2007 at 12:57:08AM +0100, Laurent Dufrechou wrote: > I have just updated the code to support multiline history recall. (pushing > up + enter will rerun a multiline call. > With that I'm quite happy. > I've found that it doesn't support ctrl+c, so I need to check this...I f > anybody want to add it :) > > >Hi Laurent, > > > >Congratulations for that. > > > > > >We have currently started an effort to get a wxPython front end for > >Ipython1, the next version of Ipython. It lives in the Ipython1 svn: > >http://projects.scipy.org/ipython/ipython/browser/ipython1/trunk/ipython1/f > ronte > >nd > >The idea is to have a readline replacement, ipreadline, that can act as > >readline, but have several front ends. We have currently started the > >wxPython front-end, but to make all this general-enough, it would be a > >good idea to start other front ends. > > > >As none of us are good wxPython coder (unlike you, as it seems), and as > >we don't have much time, it is moving along very slowly. But we sure > >could use some help. > > I've worked 6 month ago with ipython1, an excellent piece of code :) (for > code execution on multiple machine) > I'm happy to learn that you are trying to do a gui for it. > Will take a look at your svn when possible :) > > Bye, > Laurent > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://lists.ipython.scipy.org/mailman/listinfo/ipython-dev -------------- next part -------------- A non-text attachment was scrubbed... Name: wxIpython_without_old_shell.zip Type: application/x-zip-compressed Size: 7583 bytes Desc: not available URL: From laurent.dufrechou at free.fr Fri Dec 7 20:46:47 2007 From: laurent.dufrechou at free.fr (Laurent Dufrechou) Date: Sat, 8 Dec 2007 02:46:47 +0100 Subject: [IPython-dev] new wxIPython version 0.4.(polished,polished!) Message-ID: <008501c8393c$3336ceb0$99a46c10$@dufrechou@free.fr> Hello, Here is the 0.4 version. I think this will be my last release until some times. I've done a lot of polish: - Multiline history cleaning - Color corruption if click on any area solved - Added a helper widget that track the history, very usefull (well for me :) ) It evolves every time you execute a line in ipython shell. You can edit/copy/paste/delete into it so you can use it as a buffer to design on the fly a script. At the end, just copy the buffer in your favorite text editor and that's all. The only bugs remaining will be solved later, I need help from IPython devs on these I think. ### known bug: # - autoindent doesn't work # - I must find a way to handle crtl+c in case the ipython do an infinite loop # while (1): # a=1 # --> use IPython in a separate thread, but then how can I send it a Keyborard Interrupt? # - 'pass' key when call make the interpreter to stop current multiline command # class a: # def b(): # pass # def c(): # --> syntax error '):' # - '!' commands outputs result on a specific stream that I can't manage to catch! I think I'll also take a look at gael work now, as I'm quite happy with the current state of the app. And ho yes, I need to understand what is pyreadline used for, because I think some of its functions could have been used in my project... (there some magic lines like self.IP.readline_startup_hook that are in the project but don't know how does it works...) Is there a list of all IPython possible hooks and their usage? Regards, Laurent -------------- next part -------------- A non-text attachment was scrubbed... Name: wxIPython_0.4.zip Type: application/x-zip-compressed Size: 8477 bytes Desc: not available URL: From jorgen.stenarson at bostream.nu Sun Dec 9 12:22:54 2007 From: jorgen.stenarson at bostream.nu (=?ISO-8859-1?Q?J=F6rgen_Stenarson?=) Date: Sun, 09 Dec 2007 18:22:54 +0100 Subject: [IPython-dev] new wxIPython version 0.4.(polished,polished!) In-Reply-To: <008501c8393c$3336ceb0$99a46c10$@dufrechou@free.fr> References: <008501c8393c$3336ceb0$99a46c10$@dufrechou@free.fr> Message-ID: <475C246E.4090706@bostream.nu> Laurent Dufrechou skrev: > > I think I'll also take a look at gael work now, as I'm quite happy with the > current state of the app. > And ho yes, I need to understand what is pyreadline used for, because I > think some of its functions could have been used in my project... (there > some magic lines like self.IP.readline_startup_hook that are in the project > but don't know how does it works...) > Laurent, the primary purpose of pyreadline is to provide windows users with an implementation of the functionality of GNU readline which is availbale on most unix like platforms. Readline provides an improved experience when working at the interactive prompt by supporting history, command line editing, copy paste, search through the history buffer etc. Pyreadline has a console backend that can be replaced for different kinds of consoles. Currently there one for the standard windows console and an experimental one for ironpython. I'm planning to include a patch to support a callback interface which I think will be useful for inclusion of pyreadline in a gui event handler. I'm going to try to do this during the coming week. Hope this description helps. /J?rgen From laurent.dufrechou at free.fr Mon Dec 10 05:39:28 2007 From: laurent.dufrechou at free.fr (laurent.dufrechou at free.fr) Date: Mon, 10 Dec 2007 11:39:28 +0100 Subject: [IPython-dev] new wxIPython version 0.4.(polished,polished!) Message-ID: <1197283168.475d1760c08de@imp.free.fr> Hi J?rgen,All, I've updated the app this week-end and it now supports a threaded IPython :) So Ctrl+C works now, also the '!' commands works too. It was a little bit tricky because it is not easy to raise an asynchronous KeyboardInterrupt inside another thread. Hopefully I've found on the net a little piece of code that do it :) Things remaining are autoindent that still doesn't want to work, tab completion behavior that don't work on some case ( a=smtplib.+TAB don't work) and the strange bug with: Class a: Def a(): Pass The next line always fail with syntax error... (classical IPython shell work) I will release the new version soon, I want before to check some glitch on the new pager. >From my point of view, the things that were missing where first not so pyreadline related in fact. Because I used the function provided by IPython, I could find the history. I admit it would have been more easier if there was inside IPython something like historyBack, historyForward, getHistoryLength, setHistoryPos(pos). The hardest thing was to find what function needed to be redefined: IPython.genutils.page_dumb, IPython.iplib.raw_input, IPython.Shell.Term.cout /cin/cerr, IPython.genutils.shell_ori and self._IP.stdin_encoding, self._IP.exit. What my gui needed was (If I don't forget anything): getPromptCount (to get the number of prompt) getPrompt (to get the prompt) getPromptLengthRaw (with no colors text)<-- can be handled by pyreadline complete (to get completion list) <-- can be handled by pyreadline to format it knowing the width of the widget removeCurrentLine() removeCurrentPrompt() that remove multiple line until start of curren prompt write() that handle ansi colors <-- can be handled by pyreadline, but GUI toolkit related... getCurrentLineEnd() goes to end of buffer and finally a protection system for the colored text. <-- hum... don't know if could be pyreadline related, sure there could be a mechanism were pyreadline has property buffer that I could ask to check the color of a position and it it is protected or not :) (could be REALLY usefull) (Currently to do protection I check if the color is white on black :) ) I still miss autoindent, but I try to use it via IPython instance, so I think pyreadline will not be necessary on my side but more on IPython side.(that is mostly hidden for me) The problem is that most of these function are widget related and tightly related to the GUI library you use. So I don't know if it is more a pyreadline problem or a GUI designer problem... I think, (well on how I've used IPython instance and that can be done differently) that pyreadline was hiddent inside IPython. Pyreadline can be very usefull if I had reimplemented the IPython loop but as I've reused an IPython object and communicate with it via cin/cout/cerr all the pyreadline magic was hidden inside IPython. I was only reusing the pyreadline reformated cout in fact. For the copy/paste it was directly handled by the widget I derivate(?) so didn't need pyreadline there too.(but can be a good idea for other widget/toolkit). For the auto-indent, here I will really need pyreadline, because I send a command to IPython and if it returns a prompt it could (I think) be very usefull if I could ask pyreadline how much tab I need to had to it. And this is a function that is still missing for me (perhaps it exists, and I 've not found it for the moment). So to conclude, --> Property buffer to keep track of a property of a position (hummm, perhaps hard to do...) --> getIndentation function (really need it) --> getFormatedCompletion(size) Could be very useful for me! I will take a look at the console backend... Ok I've just done it :) Humm seems that in fact it looks like what I've done in my WxConsoleView class! Perhaps we could adapt it to integrate well with pyreadline??? (And make then a Wx_Console?) Laurent From vivainio at gmail.com Mon Dec 10 06:49:00 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Mon, 10 Dec 2007 13:49:00 +0200 Subject: [IPython-dev] Need help from Ipython guru... In-Reply-To: <14191929.post@talk.nabble.com> References: <14171949.post@talk.nabble.com> <46cb515a0712060433g107267ddp5cad42ce5ea4aa29@mail.gmail.com> <14191929.post@talk.nabble.com> Message-ID: <46cb515a0712100349m1beb1c5dhed19b530530e2fbf@mail.gmail.com> On Dec 6, 2007 3:13 PM, Laurent Dufrechou wrote: > Perhaps ville you know how os.environ declaration are used by IPython? > > I don't understand this line: > os.environ['TERM'] = 'dumb' > > Fernando says me to do a os.environ['PAGER'] but i don't understand how > that interact wiht ipython shell It just looks up the environment variable PAGER. To hook up paging to a function, we would need a hook but that is not yet available. Perhaps adding such a thing (all the hooks are in hooks.py) to IPython is in order? -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From fperez.net at gmail.com Tue Dec 11 01:55:24 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Mon, 10 Dec 2007 23:55:24 -0700 Subject: [IPython-dev] [Ipython-svndiff] 2885 - Do not replace # w/ \# anymore In-Reply-To: <20071207105914.5731039C22D@new.scipy.org> References: <20071207105914.5731039C22D@new.scipy.org> Message-ID: On Dec 7, 2007 3:59 AM, wrote: > Author: vivainio > Date: 2007-12-07 04:59:06 -0600 (Fri, 07 Dec 2007) > New Revision: 2885 > > Modified: > ipython/trunk/IPython/Extensions/ipy_profile_sh.py > ipython/trunk/IPython/iplib.py > Log: > Do not replace # w/ \# anymore > > Modified: ipython/trunk/IPython/Extensions/ipy_profile_sh.py > =================================================================== > --- ipython/trunk/IPython/Extensions/ipy_profile_sh.py 2007-12-02 04:42:21 UTC (rev 2884) > +++ ipython/trunk/IPython/Extensions/ipy_profile_sh.py 2007-12-07 10:59:06 UTC (rev 2885) > @@ -134,7 +134,7 @@ > # utility to expand user variables via Itpl > # xxx do something sensible with depth? > ip.IP.var_expand = lambda cmd, lvars=None, depth=2: \ > - str(ItplNS(cmd.replace('#','\#'), ip.IP.user_ns, get_locals())) > + str(ItplNS(cmd, ip.IP.user_ns, get_locals())) Mmh, when replacing code like this, which was obviously put in there explicitly and for some reason, please double check with others first. While I realize that the lack of any automatic test makes this harder, code like what was there didn't end up that way by accident, so there must have been a reason for it. In this case, the behavior that was in there was to ensure compatibility with the shell: In [1]: !echo hi # cmd hi # cmd Which is how the shell behaves: maqroll[~]> echo hi # cmd hi # cmd Now, after the above commit: In [1]: !echo hi # cmd hi I won't revert (yet) the commit because I imagine there was a reason for it, but the right approach is to find a solution for both needs, not to blindly break what was in there. Cheers, f From vivainio at gmail.com Tue Dec 11 10:04:52 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Tue, 11 Dec 2007 17:04:52 +0200 Subject: [IPython-dev] [Ipython-svndiff] 2885 - Do not replace # w/ \# anymore In-Reply-To: References: <20071207105914.5731039C22D@new.scipy.org> Message-ID: <46cb515a0712110704m3915ccd7i19eaf8b921c923a1@mail.gmail.com> On Dec 11, 2007 8:55 AM, Fernando Perez wrote: > Mmh, when replacing code like this, which was obviously put in there > explicitly and for some reason, please double check with others first. I asked about this a while ago, but you couldn't figure out back then why it was there. I figured that this early in the svn cycle, we could smoke out the real reason. And that appears to be the case :-). > In this case, the behavior that was in there was to ensure > compatibility with the shell: > > In [1]: !echo hi # cmd > hi # cmd > > Which is how the shell behaves: > > maqroll[~]> echo hi # cmd > hi # cmd > > Now, after the above commit: > > In [1]: !echo hi # cmd > hi > > I won't revert (yet) the commit because I imagine there was a reason > for it, but the right approach is to find a solution for both needs, > not to blindly break what was in there. Now that we know the root of the problem, we can make escaping # a linux-specific thing. On win32, escaping # broke all the command lines that have #. -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From vivainio at gmail.com Tue Dec 11 10:22:53 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Tue, 11 Dec 2007 17:22:53 +0200 Subject: [IPython-dev] [Ipython-svndiff] 2885 - Do not replace # w/ \# anymore In-Reply-To: <46cb515a0712110704m3915ccd7i19eaf8b921c923a1@mail.gmail.com> References: <20071207105914.5731039C22D@new.scipy.org> <46cb515a0712110704m3915ccd7i19eaf8b921c923a1@mail.gmail.com> Message-ID: <46cb515a0712110722r3ca1fd57icd9c4c6cadc241f7@mail.gmail.com> On Dec 11, 2007 5:04 PM, Ville M. Vainio wrote: > Now that we know the root of the problem, we can make escaping # a > linux-specific thing. On win32, escaping # broke all the command lines > that have #. Speaking of which, why does linux act this way? It thinks # is a comment character? If that is the case, should we not pass it to the shell as-is and require the user to escape it manually? Escaping something is quite doable, but unescaping \# is of course impossible... -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From stefan at sun.ac.za Tue Dec 11 13:39:44 2007 From: stefan at sun.ac.za (Stefan van der Walt) Date: Tue, 11 Dec 2007 20:39:44 +0200 Subject: [IPython-dev] wxIPython In-Reply-To: <002601c8385c$80a4f4a0$81eedde0$@dufrechou@free.fr> References: <20071202230454.GA3072@clipper.ens.fr> <20071205153301.GB8314@mentat.za.net> <002601c8385c$80a4f4a0$81eedde0$@dufrechou@free.fr> Message-ID: <20071211183943.GD11744@mentat.za.net> Hi Laurent On Fri, Dec 07, 2007 at 12:05:31AM +0100, Laurent Dufrechou wrote: > Hi guys, > > Here is my last version. > I've removed the bug with '?' and probably also autocompletion. > Feel free to report any other bugs. > It doesn't support all the functions of true ipython shell but it's I think > it's a good start. > I need now to do some cleaning, comments etc... > > Current boring unsolved issue: > 'Ls' command result is written in console not into the widget ?_? > If you really need ls just do... __IP.system("dir") > Or whatever your os need :) > How and crlc+c isn't supported... so don't write an infinite loop ! > > Any comment, other bug report welcome. I use a pretty standard Ubuntu (gutsy) install with wx 2.8. Whenever I press or the program freezes up. Sorry, I haven't had time to look into it more carefully. Regards St?fan From hans_meine at gmx.net Tue Dec 11 04:10:47 2007 From: hans_meine at gmx.net (Hans Meine) Date: Tue, 11 Dec 2007 10:10:47 +0100 Subject: [IPython-dev] [Ipython-svndiff] 2885 - Do not replace # w/ \# anymore In-Reply-To: References: <20071207105914.5731039C22D@new.scipy.org> Message-ID: <200712111010.47780.meine@informatik.uni-hamburg.de> Am Dienstag, 11. Dezember 2007 07:55:24 schrieb Fernando Perez: > In this case, the behavior that was in there was to ensure > compatibility with the shell: > > In [1]: !echo hi # cmd > hi # cmd > > Which is how the shell behaves: > > maqroll[~]> echo hi # cmd > hi # cmd > > Now, after the above commit: > > In [1]: !echo hi # cmd > hi In bash, there is an "interactive comments" flag IIRC (settable with 'set'); for me, the latter is the desired behaviour. (Eases copy-pasting code from scripts.) -- Ciao, / / /--/ / / ANS From laurent.dufrechou at free.fr Tue Dec 11 18:46:43 2007 From: laurent.dufrechou at free.fr (Laurent Dufrechou) Date: Wed, 12 Dec 2007 00:46:43 +0100 Subject: [IPython-dev] wxIPython 0.6 (Multithreaded and current bug list flushed) Message-ID: <000001c83c50$1633e890$429bb9b0$@dufrechou@free.fr> Hi there, Here is the last version of wxIPython. - Multithreaded: The IPython shell is instanciated inside a thread and the wxapp communicate with thread and send line to execute to it. The class can be reused as is with any other GUI toolkit :). So if there is any pyQT developpers I will be happy to help them! - Autoindent feature - new WxPager (try smtplib.re?) - '!' commands support If any linux guy could test it, (mine is currently dead :() and report any bugs. I'm concerned about locale support. Currently it works well in French... By the way, there is some hacky things inside it especially it could be really usefull if IPython could provide these hooks:(because currently I completely redefines IPython libs func(!) ) IPython.Shell.Term.cin IPython.Shell.Term.cout IPython.Shell.Term.cerr IPython.iplib.raw_input = lambda x: None IPython.genutils.IOTerm(cin=cin, cout=cout, cerr=cerr) #we replace the ipython default pager by our pager IPython.OInspect.page #we replace the ipython default shell command caller by our shell handler IPython.genutils.shell_ori #we replace the ipython default input command caller by our method IPython.iplib.raw_input_original Any comments, bug reports welcomed! Do you think (if you're ok and bugs are quite few :) ) that we could add it to IPython Gui integration example? Laurent -------------- next part -------------- A non-text attachment was scrubbed... Name: wxIPython0.6.zip Type: application/x-zip-compressed Size: 10852 bytes Desc: not available URL: From fperez.net at gmail.com Wed Dec 12 00:07:28 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Tue, 11 Dec 2007 22:07:28 -0700 Subject: [IPython-dev] Moin wiki config closed out... Message-ID: Hi all, I recently realized that despite trying to disable new user accounts, I'd never correctly set the Moin access control lists, and hence the whole site was actually open to spam. I basically spent a bunch of time locking up an attic window and left the entry door wide open... Anyway, Ryan James had been silently doing the thankless task of cleaning up the constant flow of spam, for which I'm very grateful. But I'm sure your time is better spent elsewhere (with ipython, of course :), Ryan, so I just closed up the main ACL for anonymous edits. If any of you with a normal login to the wiki can't edit, please let me know ASAP so I can fix it. And sorry for such a stupid blunder... Cheers, f From vivainio at gmail.com Wed Dec 12 00:36:30 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Wed, 12 Dec 2007 07:36:30 +0200 Subject: [IPython-dev] wxIPython 0.6 (Multithreaded and current bug list flushed) In-Reply-To: <-8440009743494299951@unknownmsgid> References: <-8440009743494299951@unknownmsgid> Message-ID: <46cb515a0712112136t4066402fve5cd53abc4307de6@mail.gmail.com> On Dec 12, 2007 1:46 AM, Laurent Dufrechou wrote: > Any comments, bug reports welcomed! > Do you think (if you're ok and bugs are quite few :) ) that we could add it > to IPython Gui integration example? Yes, I was thinking of adding it to "doc/examples" folder, though something like "contrib" folder would perhaps be more appropriate... -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From vivainio at gmail.com Wed Dec 12 00:38:28 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Wed, 12 Dec 2007 07:38:28 +0200 Subject: [IPython-dev] [Ipython-svndiff] 2885 - Do not replace # w/ \# anymore In-Reply-To: <200712111010.47780.meine@informatik.uni-hamburg.de> References: <20071207105914.5731039C22D@new.scipy.org> <200712111010.47780.meine@informatik.uni-hamburg.de> Message-ID: <46cb515a0712112138m6b2250d3wc74f51b9b2384b6@mail.gmail.com> On Dec 11, 2007 11:10 AM, Hans Meine wrote: > In bash, there is an "interactive comments" flag IIRC (settable with 'set'); > for me, the latter is the desired behaviour. (Eases copy-pasting code from > scripts.) Ok, this suggests that the current behaviour (passing # to os.system unescaped) is the correct one also for Linux. And after all, escaping is easy... -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From fperez.net at gmail.com Wed Dec 12 00:46:34 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Tue, 11 Dec 2007 22:46:34 -0700 Subject: [IPython-dev] [Ipython-svndiff] 2885 - Do not replace # w/ \# anymore In-Reply-To: <46cb515a0712112138m6b2250d3wc74f51b9b2384b6@mail.gmail.com> References: <20071207105914.5731039C22D@new.scipy.org> <200712111010.47780.meine@informatik.uni-hamburg.de> <46cb515a0712112138m6b2250d3wc74f51b9b2384b6@mail.gmail.com> Message-ID: On Dec 11, 2007 10:38 PM, Ville M. Vainio wrote: > On Dec 11, 2007 11:10 AM, Hans Meine wrote: > > > In bash, there is an "interactive comments" flag IIRC (settable with 'set'); > > for me, the latter is the desired behaviour. (Eases copy-pasting code from > > scripts.) > > Ok, this suggests that the current behaviour (passing # to os.system > unescaped) is the correct one also for Linux. And after all, escaping > is easy... I don't know if tcsh has this optional flag, but I can live with the behavior. We'll see if we get any complaints... Let's just comment the fact if possible, so we don't end up repeating the cycle in the future :) Cheers, f From hans_meine at gmx.net Wed Dec 12 03:25:00 2007 From: hans_meine at gmx.net (Hans Meine) Date: Wed, 12 Dec 2007 09:25:00 +0100 Subject: [IPython-dev] [Ipython-svndiff] 2885 - Do not replace # w/ \# anymore In-Reply-To: References: <20071207105914.5731039C22D@new.scipy.org> <46cb515a0712112138m6b2250d3wc74f51b9b2384b6@mail.gmail.com> Message-ID: <200712120925.00688.hans_meine@gmx.net> > > On Dec 11, 2007 11:10 AM, Hans Meine wrote: > > > In bash, there is an "interactive comments" flag IIRC (settable with > > > 'set'); for me, the latter is the desired behaviour. (Eases > > > copy-pasting code from scripts.) I just saw that it is even on by default; quoting bash(1): > In a non-interactive shell, or an interactive shell in which the > interactive_comments option to the shopt builtin is enabled (see SHELL > BUILTIN COMMANDS below), a word beginning with # causes that word and all > remaining characters on that line to be ignored. An interactive shell > without the interactive_comments option enabled does not allow comments. > The interactive_comments option is on by default in interactive shells. Mittwoch, 12. Dezember 2007 06:46:34 Fernando wrote: > On Dec 11, 2007 10:38 PM, Ville M. Vainio wrote: > > Ok, this suggests that the current behaviour (passing # to os.system > > unescaped) is the correct one also for Linux. And after all, escaping > > is easy... > > I don't know if tcsh has this optional flag, but I can live with the > behavior. We'll see if we get any complaints... My tcsh(1) says only: > When the shell's input is not a terminal, the character `#' is taken to > begin a comment. Each `#' and the rest of the input line on which it > appears is discarded before further parsing. No sign of interactive comments. zsh has them though. Just for the records. I am +1 on Ville's change and an according comment. Greetings, Hans From fperez.net at gmail.com Wed Dec 12 03:30:16 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Wed, 12 Dec 2007 01:30:16 -0700 Subject: [IPython-dev] [Ipython-svndiff] 2885 - Do not replace # w/ \# anymore In-Reply-To: <200712120925.00688.hans_meine@gmx.net> References: <20071207105914.5731039C22D@new.scipy.org> <46cb515a0712112138m6b2250d3wc74f51b9b2384b6@mail.gmail.com> <200712120925.00688.hans_meine@gmx.net> Message-ID: On Dec 12, 2007 1:25 AM, Hans Meine wrote: > My tcsh(1) says only: > > When the shell's input is not a terminal, the character `#' is taken to > > begin a comment. Each `#' and the rest of the input line on which it > > appears is discarded before further parsing. > > No sign of interactive comments. zsh has them though. > > Just for the records. I am +1 on Ville's change and an according comment. Yup, will do that. Thanks for the detailed info on the shells. Cheers, f From prabhu at aero.iitb.ac.in Wed Dec 12 06:47:40 2007 From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran) Date: Wed, 12 Dec 2007 17:17:40 +0530 Subject: [IPython-dev] wxIPython 0.6 (Multithreaded and current bug list flushed) In-Reply-To: <000001c83c50$1633e890$429bb9b0$@dufrechou@free.fr> References: <000001c83c50$1633e890$429bb9b0$@dufrechou@free.fr> Message-ID: <475FCA5C.7020101@aero.iitb.ac.in> Laurent Dufrechou wrote: > If any linux guy could test it, (mine is currently dead :() and report any > bugs. > I'm concerned about locale support. Currently it works well in French... This looks very cool and I'd love to be able to use this in various other tools. I am running into some problems. I'm using Ubuntu gutsy with wx.__version__ == '2.8.4.0' and the default ipython-0.8.1-1 shipped with Ubuntu. When I run wxIPython.py, I run into the following. 1. I can't see what I type, altho you have the FG set to WHITE it renders it black. So unless I change the BG (in ipython_view.py) to something else that is light I can't see the cursor or any test I type. 2. Regardless of 1 if I do type anything I get the following In [1]: print "hello" --------------------------------------------------------------------------- Traceback (most recent call last) /var/lib/python-support/python2.5/IPython/iplib.py in raw_input(self, prompt, continue_prompt) 2013 2014 try: -> 2015 line = raw_input_original(prompt).decode(self.stdin_encoding) 2016 except ValueError: 2017 warn("\n********\nYou or a %run:ed script called sys.stdin.close()" : decode() argument 1 must be string, not None In [2]: Sorry I can't debug anymore, I just thought I'd give you what errors I was running into. regards, prabhu From laurent.dufrechou at gmail.com Wed Dec 12 07:37:50 2007 From: laurent.dufrechou at gmail.com (=?us-ascii?Q?Laurent_Dufrechou?=) Date: Wed, 12 Dec 2007 13:37:50 +0100 Subject: [IPython-dev] wxIPython 0.6 (Multithreaded and current bug list flushed) In-Reply-To: <475FC9F1.6040304@aero.iitb.ac.in> References: <000001c83c50$1633e890$429bb9b0$@dufrechou@free.fr> <475FC9F1.6040304@aero.iitb.ac.in> Message-ID: <475fd630.0637560a.2d9b.2796@mx.google.com> Hi Prabhu, >When I run wxIPython.py, I run into the following. > > 1. I can't see what I type, altho you have the FG set to WHITE it >renders it black. So unless I change the BG (in ipython_view.py) to >something else that is light I can't see the cursor or any test I type. Hum seems to be a GTK problem with wx. It is probably handled differently. I'll launch my debian machine to check this. So foreground is not well set... Can you (while in black) do an error like "a"+enter at startup and tells me if traceback appears in color? > 2. Regardless of 1 if I do type anything I get the following >In [1]: print "hello" >--------------------------------------------------------------------------- > Traceback (most recent call last) > >/var/lib/python-support/python2.5/IPython/iplib.py in raw_input(self, >prompt, continue_prompt) > 2013 > 2014 try: >-> 2015 line = >raw_input_original(prompt).decode(self.stdin_encoding) > 2016 except ValueError: > 2017 warn("\n********\nYou or a %run:ed script called >sys.stdin.close()" > >: decode() argument 1 must be string, not None >In [2]: Ouch, it's a locale problem. In fact in some point of the code I try to detect the local used on the system, and in your case, it has not been detected, and set to None. Locale is set line 108: #we replace IPython default encoding by wx locale encoding self._IP.stdin_encoding = locale.getdefaultlocale()[1] Anyway, if you want to continue to try the app, just comment this line. Can you please send me the result of: Import locale print locale.getdefaultlocale()? regards, Laurent From prabhu at aero.iitb.ac.in Wed Dec 12 07:50:01 2007 From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran) Date: Wed, 12 Dec 2007 18:20:01 +0530 Subject: [IPython-dev] wxIPython 0.6 (Multithreaded and current bug list flushed) In-Reply-To: <475fd630.0637560a.2d9b.2796@mx.google.com> References: <000001c83c50$1633e890$429bb9b0$@dufrechou@free.fr> <475FC9F1.6040304@aero.iitb.ac.in> <475fd630.0637560a.2d9b.2796@mx.google.com> Message-ID: <475FD8F9.7070306@aero.iitb.ac.in> Laurent Dufrechou wrote: > Hi Prabhu, > >> When I run wxIPython.py, I run into the following. >> >> 1. I can't see what I type, altho you have the FG set to WHITE it >> renders it black. So unless I change the BG (in ipython_view.py) to >> something else that is light I can't see the cursor or any test I type. > > Hum seems to be a GTK problem with wx. It is probably handled differently. > I'll launch my debian machine to check this. > So foreground is not well set... > Can you (while in black) do an error like "a"+enter at startup and tells me > if traceback appears in color? Yes, the traceback is in color and seems to render OK. If I change the bg to white and fg to black things are better. BTW, I can see output text just fine, just the input line is broken. [...] > Ouch, it's a locale problem. In fact in some point of the code I try to > detect the local used on the system, and in your case, it has not been > detected, and set to None. > Locale is set line 108: > #we replace IPython default encoding by wx locale encoding > self._IP.stdin_encoding = locale.getdefaultlocale()[1] > > Anyway, if you want to continue to try the app, just comment this line. Ahh, cool, this fixes it and I can use Python. However, I just did this: import string string?? and get a core dump! Otherwise if I do this: ? I get the help on the terminal and not on the wx shell. If I do !ls for example the output is listed on the shell I launched the wxIPython shell from. If I hit tab, I get completion (on the wx widget) but I also loose keyboard focus. These sound like small linux specific issues. > > Can you please send me the result of: > Import locale > print locale.getdefaultlocale()? In [1]: import locale In [2]: print locale.getdefaultlocale() (None, None) HTH. cheers, prabhu From laurent.dufrechou at gmail.com Wed Dec 12 08:25:18 2007 From: laurent.dufrechou at gmail.com (=?us-ascii?Q?Laurent_Dufrechou?=) Date: Wed, 12 Dec 2007 14:25:18 +0100 Subject: [IPython-dev] wxIPython 0.6 (Multithreaded and current bug list flushed) In-Reply-To: <475FD8F9.7070306@aero.iitb.ac.in> References: <000001c83c50$1633e890$429bb9b0$@dufrechou@free.fr> <475FC9F1.6040304@aero.iitb.ac.in> <475fd630.0637560a.2d9b.2796@mx.google.com> <475FD8F9.7070306@aero.iitb.ac.in> Message-ID: <475fe140.1358560a.1efb.0673@mx.google.com> >Yes, the traceback is in color and seems to render OK. > >If I change the bg to white and fg to black things are better. BTW, I >can see output text just fine, just the input line is broken. > >[...] Ok will check this tonight. As I haven't try under linux, must have a lot of bug ;) >Ahh, cool, this fixes it and I can use Python. However, I just did this: > >import string >string?? > >and get a core dump! Otherwise if I do this: > >? > >I get the help on the terminal and not on the wx shell. If I do !ls for >example the output is listed on the shell I launched the wxIPython shell >from. If I hit tab, I get completion (on the wx widget) but I also >loose keyboard focus. Oups :) I forget to try the '??' key, will check this too. For the'!' It's a boring problem I had to solve on windows (I mean redirect output to wx...), so I need to check this for linux... Hum, seems under linux there is big issue whith std redirection... arfff, I was thinking I had solved the issue...too bad :) Ok for the TAB. > > Can you please send me the result of: > Import locale > print locale.getdefaultlocale()? In [1]: import locale In [2]: print locale.getdefaultlocale() (None, None) Ok... mine outputs: ('fr_FR', 'cp1252'). >From what I've just checked,under linux it works differently... Good :) My bug list is now almost refilled thanks to you ;) I'll tell you if I need more info! Thanks! Laurent From gael.varoquaux at normalesup.org Wed Dec 12 09:11:31 2007 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Wed, 12 Dec 2007 15:11:31 +0100 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py Message-ID: <20071212141131.GE14948@clipper.ens.fr> Hi, I am not a big proponent of setuptools, but I must admit they have really nice feature for developpers: "python setup.py develop", which does the Python equivalent of a symlink during the install so that it uses the source code from the check out to run. I would like to add these lines to the setup.py: """ try: from setuptools import setup, find_packages except ImportError: pass """ I cannot see waht harm this can do, and I can see the benefit: I install an svn checkout of ipython with: """ sudo python setup.py develop --prefix /usr/local """ and then maintain this check out by doing "svn up", and this even in windows. What do you think? Ga?l From vivainio at gmail.com Wed Dec 12 12:00:47 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Wed, 12 Dec 2007 19:00:47 +0200 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <20071212141131.GE14948@clipper.ens.fr> References: <20071212141131.GE14948@clipper.ens.fr> Message-ID: <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> On Dec 12, 2007 4:11 PM, Gael Varoquaux wrote: > I would like to add these lines to the setup.py: > > """ > try: > from setuptools import setup, find_packages > except ImportError: > pass > """ > > I cannot see waht harm this can do, and I can see the benefit: I install > an svn checkout of ipython with: > > """ > sudo python setup.py develop --prefix /usr/local > """ > > and then maintain this check out by doing "svn up", and this even in > windows. > > What do you think? This has been discussed extensively in the past, with a halfway consensus that we should not do it. However, there is no harm in your snippet iff you check that 'develop' is in sys.argv. We should still try using normal distutils for normal situations. -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From strawman at astraw.com Wed Dec 12 12:25:45 2007 From: strawman at astraw.com (Andrew Straw) Date: Wed, 12 Dec 2007 09:25:45 -0800 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> Message-ID: <47601999.9090001@astraw.com> Ville M. Vainio wrote: > On Dec 12, 2007 4:11 PM, Gael Varoquaux wrote: > > >> I would like to add these lines to the setup.py: >> >> """ >> try: >> from setuptools import setup, find_packages >> except ImportError: >> pass >> """ >> >> I cannot see waht harm this can do, and I can see the benefit: I install >> an svn checkout of ipython with: >> >> """ >> sudo python setup.py develop --prefix /usr/local >> """ >> >> and then maintain this check out by doing "svn up", and this even in >> windows. >> >> What do you think? >> > > This has been discussed extensively in the past, with a halfway > consensus that we should not do it. > > However, there is no harm in your snippet iff you check that 'develop' > is in sys.argv. We should still try using normal distutils for normal > situations. > > And don't forget you can always do: python -c "import setuptools; execfile('setup.py')" develop --prefix /usr/local From gael.varoquaux at normalesup.org Wed Dec 12 12:33:36 2007 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Wed, 12 Dec 2007 18:33:36 +0100 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <47601999.9090001@astraw.com> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> Message-ID: <20071212173336.GA21815@clipper.ens.fr> On Wed, Dec 12, 2007 at 09:25:45AM -0800, Andrew Straw wrote: > And don't forget you can always do: > python -c "import setuptools; execfile('setup.py')" develop --prefix > /usr/local Yes, I'll settle for this. Though as the command is hard to remmeber for my small brain, I'll check in a file that does the "import setuptools; execfile('setup.py')" if you don't mind. What should I call it? setup_tools.py? Obviously not setuptools.py. Cheers, Ga?l From strawman at astraw.com Wed Dec 12 12:40:46 2007 From: strawman at astraw.com (Andrew Straw) Date: Wed, 12 Dec 2007 09:40:46 -0800 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <20071212173336.GA21815@clipper.ens.fr> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> Message-ID: <47601D1E.4020004@astraw.com> Here is my "ssetup", which is on my $PATH: #!/bin/bash # ssetup - "setuptools setup" python -c "import setuptools, sys;f='setup.py';sys.argv[0]=f;execfile(f,{'__file__':f,'__name__':'__main__'})" $* (Everything that is not a comment goes on the same line.) This deals with all kinds of setup.py scripts that expect __file__ and __name__ to be set in addition to sys.argv[0] to be 'setup.py'. Note that numpy.distutils and setuptools are not completely compatible, so this may get you into trouble with numpy/scipy, so be careful there. -Andrew Gael Varoquaux wrote: > On Wed, Dec 12, 2007 at 09:25:45AM -0800, Andrew Straw wrote: >> And don't forget you can always do: > >> python -c "import setuptools; execfile('setup.py')" develop --prefix >> /usr/local > > Yes, I'll settle for this. Though as the command is hard to remmeber for > my small brain, I'll check in a file that does the > "import setuptools; execfile('setup.py')" if you don't mind. What should > I call it? setup_tools.py? Obviously not setuptools.py. > > Cheers, > > Ga?l From vivainio at gmail.com Wed Dec 12 13:46:07 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Wed, 12 Dec 2007 20:46:07 +0200 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <20071212173336.GA21815@clipper.ens.fr> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> Message-ID: <46cb515a0712121046v2f45d4eflf9a6cf2d1ca419e8@mail.gmail.com> On Dec 12, 2007 7:33 PM, Gael Varoquaux wrote: > Yes, I'll settle for this. Though as the command is hard to remmeber for > my small brain, I'll check in a file that does the > "import setuptools; execfile('setup.py')" if you don't mind. What should Why not add the 'develop' check to setup.py, provided that 'develop' is the only thing you need it for? We could also throw in bdist_egg and get rid of a couple of files in the root of the source tree as a bonus... -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From ellisonbg.net at gmail.com Wed Dec 12 14:02:27 2007 From: ellisonbg.net at gmail.com (Brian Granger) Date: Wed, 12 Dec 2007 12:02:27 -0700 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <20071212173336.GA21815@clipper.ens.fr> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> Message-ID: <6ce0ac130712121102u2cd26086scf4a40f8858b0998@mail.gmail.com> We could also simply make an setupegg.py file that import setuptools and then does an execfile on setup.py. That is how matplotlib handles it as I recall. Brian On Dec 12, 2007 10:33 AM, Gael Varoquaux wrote: > On Wed, Dec 12, 2007 at 09:25:45AM -0800, Andrew Straw wrote: > > And don't forget you can always do: > > > python -c "import setuptools; execfile('setup.py')" develop --prefix > > /usr/local > > Yes, I'll settle for this. Though as the command is hard to remmeber for > my small brain, I'll check in a file that does the > "import setuptools; execfile('setup.py')" if you don't mind. What should > I call it? setup_tools.py? Obviously not setuptools.py. > > Cheers, > > Ga?l > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://lists.ipython.scipy.org/mailman/listinfo/ipython-dev > From gael.varoquaux at normalesup.org Wed Dec 12 14:11:41 2007 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Wed, 12 Dec 2007 20:11:41 +0100 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <6ce0ac130712121102u2cd26086scf4a40f8858b0998@mail.gmail.com> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> <6ce0ac130712121102u2cd26086scf4a40f8858b0998@mail.gmail.com> Message-ID: <20071212191141.GE21815@clipper.ens.fr> On Wed, Dec 12, 2007 at 12:02:27PM -0700, Brian Granger wrote: > We could also simply make an setupegg.py file that import setuptools > and then does an execfile on setup.py. That is how matplotlib handles > it as I recall. Yes, I think this is the good solution to please everybody. Should I copy the setupegg.py file? Cheers, Ga?l From vivainio at gmail.com Wed Dec 12 14:52:36 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Wed, 12 Dec 2007 21:52:36 +0200 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <6ce0ac130712121102u2cd26086scf4a40f8858b0998@mail.gmail.com> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> <6ce0ac130712121102u2cd26086scf4a40f8858b0998@mail.gmail.com> Message-ID: <46cb515a0712121152t51a25119rc70391715b16e6a0@mail.gmail.com> On Dec 12, 2007 9:02 PM, Brian Granger wrote: > We could also simply make an setupegg.py file that import setuptools > and then does an execfile on setup.py. That is how matplotlib handles > it as I recall. Hint: there is already eggsetup.py :-) -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From gael.varoquaux at normalesup.org Wed Dec 12 15:33:26 2007 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Wed, 12 Dec 2007 21:33:26 +0100 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <46cb515a0712121152t51a25119rc70391715b16e6a0@mail.gmail.com> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> <6ce0ac130712121102u2cd26086scf4a40f8858b0998@mail.gmail.com> <46cb515a0712121152t51a25119rc70391715b16e6a0@mail.gmail.com> Message-ID: <20071212203326.GA29192@clipper.ens.fr> On Wed, Dec 12, 2007 at 09:52:36PM +0200, Ville M. Vainio wrote: > On Dec 12, 2007 9:02 PM, Brian Granger wrote: > > We could also simply make an setupegg.py file that import setuptools > > and then does an execfile on setup.py. That is how matplotlib handles > > it as I recall. > Hint: there is already eggsetup.py :-) OK, today I am being stupid on two different mailing lists! Sorry for the noise. Ga?l From robert.kern at gmail.com Wed Dec 12 15:35:39 2007 From: robert.kern at gmail.com (Robert Kern) Date: Wed, 12 Dec 2007 14:35:39 -0600 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <47601D1E.4020004@astraw.com> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> <47601D1E.4020004@astraw.com> Message-ID: Andrew Straw wrote: > Note that numpy.distutils and setuptools are not completely compatible, > so this may get you into trouble with numpy/scipy, so be careful there. I fixed an issue today where develop didn't run build_src --inplace. What other issues are you aware of? -- Robert Kern "I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth." -- Umberto Eco From stefan at sun.ac.za Wed Dec 12 16:13:04 2007 From: stefan at sun.ac.za (Stefan van der Walt) Date: Wed, 12 Dec 2007 23:13:04 +0200 Subject: [IPython-dev] [Ipython-svndiff] 2885 - Do not replace # w/ \# anymore In-Reply-To: <46cb515a0712110704m3915ccd7i19eaf8b921c923a1@mail.gmail.com> References: <20071207105914.5731039C22D@new.scipy.org> <46cb515a0712110704m3915ccd7i19eaf8b921c923a1@mail.gmail.com> Message-ID: <20071212211304.GC14894@mentat.za.net> On Tue, Dec 11, 2007 at 05:04:52PM +0200, Ville M. Vainio wrote: > Now that we know the root of the problem, we can make escaping # a > linux-specific thing. On win32, escaping # broke all the command lines > that have #. I don't know about other Linux systems, but on Ubuntu the BASH shell behaves as IPython (SVN) does: $ echo hi # cmd hi Making platform specific changes are always risky, and in this case we can avoid it. Regards St?fan From fperez.net at gmail.com Wed Dec 12 16:20:45 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Wed, 12 Dec 2007 14:20:45 -0700 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <20071212191141.GE21815@clipper.ens.fr> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> <6ce0ac130712121102u2cd26086scf4a40f8858b0998@mail.gmail.com> <20071212191141.GE21815@clipper.ens.fr> Message-ID: On Dec 12, 2007 12:11 PM, Gael Varoquaux wrote: > On Wed, Dec 12, 2007 at 12:02:27PM -0700, Brian Granger wrote: > > We could also simply make an setupegg.py file that import setuptools > > and then does an execfile on setup.py. That is how matplotlib handles > > it as I recall. > > Yes, I think this is the good solution to please everybody. Should I copy > the setupegg.py file? Swamped right now, but go for it... f From gael.varoquaux at normalesup.org Wed Dec 12 16:54:04 2007 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Wed, 12 Dec 2007 22:54:04 +0100 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> <6ce0ac130712121102u2cd26086scf4a40f8858b0998@mail.gmail.com> <20071212191141.GE21815@clipper.ens.fr> Message-ID: <20071212215404.GD29192@clipper.ens.fr> On Wed, Dec 12, 2007 at 02:20:45PM -0700, Fernando Perez wrote: > > Yes, I think this is the good solution to please everybody. Should I copy > > the setupegg.py file? > Swamped right now, but go for it... As Ville pointed out, there is already eggsetup.py, which does this job, in the ipython svn. I was just being blind. Cheers, Ga?l From robert.kern at gmail.com Wed Dec 12 17:38:27 2007 From: robert.kern at gmail.com (Robert Kern) Date: Wed, 12 Dec 2007 16:38:27 -0600 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <46cb515a0712121152t51a25119rc70391715b16e6a0@mail.gmail.com> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> <6ce0ac130712121102u2cd26086scf4a40f8858b0998@mail.gmail.com> <46cb515a0712121152t51a25119rc70391715b16e6a0@mail.gmail.com> Message-ID: Ville M. Vainio wrote: > On Dec 12, 2007 9:02 PM, Brian Granger wrote: > >> We could also simply make an setupegg.py file that import setuptools >> and then does an execfile on setup.py. That is how matplotlib handles >> it as I recall. > > Hint: there is already eggsetup.py :-) AFAICT, that's a script solely for your or Fernando's benefit, not anyone else's. For it to be generally useful, it shouldn't add your local site-packages directory and it shouldn't delete anything. And why is there also setup_bdist_egg.py? Release scripts belong in tools/. -- Robert Kern "I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth." -- Umberto Eco From vivainio at gmail.com Wed Dec 12 18:10:25 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Thu, 13 Dec 2007 01:10:25 +0200 Subject: [IPython-dev] Experimental unicode branch Message-ID: <46cb515a0712121510g5bb4a72bsd78e1f59bf422038@mail.gmail.com> For making some unicode fixes, I created a new branch at: http://ipython.scipy.org/svn/ipython/ipython/branches/vivainio/ip_unicode_cleanup It only has my name to avoid bloating the "branches" namespace - everyone is welcome to commit some risky unicode-related enhancements that they don't dare to try on trunk. It ought to be safe to use this branch, especially if you don't typically use unicode (paradoxical as it is) and can afford unicode not working properly all the time. The aim is to use unicode internally as much as possible, and only coerce to str when necessary. -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From strawman at astraw.com Wed Dec 12 18:35:03 2007 From: strawman at astraw.com (Andrew Straw) Date: Wed, 12 Dec 2007 15:35:03 -0800 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> <47601D1E.4020004@astraw.com> Message-ID: <47607027.3020209@astraw.com> Robert Kern wrote: > Andrew Straw wrote: > > >> Note that numpy.distutils and setuptools are not completely compatible, >> so this may get you into trouble with numpy/scipy, so be careful there. >> > > I fixed an issue today where develop didn't run build_src --inplace. What other > issues are you aware of? > > http://scipy.org/scipy/numpy/ticket/596 From laurent.dufrechou at gmail.com Wed Dec 12 18:43:04 2007 From: laurent.dufrechou at gmail.com (=?iso-8859-1?Q?Laurent_Dufr=E9chou?=) Date: Thu, 13 Dec 2007 00:43:04 +0100 Subject: [IPython-dev] WxIpython Linux hell :) was RE: wxIPython 0.6 (Multithreaded and current bug list flushed) In-Reply-To: <475FD8F9.7070306@aero.iitb.ac.in> References: <000001c83c50$1633e890$429bb9b0$@dufrechou@free.fr> <475FC9F1.6040304@aero.iitb.ac.in> <475fd630.0637560a.2d9b.2796@mx.google.com> <475FD8F9.7070306@aero.iitb.ac.in> Message-ID: <4760720c.0d135e0a.6bca.6d1a@mx.google.com> Hi Prabhu, I've just installed all the things on my debian. First of all, ouch! What a difference! On linux it is unusable! As you said text is black (solved now), there is an issue with locale: Can you try locale.getpreferredencoding() and tells if you've got a result? (here on my debian the current command works.. :( ) There is a lot focus loosing problem, and backspace make loose the focus anf widget go on the top line!!! Argghh. And ho yes... stdin/out is redirected to console... Most of problem I has solved on windows have not been solved on linux... Will work on it. Excuse me to have delivered a so poor linux version... More news for next week ;) Cheers, Laurent -----Message d'origine----- De?: Prabhu Ramachandran [mailto:prabhu at aero.iitb.ac.in] Envoy??: mercredi 12 d?cembre 2007 13:50 ??: Laurent Dufrechou; IPython-dev at scipy.org Objet?: Re: [IPython-dev] wxIPython 0.6 (Multithreaded and current bug list flushed) Laurent Dufrechou wrote: > Hi Prabhu, > >> When I run wxIPython.py, I run into the following. >> >> 1. I can't see what I type, altho you have the FG set to WHITE it >> renders it black. So unless I change the BG (in ipython_view.py) to >> something else that is light I can't see the cursor or any test I type. > > Hum seems to be a GTK problem with wx. It is probably handled differently. > I'll launch my debian machine to check this. > So foreground is not well set... > Can you (while in black) do an error like "a"+enter at startup and tells me > if traceback appears in color? Yes, the traceback is in color and seems to render OK. If I change the bg to white and fg to black things are better. BTW, I can see output text just fine, just the input line is broken. [...] > Ouch, it's a locale problem. In fact in some point of the code I try to > detect the local used on the system, and in your case, it has not been > detected, and set to None. > Locale is set line 108: > #we replace IPython default encoding by wx locale encoding > self._IP.stdin_encoding = locale.getdefaultlocale()[1] > > Anyway, if you want to continue to try the app, just comment this line. Ahh, cool, this fixes it and I can use Python. However, I just did this: import string string?? and get a core dump! Otherwise if I do this: ? I get the help on the terminal and not on the wx shell. If I do !ls for example the output is listed on the shell I launched the wxIPython shell from. If I hit tab, I get completion (on the wx widget) but I also loose keyboard focus. These sound like small linux specific issues. > > Can you please send me the result of: > Import locale > print locale.getdefaultlocale()? In [1]: import locale In [2]: print locale.getdefaultlocale() (None, None) HTH. cheers, prabhu From robert.kern at gmail.com Wed Dec 12 19:11:00 2007 From: robert.kern at gmail.com (Robert Kern) Date: Wed, 12 Dec 2007 18:11:00 -0600 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <47607027.3020209@astraw.com> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> <47601D1E.4020004@astraw.com> <47607027.3020209@astraw.com> Message-ID: <47607894.6060607@gmail.com> Andrew Straw wrote: > Robert Kern wrote: >> Andrew Straw wrote: >> >>> Note that numpy.distutils and setuptools are not completely compatible, >>> so this may get you into trouble with numpy/scipy, so be careful there. >>> >> I fixed an issue today where develop didn't run build_src --inplace. What other >> issues are you aware of? >> > http://scipy.org/scipy/numpy/ticket/596 Okay, I just fixed that one in the trunk. Anything else? -- Robert Kern "I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth." -- Umberto Eco From fperez.net at gmail.com Wed Dec 12 20:20:27 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Wed, 12 Dec 2007 18:20:27 -0700 Subject: [IPython-dev] [IPython-user] New IPython Article In-Reply-To: <918899A4-A93A-4871-9C5F-49C540AF477E@gmail.com> References: <918899A4-A93A-4871-9C5F-49C540AF477E@gmail.com> Message-ID: On Dec 12, 2007 5:25 PM, Noah Gift wrote: > Just an FYI, an article I wrote for IBM Developerworks is running in > the AIX/Unix section. I didn't get too heavy into IPython, as I was > covering SNMP too, which is complex enough to talk about: > > http://www.ibm.com/developerworks/aix > > Direct Link: > > http://www.ibm.com/developerworks/aix/library/au-netsnmpnipython/ Very nice! Thanks a lot, I just added mentions of it in the News and Documentation pages. By the way, the 'iterative coding' process you described is exactly the way I (and others I know) work when doing scientific computing development, so I think the appeal of your article goes beyond the SNMP audience. That style of work is why IPython was born, in a sense, and it applies to just about any field. Cheers, f From vivainio at gmail.com Thu Dec 13 03:15:48 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Thu, 13 Dec 2007 10:15:48 +0200 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> <6ce0ac130712121102u2cd26086scf4a40f8858b0998@mail.gmail.com> <46cb515a0712121152t51a25119rc70391715b16e6a0@mail.gmail.com> Message-ID: <46cb515a0712130015i636d2bbamd227643b5d534b72@mail.gmail.com> On Dec 13, 2007 12:38 AM, Robert Kern wrote: > AFAICT, that's a script solely for your or Fernando's benefit, not anyone > else's. For it to be generally useful, it shouldn't add your local site-packages > directory and it shouldn't delete anything. Apparently those were for use with "eggsetup.py develop". I don't use 'develop' myself, so I can't comment on how necessary they are (Jorgen?). Of course it could be fixed further to make it more generally useful. > And why is there also setup_bdist_egg.py? Release scripts belong in tools/. That does not appear to be used by release scripts anymore (Fernando?), so perhaps we could remove it altogether. -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From stefan at xsteve.at Thu Dec 13 10:07:05 2007 From: stefan at xsteve.at (Stefan =?iso-8859-1?Q?Reich=F6r?=) Date: Thu, 13 Dec 2007 16:07:05 +0100 Subject: [IPython-dev] wxIPython References: <20071202230454.GA3072@clipper.ens.fr> <009301c8353f$0c58d820$250a8860$@dufrechou@free.fr> <20071205153301.GB8314@mentat.za.net> <25503.5974005752$1196982395@news.gmane.org> Message-ID: "Laurent Dufrechou" writes: The wxIpython shell is really great. Thanks to all involved. > Hi guys, > > Here is my last version. > I've removed the bug with '?' and probably also autocompletion. > Feel free to report any other bugs. > It doesn't support all the functions of true ipython shell but it's I think > it's a good start. > I need now to do some cleaning, comments etc... > > Current boring unsolved issue: > 'Ls' command result is written in console not into the widget ?_? > If you really need ls just do... __IP.system("dir") > Or whatever your os need :) > How and crlc+c isn't supported... so don't write an infinite loop ! > > Any comment, other bug report welcome. > > Cheers, > Laurent > > ### known bug: > # - currently WxConsoleView is a wx.Panel, can we change it to wx.TextCtrl? > # - autoindent doesn't work > # - I must find a way to handle crtl+c with a os.fork() in case the ipython > do an infinite loop > # while (1): > # a=1 > # - 'pass' key when call make the interpreter to stop current multiline > command > # class a: > # def b(): > # pass > # def c(): > # --> syntax error '):' > # - some color glitch if userr click in a prompt color zone and start to > write text. > # - 'ls' command don't work (and probably all external commands) I tried it on Windows. The ls command and other external commands are working for me. On Linux, ls does not work as intended. And I have another (bigger) problem: The input text is shown as black on black. Setting self.default_text_color does not work on startup. When I hit tab, the correct color is applied. However, On the next input line, it starts again with black on black. I changed the default background to white to have white on black ;-) Another problem with my linux version is, that I have to set the cursor manually after entering a line. Otherwise all input is ignored. I use the following wx version: In [4]: import wx In [5]: wx.version() Out[5]: '2.8.1.1 (gtk2-ansi)' Stefan. From nirvana117 at gmail.com Thu Dec 13 10:17:23 2007 From: nirvana117 at gmail.com (pan) Date: Thu, 13 Dec 2007 23:17:23 +0800 Subject: [IPython-dev] Experimental unicode branch In-Reply-To: <46cb515a0712121510g5bb4a72bsd78e1f59bf422038@mail.gmail.com> References: <46cb515a0712121510g5bb4a72bsd78e1f59bf422038@mail.gmail.com> Message-ID: <19b355330712130717rc1a8bd7nfbac979ed803483b@mail.gmail.com> I Added a file unicode_helper.py, and did some unicode-related changes. Tested on debian(utf-8) and windows xp(cp936). -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: 2893.patch Type: application/octet-stream Size: 8707 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: unicode_helper.py Type: text/x-python Size: 1404 bytes Desc: not available URL: From laurent.dufrechou at gmail.com Thu Dec 13 11:03:52 2007 From: laurent.dufrechou at gmail.com (=?iso-8859-1?Q?Laurent_Dufr=E9chou?=) Date: Thu, 13 Dec 2007 17:03:52 +0100 Subject: [IPython-dev] wxIPython In-Reply-To: References: <20071202230454.GA3072@clipper.ens.fr> <009301c8353f$0c58d820$250a8860$@dufrechou@free.fr> <20071205153301.GB8314@mentat.za.net> <25503.5974005752$1196982395@news.gmane.org> Message-ID: <476157f0.1f205e0a.3795.ffffedb1@mx.google.com> Yes, i've developped on windows only, and linux version is quite buggy! I've currently solved the black on black (self.setForegroundColor(self.default_coloretc...) and another issue with locale. Still missing focus loss on keyboard input and stdin/out redirection problem. Thanks for your report. By the way, how about the speed? I've tried on linux in my virtual machine and it looks like very slow on my T7200. Have you noticed such problem??? -----Message d'origine----- De?: ipython-dev-bounces at scipy.org [mailto:ipython-dev-bounces at scipy.org] De la part de Stefan Reich?r Envoy??: jeudi 13 d?cembre 2007 16:07 ??: ipython-dev at scipy.net Objet?: Re: [IPython-dev] wxIPython "Laurent Dufrechou" writes: The wxIpython shell is really great. Thanks to all involved. > Hi guys, > > Here is my last version. > I've removed the bug with '?' and probably also autocompletion. > Feel free to report any other bugs. > It doesn't support all the functions of true ipython shell but it's I think > it's a good start. > I need now to do some cleaning, comments etc... > > Current boring unsolved issue: > 'Ls' command result is written in console not into the widget ?_? > If you really need ls just do... __IP.system("dir") > Or whatever your os need :) > How and crlc+c isn't supported... so don't write an infinite loop ! > > Any comment, other bug report welcome. > > Cheers, > Laurent > > ### known bug: > # - currently WxConsoleView is a wx.Panel, can we change it to wx.TextCtrl? > # - autoindent doesn't work > # - I must find a way to handle crtl+c with a os.fork() in case the ipython > do an infinite loop > # while (1): > # a=1 > # - 'pass' key when call make the interpreter to stop current multiline > command > # class a: > # def b(): > # pass > # def c(): > # --> syntax error '):' > # - some color glitch if userr click in a prompt color zone and start to > write text. > # - 'ls' command don't work (and probably all external commands) I tried it on Windows. The ls command and other external commands are working for me. On Linux, ls does not work as intended. And I have another (bigger) problem: The input text is shown as black on black. Setting self.default_text_color does not work on startup. When I hit tab, the correct color is applied. However, On the next input line, it starts again with black on black. I changed the default background to white to have white on black ;-) Another problem with my linux version is, that I have to set the cursor manually after entering a line. Otherwise all input is ignored. I use the following wx version: In [4]: import wx In [5]: wx.version() Out[5]: '2.8.1.1 (gtk2-ansi)' Stefan. _______________________________________________ IPython-dev mailing list IPython-dev at scipy.org http://lists.ipython.scipy.org/mailman/listinfo/ipython-dev From vivainio at gmail.com Thu Dec 13 15:59:32 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Thu, 13 Dec 2007 22:59:32 +0200 Subject: [IPython-dev] Experimental unicode branch In-Reply-To: <19b355330712130717rc1a8bd7nfbac979ed803483b@mail.gmail.com> References: <46cb515a0712121510g5bb4a72bsd78e1f59bf422038@mail.gmail.com> <19b355330712130717rc1a8bd7nfbac979ed803483b@mail.gmail.com> Message-ID: <46cb515a0712131259j6a242a26laa6bd4ca06a74b20@mail.gmail.com> On Dec 13, 2007 5:17 PM, pan wrote: > I Added a file unicode_helper.py, and did some unicode-related changes. > Tested on debian(utf-8) and windows xp(cp936). Ok, I'll commit these. However, I will rename ensure_str and ensure_unicode to sstr and sunicode (s is for "safe"), if you don't mind. -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From robert.kern at gmail.com Fri Dec 14 14:29:33 2007 From: robert.kern at gmail.com (Robert Kern) Date: Fri, 14 Dec 2007 13:29:33 -0600 Subject: [IPython-dev] Experimental unicode branch In-Reply-To: <46cb515a0712131259j6a242a26laa6bd4ca06a74b20@mail.gmail.com> References: <46cb515a0712121510g5bb4a72bsd78e1f59bf422038@mail.gmail.com> <19b355330712130717rc1a8bd7nfbac979ed803483b@mail.gmail.com> <46cb515a0712131259j6a242a26laa6bd4ca06a74b20@mail.gmail.com> Message-ID: Ville M. Vainio wrote: > On Dec 13, 2007 5:17 PM, pan wrote: >> I Added a file unicode_helper.py, and did some unicode-related changes. >> Tested on debian(utf-8) and windows xp(cp936). > > Ok, I'll commit these. However, I will rename ensure_str and > ensure_unicode to sstr and sunicode (s is for "safe"), if you don't > mind. How about terminal_str and terminal_unicode? These functions are for encoding to and from the terminal encoding and aren't generally useful functions. -- Robert Kern "I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth." -- Umberto Eco From jorgen.stenarson at bostream.nu Fri Dec 14 16:57:34 2007 From: jorgen.stenarson at bostream.nu (=?ISO-8859-1?Q?J=F6rgen_Stenarson?=) Date: Fri, 14 Dec 2007 22:57:34 +0100 Subject: [IPython-dev] Adding setupttols to iptyhon's setup.py In-Reply-To: <46cb515a0712130015i636d2bbamd227643b5d534b72@mail.gmail.com> References: <20071212141131.GE14948@clipper.ens.fr> <46cb515a0712120900w7a9a303fj584feeb4c7572702@mail.gmail.com> <47601999.9090001@astraw.com> <20071212173336.GA21815@clipper.ens.fr> <6ce0ac130712121102u2cd26086scf4a40f8858b0998@mail.gmail.com> <46cb515a0712121152t51a25119rc70391715b16e6a0@mail.gmail.com> <46cb515a0712130015i636d2bbamd227643b5d534b72@mail.gmail.com> Message-ID: <4762FC4E.9020201@bostream.nu> Ville M. Vainio skrev: > On Dec 13, 2007 12:38 AM, Robert Kern wrote: > >> AFAICT, that's a script solely for your or Fernando's benefit, not anyone >> else's. For it to be generally useful, it shouldn't add your local site-packages >> directory and it shouldn't delete anything. > > Apparently those were for use with "eggsetup.py develop". I don't use > 'develop' myself, so I can't comment on how necessary they are > (Jorgen?). Of course it could be fixed further to make it more > generally useful. > I haven't used them for quite a while so I don't know how well they work /J?rgen From laurent.dufrechou at gmail.com Sat Dec 15 10:17:37 2007 From: laurent.dufrechou at gmail.com (=?US-ASCII?Q?Laurent_Dufrechou?=) Date: Sat, 15 Dec 2007 16:17:37 +0100 Subject: [IPython-dev] wxIPython0.7 - Linux is a known OS :) Message-ID: <4763f01d.0c07560a.7f62.3d2c@mx.google.com> Hi IPythoners, Here the last version of WXIpython. - Problems under linux should be resolved now.(wxGtk textctrl doesn't like Freeze/Thaw...) - Added filters on history widget, you can filter ('','!' and '?' commands) - wx Pager is now thread safe ,linux is less tolerant than windows ;) Instead of being called by IPython thread it is managed by wxThread via a more complex state machine. Any better idea welcomed :) ###### for ipython devs and pyreadline too ###### Will try to put it inside SVN, Fernando,Ville where should I add it? Joergen,Gael, I would like to make the App Pyreadline aware :) Where should I start? I've got class WxConsoleView(wx.Panel): that handle all the text processing, should I try to convert it like ironpython console class in pyreadline? Is it what I should do to be pyreadline compatible?If any of you want to help no problem :) The app is divided in 3 parts: WxPython.py that handle the high level GUI. Ipython_history.py that is a simple widget that is a helper.(basically a copy/paste buffer) Ipython_view.py: 3 classes inside it: - class WxConsoleView(wx.Panel) <-- handle all the action on textctrl. - class IterableIPShell(Thread) <-- will rename it to IPythonMTShellContainer This is a thread that receive line to execute and returns history info and execution info/result. - class IPythonView(WxConsoleView) <-- create the IPython thread, the text_ctrl and handle the communication between the thread,the text_ctrl and the top level GUI via state machine on wx.Idle event. Laurent -------------- next part -------------- A non-text attachment was scrubbed... Name: wxIPython0.7.zip Type: application/x-zip-compressed Size: 12042 bytes Desc: not available URL: From fperez.net at gmail.com Sat Dec 15 14:45:46 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sat, 15 Dec 2007 12:45:46 -0700 Subject: [IPython-dev] wxIPython0.7 - Linux is a known OS :) In-Reply-To: <4763f01d.0c07560a.7f62.3d2c@mx.google.com> References: <4763f01d.0c07560a.7f62.3d2c@mx.google.com> Message-ID: On Dec 15, 2007 8:17 AM, Laurent Dufrechou wrote: > Hi IPythoners, > > Here the last version of WXIpython. > - Problems under linux should be resolved now.(wxGtk textctrl doesn't like > Freeze/Thaw...) > - Added filters on history widget, you can filter ('','!' and '?' commands) > - wx Pager is now thread safe ,linux is less tolerant than windows ;) > Instead of being called by IPython thread it is managed by wxThread via a > more complex state machine. Any better idea welcomed :) Great! > ###### for ipython devs and pyreadline too ###### > > Will try to put it inside SVN, Fernando,Ville where should I add it? I'd suggest a whole new subpackage called wx, under ipython: from IPython.wx import ... so you'd make a directory called wx/ under IPython. I hope soon we'll be using this under the ipython refactoring, so having it as self-contained as possible will also make that transition easier. But there's enough value in what you're doing *right now* that I think we should have it in for today's ipython. Cheers, f From prabhu at aero.iitb.ac.in Sun Dec 16 02:03:38 2007 From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran) Date: Sun, 16 Dec 2007 12:33:38 +0530 Subject: [IPython-dev] wxIPython0.7 - Linux is a known OS :) In-Reply-To: <4763f01d.0c07560a.7f62.3d2c@mx.google.com> References: <4763f01d.0c07560a.7f62.3d2c@mx.google.com> Message-ID: <4764CDCA.7060108@aero.iitb.ac.in> Laurent Dufrechou wrote: > Hi IPythoners, > > Here the last version of WXIpython. > - Problems under linux should be resolved now.(wxGtk textctrl doesn't like > Freeze/Thaw...) > - Added filters on history widget, you can filter ('','!' and '?' commands) > - wx Pager is now thread safe ,linux is less tolerant than windows ;) > Instead of being called by IPython thread it is managed by wxThread via a > more complex state machine. Any better idea welcomed :) Here is a quick note, not much time to look into it in greater detail. Here are some test reports for me under Linux (Ubuntu gutsy), The bad: - I cant see what I type. So the WHITE color isn't working. - If I type just '?' on the shell I get the ipython help on the terminal. The good: - !ls is nicely shown in the wxIPython shell - Tab complete shows up nicely. - something? or something?? works very nicely. So apart from a few small niggles this does appear to work very nicely. Thanks! cheers, prabhu From prabhu at aero.iitb.ac.in Sun Dec 16 02:08:15 2007 From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran) Date: Sun, 16 Dec 2007 12:38:15 +0530 Subject: [IPython-dev] wxIPython0.7 - Linux is a known OS :) In-Reply-To: <4764CDCA.7060108@aero.iitb.ac.in> References: <4763f01d.0c07560a.7f62.3d2c@mx.google.com> <4764CDCA.7060108@aero.iitb.ac.in> Message-ID: <4764CEDF.6060004@aero.iitb.ac.in> Prabhu Ramachandran wrote: > Laurent Dufrechou wrote: >> Hi IPythoners, >> >> Here the last version of WXIpython. >> - Problems under linux should be resolved now.(wxGtk textctrl doesn't like >> Freeze/Thaw...) >> - Added filters on history widget, you can filter ('','!' and '?' commands) >> - wx Pager is now thread safe ,linux is less tolerant than windows ;) >> Instead of being called by IPython thread it is managed by wxThread via a >> more complex state machine. Any better idea welcomed :) > > Here is a quick note, not much time to look into it in greater detail. > Here are some test reports for me under Linux (Ubuntu gutsy), > > The bad: > - I cant see what I type. So the WHITE color isn't working. > - If I type just '?' on the shell I get the ipython help on the terminal. Sorry missed on rather major bad point which is that wxIPython seems to chew up a huge amount of CPU. It takes about 95% of CPU when running. cheers, prabhu From laurent.dufrechou at gmail.com Sun Dec 16 06:12:44 2007 From: laurent.dufrechou at gmail.com (=?iso-8859-1?Q?Laurent_Dufr=E9chou?=) Date: Sun, 16 Dec 2007 12:12:44 +0100 Subject: [IPython-dev] wxIPython0.7 - Linux is a known OS :) In-Reply-To: <4764CEDF.6060004@aero.iitb.ac.in> References: <4763f01d.0c07560a.7f62.3d2c@mx.google.com> <4764CDCA.7060108@aero.iitb.ac.in> <4764CEDF.6060004@aero.iitb.ac.in> Message-ID: <47650836.0b38560a.1c57.3023@mx.google.com> Oups, Can you try this in ipython_view.py? Line 671, add time_loop argument = 0.1 s IP = IterableIPShell(cout=self.cout,cerr=self.cout, pager = self.pager, exit_handler = exit_handler, time_loop = 0.1) Cpu goes low with this here.(In with actual config thread is always looping like while(1): if do_execute: .... .... ^_^ With this option: While(1) If do execute: ... Sleep(0.1) For the WHITE_COLOR it is really strange... will check this... For the '?' only commands, well it is an unsupported command for now :) Will add support soon :) If you know other magic commands like these tell me them because I don't know all IPython commands :) Thanks for your report! Laurent -----Message d'origine----- De?: Prabhu Ramachandran [mailto:prabhu at aero.iitb.ac.in] Envoy??: dimanche 16 d?cembre 2007 08:08 ??: Laurent Dufrechou Cc?: IPython-dev at scipy.org Objet?: Re: [IPython-dev] wxIPython0.7 - Linux is a known OS :) Prabhu Ramachandran wrote: > Laurent Dufrechou wrote: >> Hi IPythoners, >> >> Here the last version of WXIpython. >> - Problems under linux should be resolved now.(wxGtk textctrl doesn't like >> Freeze/Thaw...) >> - Added filters on history widget, you can filter ('','!' and '?' commands) >> - wx Pager is now thread safe ,linux is less tolerant than windows ;) >> Instead of being called by IPython thread it is managed by wxThread via a >> more complex state machine. Any better idea welcomed :) > > Here is a quick note, not much time to look into it in greater detail. > Here are some test reports for me under Linux (Ubuntu gutsy), > > The bad: > - I cant see what I type. So the WHITE color isn't working. > - If I type just '?' on the shell I get the ipython help on the terminal. Sorry missed on rather major bad point which is that wxIPython seems to chew up a huge amount of CPU. It takes about 95% of CPU when running. cheers, prabhu From prabhu at aero.iitb.ac.in Sun Dec 16 15:57:02 2007 From: prabhu at aero.iitb.ac.in (Prabhu Ramachandran) Date: Mon, 17 Dec 2007 02:27:02 +0530 Subject: [IPython-dev] wxIPython0.7 - Linux is a known OS :) In-Reply-To: <47650836.0b38560a.1c57.3023@mx.google.com> References: <4763f01d.0c07560a.7f62.3d2c@mx.google.com> <4764CDCA.7060108@aero.iitb.ac.in> <4764CEDF.6060004@aero.iitb.ac.in> <47650836.0b38560a.1c57.3023@mx.google.com> Message-ID: <4765911E.9090500@aero.iitb.ac.in> Laurent Dufr?chou wrote: > Oups, > Can you try this in ipython_view.py? > Line 671, add time_loop argument = 0.1 s > IP = IterableIPShell(cout=self.cout,cerr=self.cout, > pager = self.pager, exit_handler = exit_handler, > time_loop = 0.1) > > Cpu goes low with this here.(In with actual config thread is always looping > like Yup, that reduces the CPU load a huge amount. Thanks. I noticed that large modules take forever for tab completion. Try this: import wx wx. eats lots of CPU and takes forever, perhaps because you end up printing 3000 odd lines?! > Thanks for your report! Np. Thanks! cheers, prabhu From laurent.dufrechou at gmail.com Sun Dec 16 17:30:56 2007 From: laurent.dufrechou at gmail.com (=?iso-8859-1?Q?Laurent_Dufr=E9chou?=) Date: Sun, 16 Dec 2007 23:30:56 +0100 Subject: [IPython-dev] wxIPython0.7 - Linux is a known OS :) References: <4763f01d.0c07560a.7f62.3d2c@mx.google.com> <4764CDCA.7060108@aero.iitb.ac.in> <4764CEDF.6060004@aero.iitb.ac.in> <47650836.0b38560a.1c57.3023@mx.google.com> <4765911E.9090500@aero.iitb.ac.in> Message-ID: <4765a72e.1358560a.1efb.7604@mx.google.com> Hi Prabhu my linux tester ;), Here it takes ~2.5s.(Core 2 T7200) Under linux (in a virtual machine) it tooks more times but more boring, the app more or less unusable because if you try to up or down in the scroll windows it took VERY long time. What is you processor power (just to have a comparison point)? On my windows all is ok (well if a compare to IPython Dos shell it's pity but well... :) Haaaa, find something while writing the email! Can you check with these two files?(really faster :) ) (rename them ipython_view.py) Fastest is like IPython Shell prompt '1 column autocompletion' Fastest but beautifull is little slower but multi column capable (if your window size is large enough) For your config is the last one acceptable? Is there any visible difference between the two? Do you have the BORING slowdown while trying to scroll up after a wx. Completion with GTK? Regards, Laurent Ps:yeah will use svn for next release ;) -----Message d'origine----- De?: Prabhu Ramachandran [mailto:prabhu at aero.iitb.ac.in] Envoy??: dimanche 16 d?cembre 2007 21:57 ??: Laurent Dufr?chou Cc?: IPython-dev at scipy.org Objet?: Re: [IPython-dev] wxIPython0.7 - Linux is a known OS :) Laurent Dufr?chou wrote: > Oups, > Can you try this in ipython_view.py? > Line 671, add time_loop argument = 0.1 s IP = > IterableIPShell(cout=self.cout,cerr=self.cout, > pager = self.pager, exit_handler = exit_handler, > time_loop = 0.1) > > Cpu goes low with this here.(In with actual config thread is always > looping like Yup, that reduces the CPU load a huge amount. Thanks. I noticed that large modules take forever for tab completion. Try this: import wx wx. eats lots of CPU and takes forever, perhaps because you end up printing 3000 odd lines?! > Thanks for your report! Np. Thanks! cheers, prabhu -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: ipython_view_fast_beautifull_completion.py URL: -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: ipython_view_fastest_completion.py URL: From jorgen.stenarson at bostream.nu Mon Dec 17 13:38:12 2007 From: jorgen.stenarson at bostream.nu (=?ISO-8859-1?Q?J=F6rgen_Stenarson?=) Date: Mon, 17 Dec 2007 19:38:12 +0100 Subject: [IPython-dev] pyreadline callback branch Message-ID: <4766C214.8090108@bostream.nu> Stefan, I have now created a branch of pyreadline to work on the callback stuff. I have based it on the example you provided at the beginning of the summer. My changes to that script are available in the examples folder of a checkout of the callback branch. At this time it only works for the emacs mode. I will have to decide on a way to integrate it into the vi mode as well. /J?rgen From jorgen.stenarson at bostream.nu Mon Dec 17 13:58:20 2007 From: jorgen.stenarson at bostream.nu (=?ISO-8859-1?Q?J=F6rgen_Stenarson?=) Date: Mon, 17 Dec 2007 19:58:20 +0100 Subject: [IPython-dev] wxIPython0.7 - Linux is a known OS :) In-Reply-To: <4763f01d.0c07560a.7f62.3d2c@mx.google.com> References: <4763f01d.0c07560a.7f62.3d2c@mx.google.com> Message-ID: <4766C6CC.9080704@bostream.nu> Laurent Dufrechou skrev: > > ###### for ipython devs and pyreadline too ###### > > Will try to put it inside SVN, Fernando,Ville where should I add it? > > Joergen,Gael, I would like to make the App Pyreadline aware :) > Where should I start? I've got class WxConsoleView(wx.Panel): that handle > all the text processing, should I try to convert it like ironpython console > class in pyreadline? > Is it what I should do to be pyreadline compatible?If any of you want to > help no problem :) > Laurent, I think looking at the ironpython console class is a good start you should probably also create at least a dummy class for clipboard functionality since that is ctypes/win32 based as well. I also think we will have to figure out a new way to choose backend. The way ipython in a regular console window gets benefit from pyreadline is by calling rawinput which uses pyreadline, in order to get colored output there is also a special write routine that interprets ansi color codes. Exactly how you should do this in wx I don't know. I'm no gui expert. However I have just comitted a first try of a callback interface which may be what you need (currenlty only available in emacs mode). You can get to it in the callback branch at: http://ipython.scipy.org/svn/ipython/pyreadline/branches/callback/ There is an example in pyreadline/examples/callback_examples.py Good luck and feel free to ask questions. I will be offline from thursday dec 20th until jan 2. /J?rgen From vivainio at gmail.com Thu Dec 20 02:03:43 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Thu, 20 Dec 2007 09:03:43 +0200 Subject: [IPython-dev] import mercurial.hg screws the terminal Message-ID: <46cb515a0712192303h3f5d5caem4a6bbac31962050e@mail.gmail.com> Something we just noticed: On win32, if you have mercurial installed (as a library), if you do "import mercurial.hg" on interactive prompt, you lose all the colors and the color escape sequences are just dumped to screen with gray characters. Does the same happen on Linux? -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From fperez.net at gmail.com Thu Dec 20 03:07:20 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Thu, 20 Dec 2007 01:07:20 -0700 Subject: [IPython-dev] import mercurial.hg screws the terminal In-Reply-To: <46cb515a0712192303h3f5d5caem4a6bbac31962050e@mail.gmail.com> References: <46cb515a0712192303h3f5d5caem4a6bbac31962050e@mail.gmail.com> Message-ID: On Dec 20, 2007 12:03 AM, Ville M. Vainio wrote: > Something we just noticed: > > On win32, if you have mercurial installed (as a library), if you do > "import mercurial.hg" on interactive prompt, you lose all the colors > and the color escape sequences are just dumped to screen with gray > characters. Does the same happen on Linux? No problem on my laptop (feisty) or my desktop (gutsy). I can 'import mercurial.hg' and the terminal continues behaving normally. Cheers, f From list-ener at strank.info Thu Dec 20 03:14:20 2007 From: list-ener at strank.info (Stefan Rank) Date: Thu, 20 Dec 2007 09:14:20 +0100 Subject: [IPython-dev] pyreadline callback branch In-Reply-To: <4766C214.8090108@bostream.nu> References: <4766C214.8090108@bostream.nu> Message-ID: <476A245C.4050704@strank.info> Hi J?rgen, on 17.12.2007 19:38 J?rgen Stenarson said the following: > Stefan, > > I have now created a branch of pyreadline to work on the callback stuff. > I have based it on the example you provided at the beginning of the > summer. My changes to that script are available in the examples folder > of a checkout of the callback branch. > > At this time it only works for the emacs mode. I will have to decide on > a way to integrate it into the vi mode as well. Great! I checked out the branch. Unfortunately, it will be some time until I can look into it again... I hope to have some time during the holidays. thanks, stefan From vivainio at gmail.com Fri Dec 28 07:31:24 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Fri, 28 Dec 2007 14:31:24 +0200 Subject: [IPython-dev] [IPython] #203: Missing/broken license informations In-Reply-To: <049.65a5d285456ce673e206a84360388402@scipy.org> References: <040.dd21e8261ea0234318ac5c8ae7d7b1f5@scipy.org> <049.65a5d285456ce673e206a84360388402@scipy.org> Message-ID: <46cb515a0712280431w30ebc79dhca30c9d73410276d@mail.gmail.com> On Dec 28, 2007 2:13 PM, IPython wrote: > #203: Missing/broken license informations > Ticket URL: I have now added the license notes to simplegeneric.py and Debugger.py: http://ipython.scipy.org/ipython/ipython/changeset/2902 Please verify the change before I close the ticket. The "big deal" here is changing Debugger.py to be under PSF instead of BSD. There should not be any license incompatibility issues - BSD code can import PSF code freely, but we just can't change code under PSF license to be under BSD. It' -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From vivainio at gmail.com Fri Dec 28 07:32:43 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Fri, 28 Dec 2007 14:32:43 +0200 Subject: [IPython-dev] [IPython] #203: Missing/broken license informations In-Reply-To: <46cb515a0712280431w30ebc79dhca30c9d73410276d@mail.gmail.com> References: <040.dd21e8261ea0234318ac5c8ae7d7b1f5@scipy.org> <049.65a5d285456ce673e206a84360388402@scipy.org> <46cb515a0712280431w30ebc79dhca30c9d73410276d@mail.gmail.com> Message-ID: <46cb515a0712280432r40451ddam4dc9e12ea2eb9c5a@mail.gmail.com> On Dec 28, 2007 2:31 PM, Ville M. Vainio wrote: > under BSD. It' Pressed "send" accidentally. What I meant to say is: it's much cleaner to change this part of our code to be under PSF than insist on everything being under BSD. -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From fperez.net at gmail.com Fri Dec 28 12:36:00 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Fri, 28 Dec 2007 10:36:00 -0700 Subject: [IPython-dev] [IPython] #203: Missing/broken license informations In-Reply-To: <46cb515a0712280432r40451ddam4dc9e12ea2eb9c5a@mail.gmail.com> References: <040.dd21e8261ea0234318ac5c8ae7d7b1f5@scipy.org> <049.65a5d285456ce673e206a84360388402@scipy.org> <46cb515a0712280431w30ebc79dhca30c9d73410276d@mail.gmail.com> <46cb515a0712280432r40451ddam4dc9e12ea2eb9c5a@mail.gmail.com> Message-ID: On Dec 28, 2007 5:32 AM, Ville M. Vainio wrote: > On Dec 28, 2007 2:31 PM, Ville M. Vainio wrote: > > > under BSD. It' > > Pressed "send" accidentally. What I meant to say is: > > it's much cleaner to change this part of our code to be under PSF than > insist on everything being under BSD. Yup, go for it. Making PSF code BSD was just an inadvertent mistake, nothing more. Thanks for cleaning things up. Cheers, f From ondrej at certik.cz Sun Dec 30 16:57:54 2007 From: ondrej at certik.cz (Ondrej Certik) Date: Sun, 30 Dec 2007 22:57:54 +0100 Subject: [IPython-dev] please include documentation sources in the tarball In-Reply-To: <85b5c3130712301225s3b22d3e3q61bce4f4d81cd222@mail.gmail.com> References: <85b5c3130712301225s3b22d3e3q61bce4f4d81cd222@mail.gmail.com> Message-ID: <85b5c3130712301357j38a98fdiede5988c212319b3@mail.gmail.com> Hi, while packaging ipython in Debian, we run across this problem: http://projects.scipy.org/ipython/ipython/ticket/204 do you think it could be please fixed in the next release? As written in the ticket, the easiest solution is to include the sources for the documentation in the tarball. The rest can be handled by the Debian maintainer easily. Thanks a lot! Ondrej P.S. Please CC me. From fperez.net at gmail.com Sun Dec 30 17:25:28 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 30 Dec 2007 15:25:28 -0700 Subject: [IPython-dev] please include documentation sources in the tarball In-Reply-To: <85b5c3130712301357j38a98fdiede5988c212319b3@mail.gmail.com> References: <85b5c3130712301225s3b22d3e3q61bce4f4d81cd222@mail.gmail.com> <85b5c3130712301357j38a98fdiede5988c212319b3@mail.gmail.com> Message-ID: On Dec 30, 2007 2:57 PM, Ondrej Certik wrote: > Hi, > > while packaging ipython in Debian, we run across this problem: > > http://projects.scipy.org/ipython/ipython/ticket/204 > > do you think it could be please fixed in the next release? Done: http://ipython.scipy.org/ipython/ipython/changeset/2909 Feel free to comment further on the ticket if you think I missed anything. cheers, f From ondrej at certik.cz Sun Dec 30 17:31:47 2007 From: ondrej at certik.cz (Ondrej Certik) Date: Sun, 30 Dec 2007 23:31:47 +0100 Subject: [IPython-dev] please include documentation sources in the tarball In-Reply-To: References: <85b5c3130712301225s3b22d3e3q61bce4f4d81cd222@mail.gmail.com> <85b5c3130712301357j38a98fdiede5988c212319b3@mail.gmail.com> Message-ID: <85b5c3130712301431r1d9b9f1djd89751f194e4985c@mail.gmail.com> Hi Fernando, On Dec 30, 2007 11:25 PM, Fernando Perez wrote: > On Dec 30, 2007 2:57 PM, Ondrej Certik wrote: > > Hi, > > > > while packaging ipython in Debian, we run across this problem: > > > > http://projects.scipy.org/ipython/ipython/ticket/204 > > > > do you think it could be please fixed in the next release? > > Done: > > http://ipython.scipy.org/ipython/ipython/changeset/2909 > > Feel free to comment further on the ticket if you think I missed anything. Thanks a lot! I am CCing to the respective Debian bug report: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=458383 if there are further problems, I'll let you know. Ondrej