Team Chat Logs

September 29, 2009

2009 8
Mo Tu We Th Fr Sa Su
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30        
<< <   > >>

[00:12:19.610671]<otaku42>moin
[00:19:20.630097]<passatijy>hello everybody
[00:19:56.290497]<passatijy>Is somebody alive here ?
[00:22:35.864556]<passatijy>I have a question about TracDev and docs about plugin development
[00:22:53.979186]<wildintellect>ask away, just don't expect immediate responses
[00:23:16.856724]<passatijy>The example of HelloWorld from " how to make an eggs" dont work for me :(
[00:24:01.272464]<passatijy>First i'm try to copypast from example for 9.0 version trac - and got a "no content-size" error
[00:24:46.285756]<passatijy>Second, i'm try to copypaste expample for 11.0 version and get nothing visible changes in interface
[00:28:29.795537]<passatijy>In my development environment i'm use a 12.x version (fresh from svn and may be somthing wrong with examples ?
[00:34:04.511427]<passatijy>I take examples from here - http://trac-hacks.org/wiki/EggCookingTutorial/BasicEggCooking
[00:50:07.258976]<passatijy>That is : 11:48:59 Trac[main] ERROR: Internal Server Error:
[00:51:00.367407]<passatijy>... File ... apy.py line 483 in write raise RuntimeError("No Content-Length header set")
[00:51:00.908048]<ruiwen>Hi not exactly Trac per se, but has anyone gotten trac-git to work (view revisions in the timeline) with a trac repo that was migrated from svn using git-svn?
[01:12:32.549580]<bionoid>passatijy: In 0.12 you need to send the content length header before any req.write() calls
[01:12:58.314945]<bionoid>passatijy: http://trac.edgewall.org/wiki/TracDev/ApiChanges/0.12
[01:39:59.085797]<evil_twin>t.e.o: Ticket #2647 (Support AND/OR queries) updated - <http://trac.edgewall.org/ticket/2647#comment:36> - cboos
[01:42:25.703179]<ruiwen>Hi, anyone know what's causing a LookupError: unknown encoding: utf-8 #iso-8859-15?
[01:47:30.492945]<passatijy>Thanks, bionoid, i'll try this
[01:47:43.873072]<bionoid>ruiwen: Possibly a broken python installation? Try u'test'.encode('utf-8') in your interpreter
[01:47:48.371690]<bionoid>passatijy: np
[01:51:40.924677]<ruiwen>bionoid: It's ok, found the issue. Apparently it's trac.ini not liking comments in itself on the same line as a valid value, eg. the "#iso-8859-15" bit
[01:51:50.239655]<ruiwen>bionoid: thanks though =)
[02:02:48.876315]<JosefAssad>If I am sitting on a patch which introduces a new configuration option so people can change the term "Ticket" in trac (to, say, "Issue" or whatever). Is this something there's interest in?
[02:08:39.968330]<wildint>probably, I know there's great debate over naming of "Bugs"
[02:09:47.992474]<JosefAssad>ok
[02:09:58.641704]<bionoid>JosefAssad: I know some people have asked about it; Currently I do it with a microscopic Genshi filter in a plugin. I'd be interested to see how you solved it in source. (do you handle the special case "ticket%s", $num>0?'s':'' which leaves a trailing 's' in some places with my technique?)
[02:10:11.319527]<JosefAssad>I'm still trying to resolve 5 functional test failures but otherwise it's good (against trunk)
[02:10:46.142314]<JosefAssad>bionoid: I'm not very smart. I trawled through the system code looking for all the places trac emits "Ticket" in all its capitalization and plural forms
[02:11:41.661680]<JosefAssad>bionoid: I used a ListOption consisting of Ticket,ticket,Tickets,tickets I'm afraid. But it works
[02:12:22.861345]<bionoid>Aha, ok :)
[02:13:47.909993]<bionoid>JosefAssad: http://pastebin.com/m7f93ac52 -> this is the minimal plugin I use atm
[02:14:30.716423]<JosefAssad>bionoid: saved so I can look at it later and larn one more thing. :)
[02:14:55.247740]<JosefAssad>conceptually though, it would be good if we didn't have to rewrite terms no?
[02:15:36.957496]<bionoid>JosefAssad: Yeah this is being fixed with "proper" localization in 0.12
[02:22:18.806838]<JosefAssad>er.
[02:22:53.537335]<JosefAssad>That kind of obviates the need for making the term Ticket configurable, no?
[02:23:12.454629]<JosefAssad>wait, lemme think. No it doesn't does it? People won't want to edit .po files to change the term
[02:34:42.726084]<passatijy>bionoid , it works (helloworld and content length).
[02:36:22.824270]<bionoid>JosefAssad: Well the point is they _will actually have_ somewhere to edit it, which is not the case today.
[02:36:25.995053]<bionoid>passatijy: Great :)
[02:37:19.880559]<passatijy>But I didnt understand, where I can get a real length of content. I'm try a len(req) and so other but dont guess. I'm just set a number 200 and wait for error. The page is render in browser but browser try to load more content, how I can explain progress bar.
[02:37:48.227319]<passatijy>btw, thanks :)
[02:38:20.729372]<bionoid>passatijy: 0 usually does the trick ;-)
[02:41:29.964559]<passatijy>hmm .. just a moment :)
[02:44:22.529665]<JosefAssad>bionoid: true. If they want to grep/sed through the .po though, no? Isn't it more convenient to have one place to change it?
[02:47:00.202412]<bionoid>JosefAssad: Hmmm.. not quire sure I see what you're getting at? There will be one place to change it, your point seems to be that it won't happen from trac.ini?
[02:47:42.634318]<bionoid>JosefAssad: IMHO it would make no sense to have 1900 options covering all the used terms in trac.ini
[02:48:14.846317]<bionoid>Not to mention the plural rules etc
[02:49:06.887981]<JosefAssad>bionoid: completely agree re. 1900 options
[02:49:39.787491]<passatijy>bionoid , when I pass the 0 to content length , I'm got a blank screen.
[02:49:56.739506]<bionoid>passatijy: Anything useful in the logs?
[02:50:00.447304]<JosefAssad>bionoid: all I'm thinking of is, there's going to be the word ticket in lots of separate strings (granted, in one file) in the .po file. Going through _those_ strings to change Ticket to whatever could still be a pain
[02:50:07.926508]<JosefAssad>I might be missing something of course.
[02:50:11.318121]<bionoid>passatijy: (sorry havn't done any real work with 0.12 so :\)
[02:50:59.447326]<passatijy>just a moment
[02:51:51.794529]<passatijy>I'm need to recompile with content length=0 , just 3 minutes
[02:53:12.867237]<bionoid>JosefAssad: Yeah, I (think I) get your point now ;) I still think the .po way would be better (even though the most simplistic scenario would be easier to solve with your method)
[02:54:33.855960]<passatijy>Trac[main] DEBUG: Dispatching <Request "GET u'/helloworld'">
[02:54:41.742936]<bionoid>JosefAssad: But what if you were to have three different locales active? Like one user in spain, one in china and one english. How would your approach fare for a single Trac instance with multiple users? (I think not very well); Using the 0.12 approach you can make those changes across languages..
[02:54:45.035945]<passatijy>127.0.0.1 - - [29/Sep/2009 13:54:01] "GET /test/helloworld HTTP/1.1" 200 -
[02:54:54.893195]<passatijy>Nothing interesting
[02:56:17.962307]<passatijy>Content-Type text/plain
[02:56:17.981845]<passatijy>Content-Length 0
[02:56:31.399555]<passatijy>Firebug says that content length 0
[02:57:18.242366]<passatijy>I'm think that I need to calc the really content length, but I misunderstand the concepts
[02:59:32.931274]<bionoid>passatijy: Yeah that's what you _should_ do ;) But you'll have to wait for someone 0.12-aware to show up; there are probably some standard ways of doing it.
[03:02:05.376496]<passatijy>ok thanks anyway
[03:06:00.652470]<JosefAssad>bionoid: lemme think on that. Valid point. See if I can fidn some elegant way of centralising that text without compromising on-the-fly and concurrent multilinguality
[03:09:18.879925]<JosefAssad>did I just abose the word "concurrent"? I think I did
[04:47:18.566831]<Sp4rKy>how can I force TracAccountManager not to autologin me after a logout
[04:47:41.175663]<Sp4rKy>even with LoginModule disabled, it continue to autologin me
[04:48:00.026060]<Sp4rKy>aka : If I login, logout, then reclick on "login", it doesn't ask me account informations
[04:53:52.044001]<passatijy>How i know its not a TracAccountManager issue
[05:03:12.220962]<bionoid>Sp4rKy: Like passatijy said, this is an issue with HTTP auth being stored in your browser after login.
[05:08:48.923853]<Sp4rKy>I have 5 tracs
[05:09:01.537343]<Sp4rKy>the only one with this issue is the one with accountmanager plugin
[05:15:51.928934]<bionoid>Sp4rKy: Well, I have several tracs, none with acctmgr, all exhibiting this behaviour. Whatever we may conclude from that ;)
[05:16:08.091496]<Sp4rKy>-_-
[05:16:39.432085]<Sp4rKy>so, even after explicitly press "logout", the session isn't destroyed ...
[05:16:46.563396]<Sp4rKy>what a trange behaviour
[05:17:01.623462]<Sp4rKy>anyway, I can't understand why I don't have it in other trac
[05:17:13.393759]<bionoid>The session as such is destroyed. It's just that the browser resubmits the http authentication credentials (in my case anyway)
[05:18:05.712738]<bionoid>You could monitor the DB and sniff the network traffic to determine what's happening in your case.
[05:24:00.755608]<bionoid>Sp4rKy: I just did a test for fun. Start Opera, login then logout. Hit Login again to observe the behaviour. Hit logout. Hit tools -> delete private data, then Login again and you get password request. Voila.
[05:24:32.211373]<Sp4rKy>bionoid: why I don't get this beahaviour with all my tracs but 1
[05:24:51.410727]<bionoid>Sp4rKy: TBH I don't believe you :P
[05:25:15.360084]<Sp4rKy>:)
[05:25:47.016169]<Sp4rKy>hmmm the ame trac only has "Trac 0.11.1" in its plugins list, whereas all others also have Genshi
[05:55:01.606547]<evil_twin>t.e.o: Ticket #8700 (patch for trac-svn-post-commit-hook.cmd) created - <http://trac.edgewall.org/ticket/8700> - federico.briata@…
[05:55:06.607378]<evil_twin>t.e.o: argument-fix_trac-svn-post-commit-hook_cmd.patch attached to Ticket #8700 - <http://trac.edgewall.org/attachment/ticket/8700/argument-fix_trac-svn-post-commit-hook_cmd.patch> - federico.briata@…
[06:15:01.631400]<evil_twin>t.e.o: Ticket #8700 (patch for trac-svn-post-commit-hook.cmd) closed - <http://trac.edgewall.org/ticket/8700#comment:1> - federico.briata@…
[06:20:01.647294]<evil_twin>t.e.o: Ticket #8701 (Display last n changes on the timeline) created - <http://trac.edgewall.org/ticket/8701> - Sebastian Krysmanski <sebastian@…>
[06:35:06.798731]*retracile croaks something indelicate about mornings.
[06:35:47.041189]<dunk_>!morning retra!
[06:36:02.807394]<retracile>hey dunk_ :)
[06:57:30.530019]<paseante1>is it possible to generate an URL with a diff that excludes certain changesets from the diff?
[06:58:07.270712]<paseante1>I don't want the merges to appear in the diff
[07:13:27.676052]<retracile>paseante1: no
[07:13:46.649152]<paseante1>that's too bad :(
[07:13:59.910644]<retracile>paseante1: more than that, it doesn't exist in the repository
[07:14:11.302854]<jhammel>yeah, i was going to say....that would be hard
[07:14:24.758814]<retracile>paseante1: you're wanting a diff that may not be something that can be created.
[07:15:04.269545]<retracile>paseante1: you want changesets 2,3,6,7,8, but changeset 6,7,8 may not apply cleanly to the tree after changeset 3 because of the merges that happened in 4 and 5.
[07:20:13.144736]<paseante1>retracile: you are right, altough actually the merges happened in an isolate subdir, so it could be done
[07:20:22.586921]<paseante1>I'm speaking in my case of course
[07:21:40.551047]<retracile>paseante1: Well, you could create a branch, merge the changes into the branch, then reference the resulting diff
[07:35:38.454693]<paseante1>retracile: I'm thinknig, shouldn't doing a diff between the branch and the trunk (from which the merges came) work?
[07:46:10.197173]<paseante1>it's very strange, for some reason the command line client does the diff well, but Trac doesn't
[07:46:14.358514]<retracile>paseante1: trac can do that, and it may give you the result you're lookign fore.
[07:46:28.773857]<retracile>s/lookign fore/looking for/
[07:46:28.787721]<evil_twin>retracile meant: paseante1: trac can do that, and it may give you the result you're looking for.
[08:09:17.166078]<Hossrod_>Hello all. I just setup Trac 11.5 by following the TracOnWindows instructions. Trac seems to be working fine, but whats the URL for my SVN repository?
[08:10:10.931365]<jhammel>Hossrod_: trac doesn't serve svn
[08:12:51.190846]<Hossrod_>The instructions had me setup SVN server as part of the install. Is there something more I need to do to enable the SVN server? The browse source tab in Trac shows root, but says no files found (which I expect because I've not checked in anything yet).
[08:13:29.528532]<jhammel>Hossrod_: oh, sorry, didn't read the instructions....surprised that is in there
[08:14:06.130265]<paseante1>how do I clear Trac's cache?
[08:21:04.271866]<tdn>When I edit a ticket and assign it to another user, its status does not change from New -> Assigned. This is probably because the assigned user needs to accept the ticket first. Is there any way to assign tickets without requirering the assigned user's accept?
[08:28:00.211266]<retracile>tdn: yes, by customizing the workflow.
[08:28:21.640442]<retracile>tdn: oh, I should have asked: what version of Trac?
[08:34:49.828059]<Hossrod_>Does anyone know if there are more steps needed to get SVN working from whats in the instructions in TracOnWindows? I can not figure out the repository URL to use so I can checkin code.
[08:37:33.068405]<elshaa>hi
[08:38:46.518424]<elshaa>I have a little problem with trac users authenticated through ldap.
[08:39:09.325987]<elshaa>These users have been removed from ldap but trac still tries to send updates to them, resulting in postfix errors.
[08:39:41.215349]<elshaa>I could delete the user's account with trac-admin, but I'm wondering if his name will still be referenced in the ticket he created
[08:40:02.092151]<elshaa>tickets
[08:40:50.429719]<retracile>Hossrod_: You're running svnserve as a service?
[08:41:26.979497]<retracile>elshaa: yeah, he'll still be referenced. You may be able to remove his email address from his account info though.
[08:43:23.611067]<Hossrod_>I beleive so, I started it like the tutorial said with: net start svnserver
[08:43:51.585385]<Hossrod_>I see svnserve.exe running in taskman under user SYSTEM
[08:44:50.101220]<Hossrod_>oh, and the command I used just before doing the net start command was: sc create svnserver binpath= "C:\Subversion\bin\svnserve.exe --service -r c:\svnrepositories" displayname= "Subversion Server (svnserver)" depend= Tcpip start= auto
[08:44:52.475892]<elshaa>retracile: Thank you ! Will check trac-admin options.
[08:54:03.627898]<retracile>Hossrod_: try 'svn ls svn://localhost' and see what happens
[08:55:37.941662]<Hossrod_>I get back: svn: No repository found in 'svn://localhost'
[08:57:51.042323]<retracile>Hossrod_: hm. What directory is your svn repository in?
[09:01:05.349135]<Hossrod_>D:\svnrepositories\myproject
[09:01:27.919235]<tdn>retracile, just checked the version. It is 0.11.1.
[09:01:41.789727]<tdn>retracile, how do I modify workflow?
[09:01:46.631607]<Hossrod_>Note, the command above I posted, was just copy from the trac tutorial. When I did it live, I edited it to use D:.
[09:02:21.574442]<retracile>Hossrod_: try 'svn ls svn://localhost/myproject'
[09:02:43.995895]<retracile>tdn: did you upgrade from 0.10, or has this install always been 0.11?
[09:02:54.942995]<tdn>retracile, upgrade.
[09:03:07.250666]<Hossrod_>No messages when I tried that. Just back to command prompt.
[09:03:31.099474]<retracile>!inform tdn wiki TracWorkflow
[09:03:31.114091]<evil_twin>tdn: http://trac.edgewall.org/wiki/TracWorkflow
[09:03:49.523811]<retracile>Hossrod_: ok, that is the root url for your (currently) empty repository.
[09:04:31.113465]<tdn>evil_twin, ok. So I need to change my Trac configuration to do this?
[09:04:31.124498]<evil_twin>nothing known about ok. so i need to change my trac configuration to do this
[09:05:05.369748]<tdn>Oh. A bot.
[09:05:11.200204]<tdn>retracile, ok. So I need to change my Trac configuration to do this?
[09:06:14.937633]<retracile>tdn: yeah, evil_twin's a bot, and a not too bright one at that. One of those "we oughta improve..." projects that we just haven't gotten around to.
[09:06:32.705628]<dunk_>yeah go to work retracile !!
[09:06:36.603378]<dunk_>;]
[09:06:51.867433]<retracile>tdn: For customizing the workflow, yes, you change the Trac configuration.
[09:07:22.026791]<Hossrod_>If I open http://localhost/myproject from a web browser on the same machine, I should see SVN right? Cause I just get file not found.
[09:07:23.249625]<tdn>retracile, ok. Then I'll have to look at it tomorrow.
[09:07:26.699999]<tdn>retracile, thanks.
[09:07:36.329110]<retracile>tdn: the 0.10 workflow had a number of "warts", such as "reassign ticket to X" means that the ticket winds up in the "new" state. (So obvious that when you assign a ticket, it must be new, not assigned, right? ;) )
[09:08:06.179813]<retracile>tdn: note the migration script; that will probably be your first step
[09:08:39.249357]<retracile>tdn: well, first step after doing a backup of course ;)
[09:08:42.672718]<tdn>retracile, ok. Will have to look at that.
[09:08:44.141572]<retracile>dunk_: Hehe :)
[09:08:58.147869]<dunk_>;]
[09:09:30.372879]<dunk_>is wish you had time to perfect the log analysing code
[09:09:39.178142]<Hossrod_>Note, in the command window that is running tracd, I see this when I try going to http://localhost/myproject: 127.0.0.1 - - [29/Sep/2009 09:04:44] "GET /myproject HTTP/1.1" 404 -
[09:10:51.852318]<retracile>dunk_: Yeah, I think there is some interesting potential in that. Life has had me a bit swamped of late though. :/
[09:11:05.146665]<dunk_>me too dude ;[
[09:11:15.096877]<dunk_>to
[09:11:17.194639]<dunk_>hmm
[09:11:28.260980]<dunk_>but yes there is potential
[09:11:48.887219]<retracile>dunk_: "too" means "also"
[09:12:16.065509]<dunk_>ok then is too ;]
[09:12:45.089433]<retracile>:)
[09:13:08.207599]<dunk_>if was thinking of too in too much
[09:13:51.960923]<retracile>dunk_: ah, I see. Isn't English fun? ;)
[09:14:13.592237]<dunk_>its not my native language ;]
[09:14:34.950472]<retracile>dunk_: yeah, I know
[09:37:48.078080]<mahound>hello
[09:38:15.378819]<fission6>is there a plugin out there to parse puppet classes, etcc... and create a wiki out of it
[09:38:19.681404]<mahound>is there any way to add a list of the latest commits? Like RecentChanges, but for svn/git/whatever commits?
[09:55:19.232830]<Hossrod_>I think I'm gonna skip windows and tryout a tutorial I found for setting up Ubuntu, Trac, SVN and Apache (with https). Maybe try in a virtual machine first.
[09:58:47.079162]<Hossrod_>http://en.newinstance.it/2009/08/27/tutorial-ubuntu-904-apache-with-ssl-subversion-over-http-https-and-trac/
[10:16:55.329242]<cboos>hello
[10:17:32.657536]<jhammel>cboos: hey there
[11:22:16.927489]<cboos>hey osimons
[11:22:30.831468]*cboos is doing some exploratory work on lynx ;-)
[11:22:49.035760]<cboos>"first do not harm"
[11:23:10.530708]<osimons>ah. that lynx... i see. don't mind me - i'm nowhere close...
[11:23:42.034861]<cboos>well I was talking about myself, the "do not harm" applied to me :-)
[11:24:19.842604]<cboos>so right now I'm trying to get hg on that box, would have build it myself, but... no "make" on that machine :-(
[11:24:44.339007]<cboos>IIRC, cmlenz used apt-get to install gdb, that other day
[11:25:43.545555]<cboos>or was that dpkg?
[11:26:04.050116]<cboos>(I'm not really familiar with this stuff)
[11:41:11.748270]<FireWire>how old is version .11b2?
[11:41:22.724986]<cboos>very old
[11:41:39.942029]<cboos>2 years maybe
[11:41:56.361205]<jhammel>wow, is that true?
[11:42:01.395815]<cboos>19 months precisely
[11:42:09.521988]<jhammel>wow...seems like just yesterday
[11:42:40.284981]<cboos>I was almost sure it was /more/ than 2 years ;-)
[11:43:09.542814]<retracile>heh :) Yeah, I know that feeling :)
[11:43:20.325345]<cboos>jhammel: maybe you were thinking about 0.11.2, that's less than a year ... but beta2, really old stuff
[11:44:10.216529]<FireWire>Thank you, company i work for is currently running .11b2, and im debating if i should upgrade it... should an upgrade be fairly straight forward?
[11:44:26.832905]<jhammel>cboos: i remember when 0.11 came out...yeah, seems like a year-ish
[11:44:32.221337]<jhammel>though i suppose it was longer
[11:45:01.243607]<retracile>FireWire: _should_ be. Upgrade a testing copy of the system to be sure. Backups recommended, of course.
[11:45:26.889420]<cboos>FireWire: fairly - I don't think you'll need a db upgrade, but keep a db backup in case - as eli just said ;-)
[11:45:37.671163]<cboos>you'll need a "wiki upgrade" to get recent docs, though
[11:46:02.644049]<FireWire>understood :) and the change log should give me a good overview of the major feature changes, if I need to justify the decision?
[11:47:28.827356]<cboos>lots of memory issues and speed issues have been addressed, 0.11.5 is /nearly/ satisfying (though 0.11.6 will be better)
[11:47:50.475217]<workPragmatist>can you make submitting a ticket go back to new ticket
[11:47:56.642420]<workPragmatist>instead of "show" ticket
[11:48:00.806504]<cboos>if you're using sqlite for the db, it might be worth going to 0.11.6dev even
[11:48:07.928168]<cboos>(the above was addressed to FireWire)
[11:48:13.944914]<jhammel>cboos: lots of speedups for sqlite in 0.11.6?
[11:48:37.180619]<retracile>jhammel: 'database is locked' improvements, IIRC
[11:48:43.633387]<jhammel>coolz
[11:48:58.465596]<cboos>right, the "database is locked" problem should be gone
[11:49:19.966635]<jhammel>awesome
[11:49:29.200823]<cboos>btw, prize awarded for anyone getting a "database is locked" exception with 0.11.6dev ;-)
[11:49:36.768204]<jhammel>heh, wish i didn't take many hours of effort working around
[11:50:42.705844]<cboos>sorry for that - the solution was at hand for quite a long time, just didn't get around to it until recently
[11:51:55.956712]<jhammel>cboos: np, just glad to see it fixed
[11:52:35.277827]<FireWire>alright, thanks guys, il probably be back in here with questions if i get around to doing it :)
[11:56:06.976475]<workPragmatist>is it possible to put in the wiki a list of the components available
[11:56:16.609808]<workPragmatist>actually nm I bet you can use Query
[12:34:27.523327]<agrajag>Anyone have custom query working with the DateFieldPlugin?
[13:50:09.546372]<evil_twin>t.e.o: Ticket #8702 (Different workflows when viewing own tickets/others tickets) created - <http://trac.edgewall.org/ticket/8702> - adrian.price@…
[13:50:39.582759]<jhammel>wow, that would be a big one
[13:50:40.839489]<jhammel>i think
[13:59:09.189779]<retracile>jhammel: nah, you can already do it :)
[14:02:19.758968]<retracile>jhammel: should be able to use http://trac-hacks.org/wiki/VirtualTicketPermissionsPlugin to do it, though I haven't tried
[14:02:46.338447]<jhammel>retracile: wow, that's pretty cool...hadn't noticed that plugin before
[14:03:23.941012]<retracile>hrm... actually, you need the inverse of that....
[14:03:42.342791]<jhammel>retracile: hmmm, yeah, i suppose so
[14:04:45.511352]<retracile>jhammel: there is also http://trac.edgewall.org/browser/trunk/sample-plugins/workflow/CodeReview.py which has a similar hook to it...
[14:04:55.379291]<retracile>namely, you can't code review your own tickets :)
[14:05:02.820373]<jhammel>retracile: hah!
[14:05:12.753683]*jhammel wonders who would want to
[14:05:29.575810]<jhammel>jhammel: yes, that seems a pretty solid solution; thanks for doing this!
[14:08:30.241510]<retracile>jhammel: Hehe :) Well, see, if you were 'just trying to get this done', and there was a mandatory 'code review' step in the workflow... there are those who would just mark it as reviewed so that they could get the changes merged and move on
[14:09:17.786932]<jhammel>retracile: heh; goes to prove that any system will break down if it ain't easy
[14:09:37.156683]<jhammel>retracile: so now, you have to create another user, give them appropriate permissions, and then code review
[14:09:40.744551]<jhammel>what a hassle!
[14:10:04.494020]<jhammel>jhammel1: bang-up job! great work!
[14:10:09.933257]<evil_twin>t.e.o: Ticket #8702 (Different workflows when viewing own tickets/others tickets) updated - <http://trac.edgewall.org/ticket/8702#comment:1> - ecarter
[14:10:15.050422]<retracile>Trac was designed to 'get out of your way'. Well, the customizable workflow can be used to get Trac more out of your way, but it can also be used to get _in_ your way. It depends on who is setting up the workflow. :)
[14:10:24.868514]<retracile>hehe :)
[14:10:29.210129]<jhammel>retracile: true that
[14:13:05.602828]<retracile>jhammel: If a developer has to remember to go to the ticket and mark each action so that management can follow along, then it's broken.
[14:13:15.880629]<jhammel>retracile: yeah, pretty much
[14:13:38.878026]<jhammel>retracile: i mean, there's so much like that in the world
[14:14:21.762879]<retracile>jhammel: _But_, if the developer can go to the ticket, and hit a button, and some tedious step he'd have to have done anyway just gets done for him, he'll be delighted. Oh, and it also happens to let management follow along as a _side_effect_... _then_ it's useful and helpful.
[14:15:26.059151]<jhammel>retracile: yeah, otherwise you just end up writing scripts to work around parts of the workflow that are just tedious
[14:15:44.533358]<jhammel>which of course makes the workflow moot/broken
[14:17:39.256695]<retracile>I've been in a place where it was fairly common for a task to get done, then after it was committed to trunk, it got actioned through all the various states one-after-the-other, fighting the GUI the whole way through. The workflow tool was worse than useless.
[14:19:03.614284]<jhammel>retracile: well, we don't use Trac workflow much...because I don't think many people understand it but any, probably, but in our non-electronic workflow, yeah, its come to developers fighting designers...pretty dysfunctional
[14:20:01.817000]<jhammel>retracile: after....5 months?....I think I've finally gotten enough traction on my "Look, I'm going to make my templates work with the default app. If you guys want to do crazy custom templates, then we should figure out a workflow or you are going to need to make them work correctly"
[14:20:27.151315]<jhammel>i.e. I was blamed because I couldn't intuit what their unannounced changes to custom templates were doing
[14:20:49.404342]<retracile>oh lovely. :/
[14:36:06.079411]<abernier>hi, does trac supports git SCM ?
[14:36:44.941691]<retracile>abernier: with a plugin
[14:36:58.665572]<retracile>abernier: but I haven't used it, so I won't be of much help in getting it setup
[14:37:09.872199]<retracile>abernier: Look at trac-hacks.org
[14:37:47.771508]<jhammel>abernier: http://trac-hacks.org/wiki/GitPlugin
[14:38:00.903919]<jhammel>abernier: i haven't used it either though....i'm an svn/hg man myself
[14:38:15.928602]<abernier>hg? mercurial?
[14:38:21.635192]<jhammel>abernier: yeah
[14:38:33.454319]<abernier>jhammel: is mercurial supported with trac?
[14:38:53.634228]<jhammel>abernier: again with a plugin, but that one i have setup and i can testify that it works well
[14:39:08.876094]<jhammel>afaik, svn is the only SCM supported OOTB for Trac
[14:50:10.585724]<evil_twin>t.e.o: messages.po.8389.diff attached to Ticket #5473 - <http://trac.edgewall.org/attachment/ticket/5473/messages.po.8389.diff> - joost <macjoost@…>
[16:09:30.041766]<doki_pen>hehe
[16:15:14.630336]<evil_twin>t.e.o: Ticket #8700 (patch for trac-svn-post-commit-hook.cmd) reopened - <http://trac.edgewall.org/ticket/8700#comment:2> - mike.sherov@…
[17:30:13.185850]<evil_twin>t.e.o: TracUsers edited - <http://trac.edgewall.org/wiki/TracUsers?version=874> - anonymous
[18:19:38.208909]<subopt>I just installed the bitnami trac on my Mac. It won't let me run trac-admin, saying there's no module named trac.scripts.admin. Don't see any mention of the need to adjust any of my paths, but it seems like that's the problem. Anyone seen this?
[23:17:27.369136]<paseante1>Has anyone made successful diffs between a trunk and a branch with merged changes from the trunk?
[23:53:23.934473]<jake-aus>is there an irc room for trac-hacks