From metatracker at psf.upfronthosting.co.za Mon Mar 1 10:27:27 2010 From: metatracker at psf.upfronthosting.co.za (Florent Xicluna) Date: Mon, 01 Mar 2010 09:27:27 +0000 Subject: [Tracker-discuss] [issue323] email is not obfuscated in the msg In-Reply-To: <1267435647.66.0.658210680138.issue323@psf.upfronthosting.co.za> Message-ID: <1267435647.66.0.658210680138.issue323@psf.upfronthosting.co.za> New submission from Florent Xicluna : When you reply to a message using the SMTP gateway, the e-mail of the originator of the previous message is displayed in clear text. It is a designated target for spammers. See the replies of MAL in this thread http://bugs.python.org/issue7475 for example. ---------- messages: 1571 nosy: flox priority: critical status: unread title: email is not obfuscated in the msg _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Thu Mar 11 21:50:04 2010 From: metatracker at psf.upfronthosting.co.za (Jim Baker) Date: Thu, 11 Mar 2010 20:50:04 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <1265297768.52.0.920463244949.issue316@psf.upfronthosting.co.za> Message-ID: <1268340604.99.0.317824635322.issue316@psf.upfronthosting.co.za> Jim Baker added the comment: This is impacting development. Users are slinging patches around on our mailing list, or are outright confused. ---------- nosy: +zyasoft priority: bug -> critical _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Thu Mar 11 21:57:21 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Thu, 11 Mar 2010 20:57:21 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <1265297768.52.0.920463244949.issue316@psf.upfronthosting.co.za> Message-ID: <1268341041.83.0.98491620864.issue316@psf.upfronthosting.co.za> Martin v. L?wis added the comment: When we setup this tracker, we agreed that the Jython commmunity would provide a volunteer to look into issues. IIRC, Raghuram volunteered to do that. It would speed up processing if somebody would provide a patch. ---------- nosy: +loewis _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Thu Mar 11 22:00:29 2010 From: metatracker at psf.upfronthosting.co.za (Nicholas Riley) Date: Thu, 11 Mar 2010 21:00:29 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <1265297768.52.0.920463244949.issue316@psf.upfronthosting.co.za> Message-ID: <1268341229.41.0.11876628389.issue316@psf.upfronthosting.co.za> Nicholas Riley added the comment: Isn't this just a configuration issue? _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Thu Mar 11 22:00:55 2010 From: metatracker at psf.upfronthosting.co.za (Raghuram Devarakonda) Date: Thu, 11 Mar 2010 21:00:55 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <1268341041.83.0.98491620864.issue316@psf.upfronthosting.co.za> Message-ID: <2c51ecee1003111300r6a5aa93fof9d14d49fde2166c@mail.gmail.com> Raghuram Devarakonda added the comment: On Thu, Mar 11, 2010 at 3:57 PM, Martin v. L?wis wrote: > When we setup this tracker, we agreed that the Jython commmunity would provide a volunteer to look into issues. IIRC, Raghuram volunteered to do that. It would speed up processing if somebody would provide a patch. As it happens, I have just started looking into Python wiki for tracker set up page so that I can create a local tracker instance. I will post here once I am able to set it up and recreate the issue. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Thu Mar 11 22:08:27 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Thu, 11 Mar 2010 21:08:27 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <1268341229.41.0.11876628389.issue316@psf.upfronthosting.co.za> Message-ID: <4B995BC7.8060704@v.loewis.de> Martin v. L?wis added the comment: > Isn't this just a configuration issue? The configuration is a Python source file, stored in subversion. So any configuration change can be provided as a patch. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Thu Mar 11 22:45:50 2010 From: metatracker at psf.upfronthosting.co.za (Raghuram Devarakonda) Date: Thu, 11 Mar 2010 21:45:50 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <2c51ecee1003111300r6a5aa93fof9d14d49fde2166c@mail.gmail.com> Message-ID: <2c51ecee1003111345y2293ab17y69bbf89572c49071@mail.gmail.com> Raghuram Devarakonda added the comment: On Thu, Mar 11, 2010 at 4:00 PM, Raghuram Devarakonda wrote: > As it happens, I have just started looking into Python wiki for > tracker set up page so that I can create a local tracker instance. I > will post here once I am able to set it up and recreate the issue. Martin, I created a test issue (http://bugs.jython.org/issue1570) and uploaded two files successfully. Considering that I have Developer and Coordinator role, I am guessing that the upload problem is only occurring for people who have only User roles. Does this throw any light on the problem? I started setting up the local tracker instance of python-dev (I wanted to have a working python-dev before attempting jython as TrackerDevelopment wiki page talks about python-dev) and am finding that the installation requires lot more packages than are listed in the Wiki page (eg, BeautifulSoup, M2Crypto). _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Thu Mar 11 23:09:58 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Thu, 11 Mar 2010 22:09:58 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <2c51ecee1003111345y2293ab17y69bbf89572c49071@mail.gmail.com> Message-ID: <4B996A34.4010602@v.loewis.de> Martin v. L?wis added the comment: > Martin, I created a test issue (http://bugs.jython.org/issue1570) and > uploaded two files successfully. Considering that I have Developer and > Coordinator role, I am guessing that the upload problem is only > occurring for people who have only User roles. Does this throw any > light on the problem? Not really. "User" has "Create" permission on "file", and Edit permission for the "files" property of "issue". So it all ought to work fine. > I started setting up the local tracker instance of python-dev (I > wanted to have a working python-dev before attempting jython as > TrackerDevelopment wiki page talks about python-dev) and am finding > that the installation requires lot more packages than are listed in > the Wiki page (eg, BeautifulSoup, M2Crypto). These are only needed for the OpenID support in python-dev; the jython tracker wouldn't need these. In any case, feel free to correct the Wiki. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 12 05:02:25 2010 From: metatracker at psf.upfronthosting.co.za (Stephen Turnbull) Date: Fri, 12 Mar 2010 04:02:25 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <4B995BC7.8060704@v.loewis.de> Message-ID: <87iq92rxeq.fsf@uwakimon.sk.tsukuba.ac.jp> Stephen Turnbull added the comment: Martin v. L?wis writes: > > Martin v. L?wis added the comment: > > > Isn't this just a configuration issue? > > The configuration is a Python source file, stored in subversion. So any > configuration change can be provided as a patch. And should be provided as a patch. I manage a different instance of Roundup, not Python's, but I can say it's really a great help to me in the case that a third party provides a fix or improvement if it comes as a patch generated by Subversion (or whatever VCS). Yes, it's a bit of aggravation for one-off contributors, but enough people are happy to do it that it's worth asking. ---------- nosy: +stephen _______________________________________________________ PSF Meta Tracker _______________________________________________________ From stephen at xemacs.org Fri Mar 12 05:12:13 2010 From: stephen at xemacs.org (Stephen J. Turnbull) Date: Fri, 12 Mar 2010 13:12:13 +0900 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <4B995BC7.8060704@v.loewis.de> References: <1268341229.41.0.11876628389.issue316@psf.upfronthosting.co.za> <4B995BC7.8060704@v.loewis.de> Message-ID: <87iq92rxeq.fsf@uwakimon.sk.tsukuba.ac.jp> Martin v. L?wis writes: > > Martin v. L?wis added the comment: > > > Isn't this just a configuration issue? > > The configuration is a Python source file, stored in subversion. So any > configuration change can be provided as a patch. And should be provided as a patch. I manage a different instance of Roundup, not Python's, but I can say it's really a great help to me in the case that a third party provides a fix or improvement if it comes as a patch generated by Subversion (or whatever VCS). Yes, it's a bit of aggravation for one-off contributors, but enough people are happy to do it that it's worth asking. From metatracker at psf.upfronthosting.co.za Fri Mar 12 16:27:10 2010 From: metatracker at psf.upfronthosting.co.za (Raghuram Devarakonda) Date: Fri, 12 Mar 2010 15:27:10 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <4B996A34.4010602@v.loewis.de> Message-ID: <2c51ecee1003120726k4a1b6c69o8fd93e0b3fad3252@mail.gmail.com> Raghuram Devarakonda added the comment: On Thu, Mar 11, 2010 at 5:09 PM, Martin v. L?wis wrote: >> Martin, I created a test issue (http://bugs.jython.org/issue1570) and >> uploaded two files successfully. Considering that I have Developer and >> Coordinator role, I am guessing that the upload problem is only >> occurring for people who have only User roles. Does this throw any >> light on the problem? > > Not really. "User" has "Create" permission on "file", and Edit > permission for the "files" property of "issue". So it all ought to work > fine. I have tracked down the problem to the following check-in to 'roundup-src': ------------------------------------------------------------------------ r76545 | martin.v.loewis | 2009-11-27 06:58:04 -0500 (Fri, 27 Nov 2009) | 3 lines Upgrade to 1.4.10, taken from http://pypi.python.org/pypi/roundup/1.4.10. -------------- In particular, to the following part of check-in: ---------------- Index: roundup/cgi/actions.py =================================================================== --- roundup/cgi/actions.py (revision 76544) +++ roundup/cgi/actions.py (revision 76545) @@ -544,10 +544,26 @@ Base behaviour is to check the user can edit this class. No additional property checks are made. """ + if not classname : classname = self.client.classname - return self.hasPermission('Create', classname=classname) + + if not self.hasPermission('Create', classname=classname): + return 0 + # Check Edit permission for each property, to avoid being able + # to set restricted ones on new item creation + for key in props: + if not self.hasPermission('Edit', classname=classname, + property=key): + # We restrict by default and special-case allowed properties + if key == 'date' or key == 'content': + continue + elif key == 'author' and props[key] == self.userid: + continue + return 0 + return 1 + ---------------- The value of "props" in my test case was: {'content': 'testfile\n', 'name': 'testfile.txt', 'type': 'text/plain'} When I reverted this code, I could attach a file successfully in the local instance. I will continue to investigate but I am posting this just to see if the problem is obvious to some one else from this description. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 12 18:18:27 2010 From: metatracker at psf.upfronthosting.co.za (Raghuram Devarakonda) Date: Fri, 12 Mar 2010 17:18:27 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <1265297768.52.0.920463244949.issue316@psf.upfronthosting.co.za> Message-ID: <1268414307.17.0.650924590231.issue316@psf.upfronthosting.co.za> Raghuram Devarakonda added the comment: I am attaching a patch to schema.py that fixed the problem in local jython instance. I found the fix by comparing with python-dev/schema.py. Note that there are more changes in python-dev version that are missing from jython's schema.py and I am not sure if any of them are required as well. Some of the missing changes are related to OpenID which are obviously not needed. _______________________________________________________ PSF Meta Tracker _______________________________________________________ -------------- next part -------------- A non-text attachment was scrubbed... Name: edit_permission.diff Type: text/x-patch Size: 636 bytes Desc: not available URL: From metatracker at psf.upfronthosting.co.za Fri Mar 12 21:34:12 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Fri, 12 Mar 2010 20:34:12 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <2c51ecee1003120726k4a1b6c69o8fd93e0b3fad3252@mail.gmail.com> Message-ID: <4B9AA541.30300@v.loewis.de> Martin v. L?wis added the comment: > + # Check Edit permission for each property, to avoid being able > + # to set restricted ones on new item creation > + for key in props: > + if not self.hasPermission('Edit', classname=classname, > + property=key): > + # We restrict by default and special-case allowed properties > + if key == 'date' or key == 'content': > + continue > + elif key == 'author' and props[key] == self.userid: > + continue > + return 0 > + return 1 I've traced that further, and found that the roundup authors have determined this change to be a bug, see the current svn for a resolution. I checked to see whether upgrading to 1.4.13 would be straight-forward. Alas, it wouldn't, so I'm going with your proposed change to the schema instead. I'll look into 1.4.13 when I have time, although a patch for that would be welcome, as well. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From martin at v.loewis.de Fri Mar 12 21:34:09 2010 From: martin at v.loewis.de (=?UTF-8?B?Ik1hcnRpbiB2LiBMw7Z3aXMi?=) Date: Fri, 12 Mar 2010 21:34:09 +0100 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <2c51ecee1003120726k4a1b6c69o8fd93e0b3fad3252@mail.gmail.com> References: <2c51ecee1003120726k4a1b6c69o8fd93e0b3fad3252@mail.gmail.com> Message-ID: <4B9AA541.30300@v.loewis.de> > + # Check Edit permission for each property, to avoid being able > + # to set restricted ones on new item creation > + for key in props: > + if not self.hasPermission('Edit', classname=classname, > + property=key): > + # We restrict by default and special-case allowed properties > + if key == 'date' or key == 'content': > + continue > + elif key == 'author' and props[key] == self.userid: > + continue > + return 0 > + return 1 I've traced that further, and found that the roundup authors have determined this change to be a bug, see the current svn for a resolution. I checked to see whether upgrading to 1.4.13 would be straight-forward. Alas, it wouldn't, so I'm going with your proposed change to the schema instead. I'll look into 1.4.13 when I have time, although a patch for that would be welcome, as well. From metatracker at psf.upfronthosting.co.za Fri Mar 12 21:43:01 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Fri, 12 Mar 2010 20:43:01 +0000 Subject: [Tracker-discuss] [issue316] Users can't attach files on Jython tracker In-Reply-To: <1265297768.52.0.920463244949.issue316@psf.upfronthosting.co.za> Message-ID: <1268426581.03.0.143452966392.issue316@psf.upfronthosting.co.za> Martin v. L?wis added the comment: Thanks for the patch. Committed as r78879. Raghuram, please contact me if you need another account in the jython tracker that has only User permissions (and don't want to register a second email address). _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 17 14:51:00 2010 From: metatracker at psf.upfronthosting.co.za (R David Murray) Date: Wed, 17 Mar 2010 13:51:00 +0000 Subject: [Tracker-discuss] [issue320] Missing emails from python-bugs In-Reply-To: <1266859787.97.0.670393025456.issue320@psf.upfronthosting.co.za> Message-ID: <1268833860.4.0.0454385295538.issue320@psf.upfronthosting.co.za> R David Murray added the comment: As far as I can tell this was fixed by the spambayes adjustments at the mail server end. ---------- status: chatting -> resolved _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Tue Mar 23 12:17:10 2010 From: metatracker at psf.upfronthosting.co.za (techtonik) Date: Tue, 23 Mar 2010 11:17:10 +0000 Subject: [Tracker-discuss] [issue324] Add repository info to tracker components In-Reply-To: <1269343030.93.0.79046519877.issue324@psf.upfronthosting.co.za> Message-ID: <1269343030.93.0.79046519877.issue324@psf.upfronthosting.co.za> New submission from techtonik : It would be extremely convenient to have link to source repository when you select component to know where to start to work on a patch. ---------- messages: 1585 nosy: techtonik priority: feature status: unread title: Add repository info to tracker components _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Tue Mar 23 13:47:42 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?b?5bCW5bCW6KeS?=) Date: Tue, 23 Mar 2010 12:47:42 +0000 Subject: [Tracker-discuss] [issue325] Exception KeyError: KeyError(182894117920, ) In-Reply-To: <1269348462.8.0.27510823572.issue325@psf.upfronthosting.co.za> Message-ID: <1269348462.8.0.27510823572.issue325@psf.upfronthosting.co.za> New submission from ??? : hi all: when I'm freezing a python script file into a executable file with the cx_Freeze tool(no missing modules during the freezing process), A problem goes that 'Exception KeyError: KeyError(182894117920,) in '> ignored', and the python script file only includes one sentence "import logging", ps: cxfreeze version is: 4.1.2(http://cx-freeze.sourceforge.net/) python version is: 2.6.5 ---------- messages: 1586 nosy: ??? priority: bug status: unread title: Exception KeyError: KeyError(182894117920,) _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Tue Mar 23 14:59:34 2010 From: metatracker at psf.upfronthosting.co.za (R David Murray) Date: Tue, 23 Mar 2010 13:59:34 +0000 Subject: [Tracker-discuss] [issue325] Exception KeyError: KeyError(182894117920, ) In-Reply-To: <1269348462.8.0.27510823572.issue325@psf.upfronthosting.co.za> Message-ID: <1269352774.39.0.0144121218142.issue325@psf.upfronthosting.co.za> R David Murray added the comment: This is the tracker for issues with the python.org trackers. You need to submit your question to a help channel for freeze. If it is there determined that there is a python bug involved, that would get reported to the python issue tracer at http://bugs.python.org. ---------- nosy: +r.david.murray status: unread -> resolved _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Tue Mar 23 23:06:10 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Tue, 23 Mar 2010 22:06:10 +0000 Subject: [Tracker-discuss] [issue324] Add repository info to tracker components In-Reply-To: <1269343030.93.0.79046519877.issue324@psf.upfronthosting.co.za> Message-ID: <1269381970.54.0.0264459586011.issue324@psf.upfronthosting.co.za> Martin v. L?wis added the comment: Can you please elaborate? I can't picture what you are asking for. Where do you want the link exactly be located, what should the link label be, and what the link target? ---------- nosy: +loewis status: unread -> chatting _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 24 09:41:40 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Wed, 24 Mar 2010 08:41:40 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> Message-ID: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> New submission from Martin v. L?wis : It seems that the postgres database is partially corrupted. For roundup_tracker, doing select count(*) from _file where id='2' gives ERROR: invalid page header in block 0 of relation "_file" Likewise, pg_dumpall produces the same error, as does "vacuum". Any help appreciated. ---------- assignedto: izak messages: 1589 nosy: izak, loewis, roche priority: critical status: unread title: Postgres error: invalid page header ... _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 24 23:28:25 2010 From: metatracker at psf.upfronthosting.co.za (R David Murray) Date: Wed, 24 Mar 2010 22:28:25 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> Message-ID: <1269469705.51.0.217882381344.issue326@psf.upfronthosting.co.za> R David Murray added the comment: I've done some research on this. This kind of error appears, as far as the postgresql mailing list is concerned, to be associated with hardware or kernel problems. That is, something on the system corrupted the database file. At least, I haven't found any indication that this kind of error message has ever been tracked down to anything else. Perhaps someone with postgresql internals knowledge could make something out of the data in the corrupted file, though. Is it something I could share with others (ie: can the information be considered sensitive in any way? My thought is no)? Assuming someone from the mailing list would be willing to look. In addition to this, however, we apparently also have missing pg_clog files. These are deleted by vacuum and autovacuum when they are no longer needed, apparently, but currently there is only one, that was created on the 24th...so at some point clog files were deleted that were still needed. A post I found on the mailing list opines that this should only happen if the files are deleted by something outside postgresql. Do we have a backup of that table somewhere? We may be stuck with restoring it from backup. We could also try (after making a full backup of the binary db files!) restarting postgres with zero_damaged_pages=1, which will get past the error but result in data loss, which could perhaps be patched to be less data loss working from a backup. Since this is page 0, maybe all the data lost would be in the backup. Conceivably I could try that on a test tracker install first, but I haven't set one up yet. It may be possible to finesse the missing transaction data too, but again there may be data loss. I'm concluding all this based on google-foo, so if someone with better postgresql chops can look at this that would of course be better. And however we recover, I hope we can figure out what caused the corruption to begin with, otherwise it may just reoccur. ---------- nosy: +r.david.murray status: unread -> chatting _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 24 23:36:16 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Wed, 24 Mar 2010 22:36:16 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269469705.51.0.217882381344.issue326@psf.upfronthosting.co.za> Message-ID: <4BAA93DE.7060902@v.loewis.de> Martin v. L?wis added the comment: > Do we have a backup of that table somewhere? We may be stuck with > restoring it from backup. We don't have a postgres dump. However, there are backups, in ~roundup/backup (and probably to-tape backups of that from earlier dates; I would also hope that there are backups of the raw database tables). > And however we recover, I hope we can figure out what caused the > corruption to begin with, otherwise it may just reoccur. That would be for Upfront to answer. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Thu Mar 25 16:28:06 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Roch=C3=A9_Compaan?=) Date: Thu, 25 Mar 2010 15:28:06 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <4BAA93DE.7060902@v.loewis.de> Message-ID: <1269530868.2548.383.camel@didi> Roch? Compaan added the comment: On Wed, 2010-03-24 at 22:36 +0000, Martin v. L?wis wrote: > Martin v. L?wis added the comment: > > > Do we have a backup of that table somewhere? We may be stuck with > > restoring it from backup. > > We don't have a postgres dump. However, there are backups, in > ~roundup/backup (and probably to-tape backups of that from earlier > dates; I would also hope that there are backups of the raw database tables). > > > And however we recover, I hope we can figure out what caused the > > corruption to begin with, otherwise it may just reoccur. > > That would be for Upfront to answer. I have asked Izak to have a look. Apologies for making you wait. I'm currently on holiday and not checking my mail frequently and our whole team was out of the office yesterday. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Thu Mar 25 16:47:07 2010 From: metatracker at psf.upfronthosting.co.za (Izak Burger) Date: Thu, 25 Mar 2010 15:47:07 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <4BAA93DE.7060902@v.loewis.de> Message-ID: <4BAB8573.1040408@upfrontsystems.co.za> Izak Burger added the comment: Martin v. L?wis wrote: >> And however we recover, I hope we can figure out what caused the >> corruption to begin with, otherwise it may just reoccur. > > That would be for Upfront to answer. We didn't change the kernel or the hardware. We did upgrade the virtual host on which psf runs, including the postgres installation. I suspect we need only shut it down, run a fsck on the filesystem to make sure that is in good shape. Let me know when would be a good time. I suppose any time and as soon as possible? Unfortunately I am not an expert in postgresql on THAT low a level, so I would also recommend a restore from backup. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 00:34:08 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Thu, 25 Mar 2010 23:34:08 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <4BAB8573.1040408@upfrontsystems.co.za> Message-ID: <4BABF2EE.6000304@v.loewis.de> Martin v. L?wis added the comment: > We didn't change the kernel or the hardware. We did upgrade the virtual > host on which psf runs, including the postgres installation. I suspect > we need only shut it down, run a fsck on the filesystem to make sure > that is in good shape. Let me know when would be a good time. I suppose > any time and as soon as possible? Sure, go ahead. > Unfortunately I am not an expert in postgresql on THAT low a level, so I > would also recommend a restore from backup. I'd rather avoid that. The backup is not current enough, and new backups aren't being made because of the corruption. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 07:46:55 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Fri, 26 Mar 2010 06:46:55 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> Message-ID: <1269586015.3.0.299610543092.issue326@psf.upfronthosting.co.za> Martin v. L?wis added the comment: I have now restarted postgres; that alone seems to have resolved some of the problems. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 08:35:12 2010 From: metatracker at psf.upfronthosting.co.za (Izak Burger) Date: Fri, 26 Mar 2010 07:35:12 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269586015.3.0.299610543092.issue326@psf.upfronthosting.co.za> Message-ID: <4BAC65AD.9010907@upfrontsystems.co.za> Izak Burger added the comment: I've checked the filesystem, and though it wasn't checked in 420 days, it was still fine. I cleaned up some of the startup scripts that do nothing on a virtual host (things that set the hardware clock etc) and then restarted it. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 16:43:32 2010 From: metatracker at psf.upfronthosting.co.za (R David Murray) Date: Fri, 26 Mar 2010 15:43:32 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> Message-ID: <1269618212.08.0.890858557738.issue326@psf.upfronthosting.co.za> R David Murray added the comment: If someone can restore one of the tape backups of file.csv to somewhere on the machine (the one in ~roundup/backup is now empty), I will try to make time this weekend to experiment with ways to recover that table. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 18:33:07 2010 From: metatracker at psf.upfronthosting.co.za (Ezio Melotti) Date: Fri, 26 Mar 2010 17:33:07 +0000 Subject: [Tracker-discuss] [issue265] Search in the top search box broken for empty input In-Reply-To: <1239148841.99.0.456266039624.issue265@psf.upfronthosting.co.za> Message-ID: <1269624787.84.0.575146073853.issue265@psf.upfronthosting.co.za> Ezio Melotti added the comment: This is now fixed. ---------- nosy: +ezio.melotti, loewis status: unread -> resolved _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 18:42:44 2010 From: metatracker at psf.upfronthosting.co.za (Ezio Melotti) Date: Fri, 26 Mar 2010 17:42:44 +0000 Subject: [Tracker-discuss] [issue309] Problems with wrap="hard" and konqueror In-Reply-To: <1262459459.14.0.698323309061.issue309@psf.upfronthosting.co.za> Message-ID: <1269625364.56.0.70273962173.issue309@psf.upfronthosting.co.za> Ezio Melotti added the comment: The messages seem fine even without CSS wrapping, so I'm closing the issue. This also fixed #277. ---------- status: chatting -> resolved _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 18:44:57 2010 From: metatracker at psf.upfronthosting.co.za (Ezio Melotti) Date: Fri, 26 Mar 2010 17:44:57 +0000 Subject: [Tracker-discuss] [issue277] Line wrapping in comment box is annoying In-Reply-To: <1240785838.71.0.559804829512.issue277@psf.upfronthosting.co.za> Message-ID: <1269625497.23.0.254689885055.issue277@psf.upfronthosting.co.za> Ezio Melotti added the comment: This should be fixed now (see #309). Kurt, if you are still having problems you can re-open the issue. ---------- nosy: +ezio.melotti status: chatting -> resolved _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 19:04:29 2010 From: metatracker at psf.upfronthosting.co.za (Ezio Melotti) Date: Fri, 26 Mar 2010 18:04:29 +0000 Subject: [Tracker-discuss] [issue308] xmlrpc server does not seem to be using allow_none=True In-Reply-To: <1261282344.81.0.701981844474.issue308@psf.upfronthosting.co.za> Message-ID: <1269626669.07.0.190763718664.issue308@psf.upfronthosting.co.za> Ezio Melotti added the comment: Now it should have 2.5. ---------- nosy: +ezio.melotti _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 19:53:51 2010 From: metatracker at psf.upfronthosting.co.za (Daniel Diniz) Date: Fri, 26 Mar 2010 18:53:51 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <4BABF2EE.6000304@v.loewis.de> Message-ID: <2d75d7661003261153g3eb2508k82f008b30fbb0fe2@mail.gmail.com> Daniel Diniz added the comment: Martin v. L?wis wrote: >> Unfortunately I am not an expert in postgresql on THAT low a level, so I >> would also recommend a restore from backup. > > I'd rather avoid that. The backup is not current enough, and new backups > aren't being made because of the corruption. Martin, are we positive there's missing data or is it possible that we just can't access the table? I've found a message[1] suggesting we can still get the data if we skip accessing the header on _file. If you issue "SET enable_seqscan TO 'off'; ", does "select * from _file where __retired__=0" work? And does the dump work? HTH, Daniel [1] http://osdir.com/ml/pgsql-general/2010-02/msg00380.html ---------- nosy: +ajaksu2 _______________________________________________________ PSF Meta Tracker _______________________________________________________ From ajaksu at gmail.com Fri Mar 26 19:53:29 2010 From: ajaksu at gmail.com (Daniel Diniz) Date: Fri, 26 Mar 2010 15:53:29 -0300 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <4BABF2EE.6000304@v.loewis.de> References: <4BAB8573.1040408@upfrontsystems.co.za> <4BABF2EE.6000304@v.loewis.de> Message-ID: <2d75d7661003261153g3eb2508k82f008b30fbb0fe2@mail.gmail.com> Martin v. L?wis wrote: >> Unfortunately I am not an expert in postgresql on THAT low a level, so I >> would also recommend a restore from backup. > > I'd rather avoid that. The backup is not current enough, and new backups > aren't being made because of the corruption. Martin, are we positive there's missing data or is it possible that we just can't access the table? I've found a message[1] suggesting we can still get the data if we skip accessing the header on _file. If you issue "SET enable_seqscan TO 'off'; ", does "select * from _file where __retired__=0" work? And does the dump work? HTH, Daniel [1] http://osdir.com/ml/pgsql-general/2010-02/msg00380.html From metatracker at psf.upfronthosting.co.za Fri Mar 26 20:22:46 2010 From: metatracker at psf.upfronthosting.co.za (R David Murray) Date: Fri, 26 Mar 2010 19:22:46 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> Message-ID: <1269631366.24.0.443181392684.issue326@psf.upfronthosting.co.za> R David Murray added the comment: Unfortunately, no: roundup_tracker=# set enable_seqscan to 'off'; SET roundup_tracker=# select * from _file where __retired__=0; ERROR: invalid page header in block 0 of relation "_file" I'm hoping that if there is any missing data it is old data that we have a backup of (somewhere), and that the rest will be readable after zeroing the invalid page. But I'm not going to test that theory on the live database, so I need to set up a test instance in order to test my theory... _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 20:33:23 2010 From: metatracker at psf.upfronthosting.co.za (R David Murray) Date: Fri, 26 Mar 2010 19:33:23 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> Message-ID: <1269632003.77.0.0643774700037.issue326@psf.upfronthosting.co.za> R David Murray added the comment: Hmm. What I just reported makes me wonder, though. Accessing the data through that index should have worked :( Using file_retired=1 does not produce the header error, but gives an empty table. (I don't know what values are valid). Maybe retired=0 is trying to read an item from the first database page, and thus gets the error. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From ajaksu at gmail.com Fri Mar 26 20:47:41 2010 From: ajaksu at gmail.com (Daniel Diniz) Date: Fri, 26 Mar 2010 16:47:41 -0300 Subject: [Tracker-discuss] [issue323] email is not obfuscated in the msg In-Reply-To: <1267435647.66.0.658210680138.issue323@psf.upfronthosting.co.za> References: <1267435647.66.0.658210680138.issue323@psf.upfronthosting.co.za> <1267435647.66.0.658210680138.issue323@psf.upfronthosting.co.za> Message-ID: <2d75d7661003261247w4c1f22b4v89ef8b4f537ebd39@mail.gmail.com> Leaving the disregard for some basic netiquette aside, we could enforce anonymity by either: 1. sending notification emails that omit the actor's email (one liner), or 2. obscuring emails in tracker messages (either by processing on display or sanitizing on store). While 1 is easier to implement and maintain (and would protect emails in the list[1]), I think it hinders communication (i.e., how do I reply off-list?). I don't like 2 either, I'd rather sanitize on store if we absolutely had to get this done but this would leave the emails in the open on the lists. I think leaving things as they stand is an option. Most of the tracker users will have their email shown in other Python mailing lists anyway, in an easier to parse format. [1] http://mail.python.org/pipermail/tracker-discuss/2010-March/002364.html From metatracker at psf.upfronthosting.co.za Fri Mar 26 20:48:02 2010 From: metatracker at psf.upfronthosting.co.za (Daniel Diniz) Date: Fri, 26 Mar 2010 19:48:02 +0000 Subject: [Tracker-discuss] [issue323] email is not obfuscated in the msg In-Reply-To: <1267435647.66.0.658210680138.issue323@psf.upfronthosting.co.za> Message-ID: <2d75d7661003261247w4c1f22b4v89ef8b4f537ebd39@mail.gmail.com> Daniel Diniz added the comment: Leaving the disregard for some basic netiquette aside, we could enforce anonymity by either: 1. sending notification emails that omit the actor's email (one liner), or 2. obscuring emails in tracker messages (either by processing on display or sanitizing on store). While 1 is easier to implement and maintain (and would protect emails in the list[1]), I think it hinders communication (i.e., how do I reply off-list?). I don't like 2 either, I'd rather sanitize on store if we absolutely had to get this done but this would leave the emails in the open on the lists. I think leaving things as they stand is an option. Most of the tracker users will have their email shown in other Python mailing lists anyway, in an easier to parse format. [1] http://mail.python.org/pipermail/tracker-discuss/2010-March/002364.html ---------- nosy: +ajaksu2 status: unread -> chatting _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 21:07:55 2010 From: metatracker at psf.upfronthosting.co.za (Daniel Diniz) Date: Fri, 26 Mar 2010 20:07:55 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269632003.77.0.0643774700037.issue326@psf.upfronthosting.co.za> Message-ID: <2d75d7661003261307k54c5a0cdga9cb2d717d95d3a3@mail.gmail.com> Daniel Diniz added the comment: David, What does "EXPLAIN select * from _file where __retired__=0;" give you? I get "Index Scan using _file_retired_idx on _file ..." after setting enable_seqscan to off, but "Seq Scan on _file ..." in the default case. IIRC, __retired__ is an integer used as a bool, defaulting to 0 (file is 'alive'). You could also try to filter by id to avoid hitting page 0, something like this seems to perform an index scan too: select * from _file where __retired__=0 and id>15000; _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 21:28:57 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Fri, 26 Mar 2010 20:28:57 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <2d75d7661003261307k54c5a0cdga9cb2d717d95d3a3@mail.gmail.com> Message-ID: <4BAD1907.20504@v.loewis.de> Martin v. L?wis added the comment: > What does "EXPLAIN select * from _file where __retired__=0;" give you? > > I get "Index Scan using _file_retired_idx on _file ..." after setting > enable_seqscan to off, but "Seq Scan on _file ..." in the default > case. Same here. > select * from _file where __retired__=0 and id>15000; That does indeed produce a result. Not sure how this is going to help, though. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 21:57:59 2010 From: metatracker at psf.upfronthosting.co.za (Daniel Diniz) Date: Fri, 26 Mar 2010 20:57:59 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <4BAD1907.20504@v.loewis.de> Message-ID: <2d75d7661003261357n124fcde7pe92982ab8ed772d5@mail.gmail.com> Daniel Diniz added the comment: Martin v. L?wis wrote: >> select * from _file where __retired__=0 and id>15000; > > That does indeed produce a result. Not sure how this is going to help, > though. Not sure it will help. It gives us some info on whether we lost data (and if so, where) or just can't access records easily. It seems to be possible to get a (lossy?) dump of the DB under zero_damaged_pages='on', as David proposed. If it makes any sense, this might help getting most of the data from _file before zeroing pages, if only to know how useful an ancient backup for _file would be. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 22:10:20 2010 From: metatracker at psf.upfronthosting.co.za (R David Murray) Date: Fri, 26 Mar 2010 21:10:20 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> Message-ID: <1269637820.42.0.424973999266.issue326@psf.upfronthosting.co.za> R David Murray added the comment: That query works the same whether you use retired or not. Looks like we at least have the current data from id 10336 on (creation date 2008-05-16). If we have a backup of the earlier data we can at least recover the most of the data. We'd still be missing any changes made to older objects, but I don't think file information changes very often. We could also ask on the postgres forum to see if anyone is knowledgeable enough to actually repair the page for us. It's possible just some bits got twiddled rather than wholesale corruption of the block, but I don't know how to tell. I won't have time to get subscribed and put together a post until tomorrow sometime. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 22:24:46 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Fri, 26 Mar 2010 21:24:46 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269637820.42.0.424973999266.issue326@psf.upfronthosting.co.za> Message-ID: <4BAD261B.2080505@v.loewis.de> Martin v. L?wis added the comment: > That query works the same whether you use retired or not. Looks like > we at least have the current data from id 10336 on (creation date > 2008-05-16). If we have a backup of the earlier data we can at least > recover the most of the data. We'd still be missing any changes made > to older objects, but I don't think file information changes very > often. I'm tempted to create the lossy dump right away, drop the table, and recreate it. We shouldn't invest too much time into this: the actual file data are on disk, and hopefully Upfront can provide us with some old backup within a few days (certainly not over the weekend, though). As for changes, we also have file__journal, which seems to dump fine. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Fri Mar 26 23:49:55 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Fri, 26 Mar 2010 22:49:55 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> Message-ID: <1269643795.98.0.431609137444.issue326@psf.upfronthosting.co.za> Martin v. L?wis added the comment: I have now created a full database tgz if anybody wants to take a look. I have then enabled the zero_damaged_pages setting, and vacuum'ed. That produced these messages: invalid page header in block 0 of relation "_file"; zeroing out page invalid page header in block 8 of relation "pg_type_typname_nsp_index"; zeroing out page Dumping the database still failed with a missing pg_clog file. I created one that was zeroed out; this let me proceed. Postgres didn't decide to delete the log file afterwards; this worries me somewhat. Dumping the database still doesn't work; it gives this error message: ERROR: invalid memory alloc request size 18446744073709551613 This happens when doing "select *" or "copy" on _file. Notice that this number is 0xfffffffffffffffd. izak: can you please provide a copy of the latest backup of /home/roundup/backups/tracker-export where file.csv is not empty? _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Sat Mar 27 02:30:32 2010 From: metatracker at psf.upfronthosting.co.za (Daniel Diniz) Date: Sat, 27 Mar 2010 01:30:32 +0000 Subject: [Tracker-discuss] [issue321] Quoted sections disappear when issue updates are sent by e-mail In-Reply-To: <1266958137.47.0.48089118165.issue321@psf.upfronthosting.co.za> Message-ID: <1269653432.14.0.438307882589.issue321@psf.upfronthosting.co.za> Daniel Diniz added the comment: I have a patch for this at both issue 264 and the upstream tracker[1], but it needs to be tested with an up-to-date env. [1] http://issues.roundup-tracker.org/issue2550535 ---------- nosy: +ajaksu2 status: unread -> chatting _______________________________________________________ PSF Meta Tracker _______________________________________________________ From ziade.tarek at gmail.com Sun Mar 28 16:44:40 2010 From: ziade.tarek at gmail.com (=?ISO-8859-1?Q?Tarek_Ziad=E9?=) Date: Sun, 28 Mar 2010 16:44:40 +0200 Subject: [Tracker-discuss] A "gsoc" keyword Message-ID: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> Hello I've started to create a serie of issues that could be done during gsoc. Could we add a 'gsoc' keyword in the tracker ? or something similar already exists ? Thanks! Regards Tarek -- Tarek Ziad? | http://ziade.org From brett at python.org Mon Mar 29 02:09:10 2010 From: brett at python.org (Brett Cannon) Date: Sun, 28 Mar 2010 17:09:10 -0700 Subject: [Tracker-discuss] A "gsoc" keyword In-Reply-To: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> References: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> Message-ID: On Sun, Mar 28, 2010 at 07:44, Tarek Ziad? wrote: > Hello > > I've started to create a serie of issues that could be done during gsoc. > Could we add a 'gsoc' keyword in the tracker ? or something similar > already exists ? > > If they are being created specifically for GSoC, couldn't that simply be put in the title? That way we don't need to add another keyword. Maybe others disagree, but we are ending up with a lot of keywords IMO. -Brett > Thanks! > > Regards > Tarek > > -- > Tarek Ziad? | http://ziade.org > _______________________________________________ > Tracker-discuss mailing list > Tracker-discuss at python.org > http://mail.python.org/mailman/listinfo/tracker-discuss > -------------- next part -------------- An HTML attachment was scrubbed... URL: From ajaksu at gmail.com Mon Mar 29 02:20:17 2010 From: ajaksu at gmail.com (Daniel Diniz) Date: Sun, 28 Mar 2010 21:20:17 -0300 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269643795.98.0.431609137444.issue326@psf.upfronthosting.co.za> References: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> <1269643795.98.0.431609137444.issue326@psf.upfronthosting.co.za> Message-ID: <2d75d7661003281720g3751e564u9ea86d71713a657d@mail.gmail.com> > Dumping the database still doesn't work; it gives this error message: > > ERROR: ?invalid memory alloc request size 18446744073709551613 > > This happens when doing "select *" or "copy" on _file. Notice that this number is 0xfffffffffffffffd. Ugh. It seems we're left with copy-ing the other tables to files, salvaging _file and rebuilding it with a backup and/or file_journal. I could try to roll a script to perform these tasks if you're interested. From metatracker at psf.upfronthosting.co.za Mon Mar 29 02:20:39 2010 From: metatracker at psf.upfronthosting.co.za (Daniel Diniz) Date: Mon, 29 Mar 2010 00:20:39 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269643795.98.0.431609137444.issue326@psf.upfronthosting.co.za> Message-ID: <2d75d7661003281720g3751e564u9ea86d71713a657d@mail.gmail.com> Daniel Diniz added the comment: > Dumping the database still doesn't work; it gives this error message: > > ERROR: ?invalid memory alloc request size 18446744073709551613 > > This happens when doing "select *" or "copy" on _file. Notice that this number is 0xfffffffffffffffd. Ugh. It seems we're left with copy-ing the other tables to files, salvaging _file and rebuilding it with a backup and/or file_journal. I could try to roll a script to perform these tasks if you're interested. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From ziade.tarek at gmail.com Mon Mar 29 08:18:23 2010 From: ziade.tarek at gmail.com (=?ISO-8859-1?Q?Tarek_Ziad=E9?=) Date: Mon, 29 Mar 2010 08:18:23 +0200 Subject: [Tracker-discuss] A "gsoc" keyword In-Reply-To: References: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> Message-ID: <94bdd2611003282318w3af9a7c6xc9e29f9ba9438be4@mail.gmail.com> On Mon, Mar 29, 2010 at 2:09 AM, Brett Cannon wrote: > On Sun, Mar 28, 2010 at 07:44, Tarek Ziad? wrote: >> >> Hello >> >> I've started to create a serie of issues that could be done during gsoc. >> Could we add a 'gsoc' keyword in the tracker ? or something similar >> already exists ? >> > > If they are being created specifically for GSoC, couldn't that simply be put > in the title? Some may be added or removed from the GSOC list over time. > That way we don't need to add another keyword. Maybe others > disagree, but we are ending up with a lot of keywords IMO. No worries, I didn't know having more keywords was an issue. I guess I could simply add the gsoc word in the text, and do the right queries Regards Tarek -- Tarek Ziad? | http://ziade.org From metatracker at psf.upfronthosting.co.za Mon Mar 29 16:13:48 2010 From: metatracker at psf.upfronthosting.co.za (R David Murray) Date: Mon, 29 Mar 2010 14:13:48 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> Message-ID: <1269872028.45.0.259909824831.issue326@psf.upfronthosting.co.za> R David Murray added the comment: I think any help will be welcome, Daniel :) It turns out I can't even play around with this because I don't have a 64bit linux on which to load the binary copy of the database. I should be able to set up a KVM instance (which I want to do anyway), but that will take me a bit. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From techtonik at gmail.com Mon Mar 29 18:41:53 2010 From: techtonik at gmail.com (anatoly techtonik) Date: Mon, 29 Mar 2010 19:41:53 +0300 Subject: [Tracker-discuss] A "gsoc" keyword In-Reply-To: <94bdd2611003282318w3af9a7c6xc9e29f9ba9438be4@mail.gmail.com> References: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> <94bdd2611003282318w3af9a7c6xc9e29f9ba9438be4@mail.gmail.com> Message-ID: I would say that "keywords" concept is flawed in Python tracker. Usually tracker users are free to add whatever keywords they need. In Python tracker keywords are: 1. limited to devs only 2. limited to a rigid list So, speaking about keywords, I would rename them to devtags and introduced real space separated keywords concept. -- anatoly t. On Mon, Mar 29, 2010 at 9:18 AM, Tarek Ziad? wrote: > On Mon, Mar 29, 2010 at 2:09 AM, Brett Cannon wrote: >> On Sun, Mar 28, 2010 at 07:44, Tarek Ziad? wrote: >>> >>> Hello >>> >>> I've started to create a serie of issues that could be done during gsoc. >>> Could we add a 'gsoc' keyword in the tracker ? or something similar >>> already exists ? >>> >> >> If they are being created specifically for GSoC, couldn't that simply be put >> in the title? > > Some may be added or removed from the GSOC list over time. > >> That way we don't need to add another keyword. Maybe others >> disagree, but we are ending up with a lot of keywords IMO. > > No worries, I didn't know having more keywords was an issue. > > I guess I could simply add the gsoc word in the text, and do the right queries > > Regards > Tarek > > -- > Tarek Ziad? | http://ziade.org > _______________________________________________ > Tracker-discuss mailing list > Tracker-discuss at python.org > http://mail.python.org/mailman/listinfo/tracker-discuss > From brett at python.org Mon Mar 29 20:32:49 2010 From: brett at python.org (Brett Cannon) Date: Mon, 29 Mar 2010 11:32:49 -0700 Subject: [Tracker-discuss] A "gsoc" keyword In-Reply-To: References: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> <94bdd2611003282318w3af9a7c6xc9e29f9ba9438be4@mail.gmail.com> Message-ID: Well, it's something I don't want to overdue, but others might disagree. I'm only a -1 on adding the keyword, but then again maybe it would help identify more viable projects throughout the year. Anyone else have an opinion? On Mar 28, 2010 11:18 PM, "Tarek Ziad?" wrote: On Mon, Mar 29, 2010 at 2:09 AM, Brett Cannon wrote: > On Sun, Mar 28, 2010 at 07... Some may be added or removed from the GSOC list over time. > That way we don't need to add another keyword. Maybe others > disagree, but we are ending up with... No worries, I didn't know having more keywords was an issue. I guess I could simply add the gsoc word in the text, and do the right queries Regards Tarek -- Tarek Ziad? | http://ziade.org -------------- next part -------------- An HTML attachment was scrubbed... URL: From brett at python.org Mon Mar 29 21:14:04 2010 From: brett at python.org (Brett Cannon) Date: Mon, 29 Mar 2010 12:14:04 -0700 Subject: [Tracker-discuss] A "gsoc" keyword In-Reply-To: References: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> <94bdd2611003282318w3af9a7c6xc9e29f9ba9438be4@mail.gmail.com> Message-ID: On Mon, Mar 29, 2010 at 11:32, Brett Cannon wrote: > Well, it's something I don't want to overdue, but others might disagree. > I'm only a -1 on adding the keyword, but then again maybe it would help > identify more viable projects throughout the year. Anyone else have an > opinion? > Sorry, that should have said -0; I was on my phone. -Brett > On Mar 28, 2010 11:18 PM, "Tarek Ziad?" wrote: > > On Mon, Mar 29, 2010 at 2:09 AM, Brett Cannon wrote: > > On Sun, Mar 28, 2010 at 07... > > Some may be added or removed from the GSOC list over time. > > > > That way we don't need to add another keyword. Maybe others > > disagree, but we are ending up with... > > No worries, I didn't know having more keywords was an issue. > > I guess I could simply add the gsoc word in the text, and do the right > queries > > > Regards > Tarek > > -- > Tarek Ziad? | http://ziade.org > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From ziade.tarek at gmail.com Mon Mar 29 21:30:38 2010 From: ziade.tarek at gmail.com (=?ISO-8859-1?Q?Tarek_Ziad=E9?=) Date: Mon, 29 Mar 2010 21:30:38 +0200 Subject: [Tracker-discuss] A "gsoc" keyword In-Reply-To: References: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> <94bdd2611003282318w3af9a7c6xc9e29f9ba9438be4@mail.gmail.com> Message-ID: <94bdd2611003291230h53593ca5l5abf96ab5c5a5148@mail.gmail.com> On Mon, Mar 29, 2010 at 9:14 PM, Brett Cannon wrote: > > > On Mon, Mar 29, 2010 at 11:32, Brett Cannon wrote: >> >> Well, it's something I don't want to overdue, but others might disagree. >> I'm only a -1 on adding the keyword, but then again maybe it would help >> identify more viable projects throughout the year. Anyone else have an >> opinion? > > Sorry, that should have said -0; I was on my phone. I can work without that keyword, it's no big deal really :) Tarek -- Tarek Ziad? | http://ziade.org From metatracker at psf.upfronthosting.co.za Mon Mar 29 22:25:27 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Mon, 29 Mar 2010 20:25:27 +0000 Subject: [Tracker-discuss] [issue326] Postgres error: invalid page header ... In-Reply-To: <1269420100.81.0.0570912412301.issue326@psf.upfronthosting.co.za> Message-ID: <1269894327.72.0.217668972291.issue326@psf.upfronthosting.co.za> Martin v. L?wis added the comment: I'm going to try to write a script that selects each individual _file record, dumps them all into a pickle, so that I can drop the table afterwards. I hope that this will then allow to dump the database. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From martin at v.loewis.de Mon Mar 29 22:35:50 2010 From: martin at v.loewis.de (=?UTF-8?B?Ik1hcnRpbiB2LiBMw7Z3aXMi?=) Date: Mon, 29 Mar 2010 22:35:50 +0200 Subject: [Tracker-discuss] A "gsoc" keyword In-Reply-To: References: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> <94bdd2611003282318w3af9a7c6xc9e29f9ba9438be4@mail.gmail.com> Message-ID: <4BB10F26.2060900@v.loewis.de> Brett Cannon wrote: > Well, it's something I don't want to overdue, but others might disagree. > I'm only a -1 on adding the keyword, but then again maybe it would help > identify more viable projects throughout the year. Anyone else have an > opinion? I wouldn't mind adding it. We can review the list of keywords from time to time and discard the least frequently used one. Regards, Martin From ezio.melotti at gmail.com Mon Mar 29 23:25:41 2010 From: ezio.melotti at gmail.com (Ezio Melotti) Date: Tue, 30 Mar 2010 00:25:41 +0300 Subject: [Tracker-discuss] A "gsoc" keyword In-Reply-To: <4BB10F26.2060900@v.loewis.de> References: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> <94bdd2611003282318w3af9a7c6xc9e29f9ba9438be4@mail.gmail.com> <4BB10F26.2060900@v.loewis.de> Message-ID: <4BB11AD5.6000408@gmail.com> On 29/03/2010 23.35, "Martin v. L?wis" wrote: > Brett Cannon wrote: > >> Well, it's something I don't want to overdue, but others might disagree. >> I'm only a -1 on adding the keyword, but then again maybe it would help >> identify more viable projects throughout the year. Anyone else have an >> opinion? >> > I wouldn't mind adding it. We can review the list of keywords from time > to time and discard the least frequently used one. > > Regards, > Martin > _______________________________________________ > Tracker-discuss mailing list > Tracker-discuss at python.org > http://mail.python.org/mailman/listinfo/tracker-discuss > > See also http://psf.upfronthosting.co.za/roundup/meta/issue253 about "outdated" keywords. From facundobatista at gmail.com Wed Mar 31 01:12:42 2010 From: facundobatista at gmail.com (Facundo Batista) Date: Tue, 30 Mar 2010 19:12:42 -0400 Subject: [Tracker-discuss] Problem with issue 1251 Message-ID: If you go to... http://bugs.python.org/issue1251 ...you get... """ An error has occurred A problem was encountered processing your request. The tracker maintainers have been notified of the problem. """ Regards, -- . Facundo Blog: http://www.taniquetil.com.ar/plog/ PyAr: http://www.python.org/ar/ From rdmurray at bitdance.com Wed Mar 31 04:20:22 2010 From: rdmurray at bitdance.com (R. David Murray) Date: Tue, 30 Mar 2010 22:20:22 -0400 Subject: [Tracker-discuss] Problem with issue 1251 In-Reply-To: References: Message-ID: <20100331022022.8E2671FBF0D@kimball.webabinitio.net> On Tue, 30 Mar 2010 19:12:42 -0400, Facundo Batista wrote: > http://bugs.python.org/issue1251 > > ...you get... > > """ > An error has occurred > > A problem was encountered processing your request. The tracker > maintainers have been notified of the problem. > """ Most likely this is a manifestation of the _file table corruption problem that Martin is working on. -- R. David Murray www.bitdance.com From metatracker at psf.upfronthosting.co.za Wed Mar 31 07:56:57 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Wed, 31 Mar 2010 05:56:57 +0000 Subject: [Tracker-discuss] [issue327] Please provide backups In-Reply-To: <1270015017.64.0.286941434615.issue327@psf.upfronthosting.co.za> Message-ID: <1270015017.64.0.286941434615.issue327@psf.upfronthosting.co.za> New submission from Martin v. L?wis : Izak, can you please provide some sort of backup of the PSF tracker data from before the postgres corruption? For the moment, an old version of file.csv of the "tracker" instance would be sufficient. ---------- assignedto: izak messages: 1616 nosy: izak, loewis priority: critical status: unread title: Please provide backups _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 31 08:13:16 2010 From: metatracker at psf.upfronthosting.co.za (techtonik) Date: Wed, 31 Mar 2010 06:13:16 +0000 Subject: [Tracker-discuss] [issue328] More intuitive "stage" and "resolution" In-Reply-To: <1270015996.05.0.585266502058.issue328@psf.upfronthosting.co.za> Message-ID: <1270015996.05.0.585266502058.issue328@psf.upfronthosting.co.za> New submission from techtonik : Right now we have separate "stage" and "resolution" fields. The former is useless when a bug is closed, the latter has no value when the bug is worked on. I propose to merge them to reduce clutter and amount of irrelevant bug information. This is similar to the way Trac works. ---------- messages: 1617 nosy: techtonik priority: feature status: unread title: More intuitive "stage" and "resolution" _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 31 08:14:59 2010 From: metatracker at psf.upfronthosting.co.za (techtonik) Date: Wed, 31 Mar 2010 06:14:59 +0000 Subject: [Tracker-discuss] [issue329] "Your Issues" list is empty In-Reply-To: <1270016099.97.0.990047736991.issue329@psf.upfronthosting.co.za> Message-ID: <1270016099.97.0.990047736991.issue329@psf.upfronthosting.co.za> New submission from techtonik : My "Your Issues" list is empty, although I have opened tickets where I am the author. I.e. http://bugs.python.org/issue7585 ---------- messages: 1618 nosy: techtonik priority: bug status: unread title: "Your Issues" list is empty _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 31 08:19:13 2010 From: metatracker at psf.upfronthosting.co.za (techtonik) Date: Wed, 31 Mar 2010 06:19:13 +0000 Subject: [Tracker-discuss] [issue330] Users should be able to set bug dependencies In-Reply-To: <1270016353.69.0.59719876843.issue330@psf.upfronthosting.co.za> Message-ID: <1270016353.69.0.59719876843.issue330@psf.upfronthosting.co.za> New submission from techtonik : Users are not able to work with bugs in tracker and can't help with triaging bugs. Core developers waste valuable time reading doing this. As they are the busies people here, they do not have much time, therefore tracker items often do not have correct dependencies, and doesn't allow core developers to see the bottlenecks. ---------- messages: 1619 nosy: techtonik priority: bug status: unread title: Users should be able to set bug dependencies _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 31 08:24:58 2010 From: metatracker at psf.upfronthosting.co.za (techtonik) Date: Wed, 31 Mar 2010 06:24:58 +0000 Subject: [Tracker-discuss] [issue253] Remove outdated keywords In-Reply-To: <1237513464.32.0.072891570662.issue253@psf.upfronthosting.co.za> Message-ID: <1270016698.11.0.931232078133.issue253@psf.upfronthosting.co.za> techtonik added the comment: Does anybody have stats for keyword usage? ---------- nosy: +techtonik _______________________________________________________ PSF Meta Tracker _______________________________________________________ From ajaksu at gmail.com Wed Mar 31 08:25:21 2010 From: ajaksu at gmail.com (Daniel Diniz) Date: Wed, 31 Mar 2010 03:25:21 -0300 Subject: [Tracker-discuss] [issue329] "Your Issues" list is empty In-Reply-To: <1270016099.97.0.990047736991.issue329@psf.upfronthosting.co.za> References: <1270016099.97.0.990047736991.issue329@psf.upfronthosting.co.za> <1270016099.97.0.990047736991.issue329@psf.upfronthosting.co.za> Message-ID: As explained in issue 299: '"Your issues" is the list of issues that have been assigned to you[...].' And I think Ezio has the best solution: ' It would even be better to have "Issues opened by you" for both devs and non-devs and "Your Issues" for devs only.' Leaving open because so far it's a much more objective issue when compared to the original ;) From metatracker at psf.upfronthosting.co.za Wed Mar 31 08:25:48 2010 From: metatracker at psf.upfronthosting.co.za (Daniel Diniz) Date: Wed, 31 Mar 2010 06:25:48 +0000 Subject: [Tracker-discuss] [issue329] "Your Issues" list is empty In-Reply-To: <1270016099.97.0.990047736991.issue329@psf.upfronthosting.co.za> Message-ID: Daniel Diniz added the comment: As explained in issue 299: '"Your issues" is the list of issues that have been assigned to you[...].' And I think Ezio has the best solution: ' It would even be better to have "Issues opened by you" for both devs and non-devs and "Your Issues" for devs only.' Leaving open because so far it's a much more objective issue when compared to the original ;) ---------- nosy: +ajaksu2 status: unread -> chatting _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 31 08:31:40 2010 From: metatracker at psf.upfronthosting.co.za (techtonik) Date: Wed, 31 Mar 2010 06:31:40 +0000 Subject: [Tracker-discuss] [issue329] "Your Issues" list is empty In-Reply-To: <1270016099.97.0.990047736991.issue329@psf.upfronthosting.co.za> Message-ID: <1270017100.14.0.684273387671.issue329@psf.upfronthosting.co.za> techtonik added the comment: Having to read issue 299 to figure this out is a little bit unfriendly in the terms of interface. I would expect on this page all of: 1. Clear message "There is no issues assigned to you" 2. A list of issues reported by me 3. A list of issues I participated Of course this will require pagination, but 2 and 3 can be at least links on this page. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From ajaksu at gmail.com Wed Mar 31 08:45:17 2010 From: ajaksu at gmail.com (Daniel Diniz) Date: Wed, 31 Mar 2010 03:45:17 -0300 Subject: [Tracker-discuss] [issue330] Users should be able to set bug dependencies In-Reply-To: <1270016353.69.0.59719876843.issue330@psf.upfronthosting.co.za> References: <1270016353.69.0.59719876843.issue330@psf.upfronthosting.co.za> <1270016353.69.0.59719876843.issue330@psf.upfronthosting.co.za> Message-ID: I like the compromise of giving extra powers to people interested in helping and keeping the UI simple for casual reporters, and dependencies might confuse the later. We do have auto-linking of issues cited in messages, so most users can mention dependencies (or duplicates, etc.) to help. But maybe we should have a role that sits between User and Developer for power users. I don't think dependencies/superseder are reason enough for that, but the principle of allowing people to help more might be. +1 on getting more helping hands -0 on adding Dependencies to the default User role From metatracker at psf.upfronthosting.co.za Wed Mar 31 08:45:38 2010 From: metatracker at psf.upfronthosting.co.za (Daniel Diniz) Date: Wed, 31 Mar 2010 06:45:38 +0000 Subject: [Tracker-discuss] [issue330] Users should be able to set bug dependencies In-Reply-To: <1270016353.69.0.59719876843.issue330@psf.upfronthosting.co.za> Message-ID: Daniel Diniz added the comment: I like the compromise of giving extra powers to people interested in helping and keeping the UI simple for casual reporters, and dependencies might confuse the later. We do have auto-linking of issues cited in messages, so most users can mention dependencies (or duplicates, etc.) to help. But maybe we should have a role that sits between User and Developer for power users. I don't think dependencies/superseder are reason enough for that, but the principle of allowing people to help more might be. +1 on getting more helping hands -0 on adding Dependencies to the default User role ---------- nosy: +ajaksu2 status: unread -> chatting _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 31 09:05:52 2010 From: metatracker at psf.upfronthosting.co.za (techtonik) Date: Wed, 31 Mar 2010 07:05:52 +0000 Subject: [Tracker-discuss] [issue330] Users should be able to set bug dependencies In-Reply-To: <1270016353.69.0.59719876843.issue330@psf.upfronthosting.co.za> Message-ID: <1270019152.21.0.0211737361146.issue330@psf.upfronthosting.co.za> techtonik added the comment: The Dependencies/Superseder field unleashes all its confusion powers just by presence in UI. And it is currently present for casual reporters. NOTE: We speak about bug modification form - not about bug reporting form. The "power user" concept is flawed. Nobody is monitoring who is eligible for "super powers" and people usually do not ask. Either they do not know how/where/whom to ask or do not know about the privileges at all. Keeping privileges under tight control does more harm than good. There is no irreversible changes in tracker just like in version control system. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 31 09:09:53 2010 From: metatracker at psf.upfronthosting.co.za (Daniel Diniz) Date: Wed, 31 Mar 2010 07:09:53 +0000 Subject: [Tracker-discuss] [issue329] "Your Issues" list is empty In-Reply-To: <1270017100.14.0.684273387671.issue329@psf.upfronthosting.co.za> Message-ID: Daniel Diniz added the comment: Adding some context for when this gets done :) The "Your Issues" page is just a issue list with a custom title, so pagination comes free. The URL is in this format: http://bugs.python.org/issue?[status and sort] &@dispname=Your%20Issues &[...] &@filter=status,assignee[... ]&assignee=[User ID] So you can set the 'no issues' message in the link. The link is generated in page.html by this code:
  • Your Issues
    • So the easiest way to get all three listings would be to add the two other as links (copy and paste, filter on creator and nosy instead of assignee). Hide what is currently 'Your Issues' from non-Developers and figure three neat names for these reports and we're set. Making 'Your Issues' a special page that links to the different reports (or a default query) is doable, just a bit more involved. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 31 09:27:53 2010 From: metatracker at psf.upfronthosting.co.za (techtonik) Date: Wed, 31 Mar 2010 07:27:53 +0000 Subject: [Tracker-discuss] [issue329] "Your Issues" list is empty In-Reply-To: <1270016099.97.0.990047736991.issue329@psf.upfronthosting.co.za> Message-ID: <1270020473.25.0.458041858083.issue329@psf.upfronthosting.co.za> techtonik added the comment: I would avoid having too much queries on the interface if it is doable to have a single page. For example, search link is now completely lost in navigation menu. The same might happen with "Your details" link. The names I have for my custom queries are: 1. Participated 2. Patches 3. Submitted They would be better seen and sorted as: 1. Reported 2. Participated 3. My patches _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 31 10:18:21 2010 From: metatracker at psf.upfronthosting.co.za (Ezio Melotti) Date: Wed, 31 Mar 2010 08:18:21 +0000 Subject: [Tracker-discuss] [issue330] Users should be able to set bug dependencies In-Reply-To: <1270016353.69.0.59719876843.issue330@psf.upfronthosting.co.za> Message-ID: <1270023501.97.0.49973653737.issue330@psf.upfronthosting.co.za> Ezio Melotti added the comment: Note that users with the "Developer" role on the tracker are not necessarily Python Core Developers. If people want to help out with the triaging of issues they can still ask for the "Developer" role, as explained in the second paragraph of this page: http://www.python.org/dev/contributing/#triaging-issues . I also agree that the "dependencies" and "superseder" fields are confusing, and I think it's better if "basic" users just leave a comment instead of editing the fields with the risk of setting them wrong (many users are already confused enough by other simpler fields, e.g. "type" and "stage"). ---------- nosy: +ezio.melotti _______________________________________________________ PSF Meta Tracker _______________________________________________________ From rdmurray at bitdance.com Wed Mar 31 14:30:16 2010 From: rdmurray at bitdance.com (R. David Murray) Date: Wed, 31 Mar 2010 08:30:16 -0400 Subject: [Tracker-discuss] comments and questions on your tracker wiki update In-Reply-To: References: <20100330140542.C2CAD1FEEA9@kimball.webabinitio.net> Message-ID: <20100331123016.C9A8B1FF016@kimball.webabinitio.net> On Wed, 31 Mar 2010 09:08:40 +0300, anatoly techtonik wrote: > How about crossposting to Tracker-discuss? And subscribing Sure. > Tracker-discuss to relevant wiki pages. Might not be a bad idea, but I'm not sure how that would get set up. > On Tue, Mar 30, 2010 at 5:05 PM, R. David Murray wrote: > >> + * confusing, 'fixed' issue is neither 'committed' nor 'rejected' > >> at all times -- anatoly techtonik > > > Could you explain more about what you find confusing? I don't understand > > it from this comment. > > It is better to have link here or quote, so with you permission: > "When an issue is closed stage should be automatically set to > 'committed/rejected'." > > This is confusing. The state of closed bug should say explicitly - > committed or rejected, and both should be a closed state. > Committed/rejected doesn't give you a glue about why the issue was > closed. You would have to reread the thread to understand this. I do > not have much time to read comments. OK, so what you find confusing is that the final stage is named 'committed/rejected'. Perhaps 'finalized' or 'completed' would be more descriptive. This stage is redundant with the status 'closed': they mean the same thing. But it would be weird and even more confusing to leave an item in some other stage when it is closed, so we have that final stage for when it is closed. In many ways 'stage' as a whole is redundant with 'status'. It might be better to combine them. That's an idea that would need to be discussed carefully, I think. Alternatively, if some of the other proposals on http://wiki.python.org/moin/DesiredTrackerFeatures are implemented, perhaps stage could be dropped entirely. > >> 1) merge the nosy list from this issue into the specified issue > >> + * seems a bad idea to me -- anatoly techtonik > > > > Why does it seem like a bad idea? > > Because signal to noise ratio is reduced. You time to read through the > comments. It is harder to follow the thread. I don't understand why this would increase the noise. Perhaps you are misunderstanding what I'm suggesting? The idea is to have anyone who made themselves nosy on the issue in question gets added to the nosy list for the issue that this issue is a duplicate of. One presumes that if they were interested in the duplicate, they will be interested in the original and will want to see any new traffic on it. They can always remove themselves. Merging the nosy lists is something triage people sometimes do manually when closing an issue as a duplicate, but don't do as often as we should because it is a bit of a pain. Thus the suggestion to automate it. > >> 3) close the issue (setting stage to committed/rejected per above) > >> + * `duplicate` would be better -- anatoly techtonik > > > > Duplicate is not a stage (and I mean that in terms of the English > > language, not the fact that it isn't currently listed as a stage in > > the tracker). Duplicate is a resolution, and I already suggest setting > > the resolution to duplicate in the previous item. > > OMG. There is "stage", "resolution", "status" - it is too much. Take a > look at Trac - there no resolution field until bug is closed. I > propose to leave "stage" at the place of resolution if the bug is > opened and transform in to "resolution" when closed. Well, feel free to propose a code patch, but I suspect that having a field mutate its identity is distinctly non-trivial in Roundup. --David From stephen at xemacs.org Wed Mar 31 15:59:12 2010 From: stephen at xemacs.org (Stephen J. Turnbull) Date: Wed, 31 Mar 2010 22:59:12 +0900 Subject: [Tracker-discuss] A "gsoc" keyword In-Reply-To: References: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> <94bdd2611003282318w3af9a7c6xc9e29f9ba9438be4@mail.gmail.com> Message-ID: <87tyrwsibz.fsf@uwakimon.sk.tsukuba.ac.jp> anatoly techtonik writes: > I would say that "keywords" concept is flawed in Python tracker. > Usually tracker users are free to add whatever keywords they need. In > Python tracker keywords are: > 1. limited to devs only > 2. limited to a rigid list Keywords are only useful if they allow collecting closely-related ideas under one rubric. (2) is therefore essential, so that people know what keywords are available. (1) can be relaxed by giving more people tracker privileges (I believe the criteria for tracker privileges are already somewhat looser than "developer", ie, committer). But it should not be relaxed to allowing arbitrary keywords; I've been there, done that, and the T-shirt is now used for mopping up kitchen spills. At the very least, users should be made to add to the rigid list before being allowed to use a new keyword. But in my experience that didn't work very well; at one point over 30% of the keywords in my tracker were near-dupes (two of which were simply typos and one of which was a company's name!) Clearly many users misunderstood the use of keywords, or perhaps intentionally abused them. > So, speaking about keywords, I would rename them to devtags and > introduced real space separated keywords concept. If I understand your suggestion correctly, it is nearly pointless when we have full-text searches. From stephen at xemacs.org Wed Mar 31 16:04:04 2010 From: stephen at xemacs.org (Stephen J. Turnbull) Date: Wed, 31 Mar 2010 23:04:04 +0900 Subject: [Tracker-discuss] A "gsoc" keyword In-Reply-To: References: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> Message-ID: <87sk7gsi3v.fsf@uwakimon.sk.tsukuba.ac.jp> Brett Cannon writes: > On Sun, Mar 28, 2010 at 07:44, Tarek Ziad? wrote: > > I've started to create a serie of issues that could be done during gsoc. > > Could we add a 'gsoc' keyword in the tracker ? or something similar > > already exists ? > If they are being created specifically for GSoC, couldn't that simply be put > in the title? That way we don't need to add another keyword. Maybe others > disagree, but we are ending up with a lot of keywords IMO. I think this is an excellent use for a keyword. GSoC has nothing to do with the title, and this is the kind of association we might want to retire all at once. I agree with Martin (IIRC): if you think there are too many keywords, retire a few. From stephen at xemacs.org Wed Mar 31 18:37:59 2010 From: stephen at xemacs.org (Stephen J. Turnbull) Date: Thu, 01 Apr 2010 01:37:59 +0900 Subject: [Tracker-discuss] comments and questions on your tracker wiki update In-Reply-To: <20100331123016.C9A8B1FF016@kimball.webabinitio.net> References: <20100330140542.C2CAD1FEEA9@kimball.webabinitio.net> <20100331123016.C9A8B1FF016@kimball.webabinitio.net> Message-ID: <87fx3gsazc.fsf@uwakimon.sk.tsukuba.ac.jp> R. David Murray writes: > > OMG. There is "stage", "resolution", "status" - it is too > > much. Take a look at Trac - there no resolution field until bug > > is closed. I propose to leave "stage" at the place of resolution > > if the bug is opened and transform in to "resolution" when > > closed. > > Well, feel free to propose a code patch, but I suspect that having a > field mutate its identity is distinctly non-trivial in Roundup. IIUC, this would actually be quite simple, because you don't need or want the field's identity to change. It simply requires a bit of TAL that presents the word "Stage" as the tag for the variable named "stage" if its value is unresolved, and presents "Resolution" for that variable if its value is one of the resolved values. However, it's really important to create a "resolved" predicate for this purpose if you don't have one already. (That's one of the real annoyances I feel in configuring my own Roundups -- too much of the standard template depends on the particular values they chose for certain fields.) From metatracker at psf.upfronthosting.co.za Wed Mar 31 19:33:21 2010 From: metatracker at psf.upfronthosting.co.za (Brett C.) Date: Wed, 31 Mar 2010 17:33:21 +0000 Subject: [Tracker-discuss] [issue253] Remove outdated keywords In-Reply-To: <1237513464.32.0.072891570662.issue253@psf.upfronthosting.co.za> Message-ID: <1270056801.75.0.154443337522.issue253@psf.upfronthosting.co.za> Brett C. added the comment: I don't know if anyone has taken the time to gather the stats, but doing a simple search where you sort on activity (descending order), don't care about the status, and choose the keyword you care about, and you can find out how useful the keyword has been. For instance, 26backport has been set on 72 issues ever and the last use was yesterday, but with usage only once every week or two recently, and then spanning up to two months with no usage. _______________________________________________________ PSF Meta Tracker _______________________________________________________ From brett at python.org Wed Mar 31 19:36:26 2010 From: brett at python.org (Brett Cannon) Date: Wed, 31 Mar 2010 10:36:26 -0700 Subject: [Tracker-discuss] A "gsoc" keyword In-Reply-To: <87sk7gsi3v.fsf@uwakimon.sk.tsukuba.ac.jp> References: <94bdd2611003280744o5d343be3ia520e43e24ca17e9@mail.gmail.com> <87sk7gsi3v.fsf@uwakimon.sk.tsukuba.ac.jp> Message-ID: On Wed, Mar 31, 2010 at 07:04, Stephen J. Turnbull wrote: > Brett Cannon writes: > > On Sun, Mar 28, 2010 at 07:44, Tarek Ziad? > wrote: > > > > I've started to create a serie of issues that could be done during > gsoc. > > > Could we add a 'gsoc' keyword in the tracker ? or something similar > > > already exists ? > > > If they are being created specifically for GSoC, couldn't that simply be > put > > in the title? That way we don't need to add another keyword. Maybe > others > > disagree, but we are ending up with a lot of keywords IMO. > > I think this is an excellent use for a keyword. GSoC has nothing to > do with the title, and this is the kind of association we might want > to retire all at once. > > I agree with Martin (IIRC): if you think there are too many keywords, > retire a few. > Since I seem to be the only one at all worried, I have overruled myself and added the gsoc keyword. I originally called it GSoC, but no other keyword is capitalized so it sorted in an odd fashion. -------------- next part -------------- An HTML attachment was scrubbed... URL: From metatracker at psf.upfronthosting.co.za Wed Mar 31 22:21:25 2010 From: metatracker at psf.upfronthosting.co.za (=?utf-8?q?Martin_v=2E_L=C3=B6wis?=) Date: Wed, 31 Mar 2010 20:21:25 +0000 Subject: [Tracker-discuss] [issue328] More intuitive "stage" and "resolution" In-Reply-To: <1270015996.05.0.585266502058.issue328@psf.upfronthosting.co.za> Message-ID: <1270066885.34.0.255147476349.issue328@psf.upfronthosting.co.za> Martin v. L?wis added the comment: -1. Having a drop down box where some values are valid only at some point would be highly confusing; it is better to have them at separate fields. ---------- nosy: +loewis status: unread -> chatting _______________________________________________________ PSF Meta Tracker _______________________________________________________ From metatracker at psf.upfronthosting.co.za Wed Mar 31 22:59:22 2010 From: metatracker at psf.upfronthosting.co.za (techtonik) Date: Wed, 31 Mar 2010 20:59:22 +0000 Subject: [Tracker-discuss] [issue328] More intuitive "stage" and "resolution" In-Reply-To: <1270015996.05.0.585266502058.issue328@psf.upfronthosting.co.za> Message-ID: <1270069162.74.0.972434901399.issue328@psf.upfronthosting.co.za> techtonik added the comment: You do not need to manipulate the "low-level" state of the issue using "stage" and "resolution" 'CPU registers'. Instead you choose an action, either: 1. close as ... (fixed, invalid, duplicate) 2. change status ... (need patch, need review, need test, need ...) Take a look at Trac - http://trac.edgewall.org/demo-0.12/ticket/157#no3 for an example. _______________________________________________________ PSF Meta Tracker _______________________________________________________