Team Chat Logs

September 24, 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        
<< <   > >>

[00:03:30.846363]<k2_soft>i fixed my problem with: export PYTHONPATH=/usr/lib64/python2.3/site-packages:/usr/lib/python2.3/site-packages/
[01:09:06.065770]<gamla_kossan>hi people
[01:09:26.801406]<gamla_kossan>is trac really available for SLES?
[01:09:35.618225]<gamla_kossan>following the link on http://trac.edgewall.org/wiki/TracOnSuSE
[01:09:47.717610]<gamla_kossan>I
[01:10:16.642901]<gamla_kossan>'ve updated yast with the repo specified there, but I can only find setuptools, but no trac package
[01:11:58.083271]<gamla_kossan>oh, now I see - trac isn't available for SLES 10, but for opensuse 10..
[01:23:19.162476]<evil_twin>t.e.o: Ticket #8690 (Missing changeset link) updated - <http://trac.edgewall.org/ticket/8690#comment:2> - cboos
[01:23:24.163060]<evil_twin>t.e.o: 8690_worksforme.png attached to Ticket #8690 - <http://trac.edgewall.org/attachment/ticket/8690/8690_worksforme.png> - cboos
[01:40:06.243260]<jdb>is it possible to hide some wiki pages for anonymous users?
[02:24:20.707447]<JosefAssad>erm. When the docs say actionname.set_owner may optionally be set to a comma delimited list or a single value does this mean python list [foo, bar] or just a list like: foo, bar
[02:50:21.006850]<asmodai>osimons: Awake by chance?
[02:50:31.267490]<osimons>oh yeah, asmodai
[02:51:30.870239]<asmodai>osimons: trying to get this xmlrpc and httpauth thinghy working so I can use Mylyn in Eclipse.
[02:51:43.058827]<asmodai>osimons: still stumped by authentication though
[02:52:50.467263]<asmodai>osimons: the server URL I pass along needs to have the /xmlrpc tacked on to it, right?
[02:53:16.862501]<osimons>i don't really know either since i don't use account manager... the general idea i think is that httpauth should trap /login/xmlrpc and use basic/digest/something
[02:55:21.868615]<asmodai>oh wait, maybe it's a small typo
[02:56:30.487174]<asmodai>mmm, seems that at least auth now works. Now to see where it is pulling it's info from
[02:57:19.731193]<asmodai>Yup, the 404 is now at least changed to asking for credentials.
[02:57:36.139303]<asmodai>Ah, so now I need to set up those paths with Apache's authentication?
[03:04:29.463099]<asmodai>looks like it
[03:11:40.338620]<asmodai>ah no, it uses acctmgr for that
[03:13:37.387419]<asmodai>Yet it is not validating. o_O
[03:24:26.627819]<asmodai>osimons: got it
[03:24:59.124070]<osimons>goodie, asmodai
[03:26:04.195094]<asmodai>osimons: thing is
[03:26:15.417356]<asmodai>osimons: httpauth uses acctmgr to check the password
[03:26:25.045758]<asmodai>so if you use the standard htpasswdstore method, no problem
[03:26:40.830478]<asmodai>you generally configure it to use httpauth for /xmlrpc, /login/xmlrpc
[03:26:48.877166]<asmodai>thing was, it kept asking me for a password over and over
[03:27:01.034572]<asmodai>turns out it needed WSGIPassAuthorization On
[03:28:02.901879]<osimons>something for adding to the docs section on the plugin? as i said, i've never used or seen acctmgr or httpauth, so i know nothing about it...
[03:28:32.178145]<osimons>feel free to improve the plugin auth docs, asmodai :-)
[03:28:36.471753]<asmodai>I'll write a blogpost on it at least
[03:28:44.638068]<asmodai>osimons: yeah, too much on my list :(
[03:28:51.733273]<asmodai>And too damn busy with stuff.
[03:29:11.133402]<asmodai>osimons: Got 5 Trac environments here with acctmgr and using Eclipse :)
[03:29:17.153712]<asmodai>So figured, hey, can as well use Mylyn
[03:30:00.866102]<osimons>it is quite nice, actually - got it up and running just for test, and looks and works really neat
[03:30:06.197437]<asmodai>Aye.
[03:30:14.373377]<asmodai>Personally I think acctmgr ought to be a base feature
[04:02:06.743372]<jdb>is it possible to hide some wiki pages for anonymous users?
[04:27:31.566270]<ismael_>is there extension point to change the square that says when was the last revision. In the ticket view window??
[05:23:23.494292]<evil_twin>t.e.o: Ticket #8691 (separator in trac.log for reload/restart) created - <http://trac.edgewall.org/ticket/8691> - anatoly techtonik <techtonik@…>
[05:26:17.623702]<jdb>should i do something specific when installing a plugin via easy_install? it doesnt show up at the admin page
[05:29:41.142854]<techtonik>i think it is bad idea installing plugins via easy_install system-wide.. it is not that easy anymore if you have multiple Pythons installed.
[05:30:29.335582]<techtonik>you may need to restart server and check that you've installed it for the right python
[05:35:18.131099]<retracile__>jdb: you need to enable the plugin in the [components] section if you install system-wide
[05:35:47.467060]<jdb>thanks all
[05:38:23.641821]<evil_twin>t.e.o: Ticket #8692 (KeyError: u'complete') created - <http://trac.edgewall.org/ticket/8692> - pac.solution@…
[05:48:56.829676]<ismael_>How can I change something in the right side square for revisions in a ticket view
[05:48:58.889109]<ismael_>?
[05:49:25.302836]<ismael_>with a plugin I mean
[05:53:24.109741]<evil_twin>t.e.o: Ticket #8692 (KeyError: u'complete') closed - <http://trac.edgewall.org/ticket/8692#comment:1> - eblot
[06:18:24.593634]<evil_twin>t.e.o: SandBox edited - <http://trac.edgewall.org/wiki/SandBox?version=943> - r.castillo at ziv.es
[06:31:13.176248]<techtonik>@pacopablo: Is there any workaround in Account Manager to validate users from several password stores?
[06:31:13.190835]<evil_twin>nothing known about pacopablo: is there any workaround in account manager to validate users from several password stores
[06:43:24.937821]<evil_twin>t.e.o: Ticket #8693 (inconsistent section naming in trac.ini) created - <http://trac.edgewall.org/ticket/8693> - anatoly techtonik <techtonik@…>
[06:47:52.268321]<JosefAssad>I could be completely failing to understand an option in trac.ini but I'm not sure how to get more detail in timeline. I have ticket_show_details = true which is lovely, but it only lists for example owner changed. Is there a way to show owner changed from who to who instead of just changed?
[06:47:57.685145]<JosefAssad>(that anyone knows of?)
[06:51:09.711431]<collink>I'm getting this from trac when I try to log out, but nothing in apache's error logs: TypeError: __init__() takes at least 2 arguments (1 given)
[06:51:19.365306]<collink>anyone have any suggestions?
[07:23:25.437449]<evil_twin>t.e.o: Ticket #2647 (Support AND/OR queries) updated - <http://trac.edgewall.org/ticket/2647#comment:29> - ebray
[08:05:32.127229]<retracile>collink: you should have a full traceback in your trac log file...
[08:05:42.577271]<collink>I didn't. I figured out the problem though
[08:06:39.111284]<retracile>what was the problem?
[08:17:41.466957]<workPragmatist>is it possible to make track auto assign tickets to the component owner?
[08:19:03.534673]<retracile>workPragmatist: they will be automatically owned by the component owner, but not automatically assigned.
[08:19:06.774974]<techtonik>Everything is possible with the power of plugins!
[08:19:16.446223]<techtonik>Too late.
[08:19:31.801869]<retracile>techtonik: heh
[08:20:02.579313]<retracile>techtonik: but in this case, plugins don't have a hook available to them... or at least, not one intended for that use
[08:20:51.290516]<techtonik>retracile: does that mean there is no event when a plugin can get a change to assign a newly created ticket? filter the incoming form, for example?
[08:21:02.965269]<techtonik>s/change/chance/
[08:21:02.978079]<evil_twin>techtonik meant: retracile: does that mean there is no event when a plugin can get a chance to assign a newly created ticket? filter the incoming form, for example?
[08:21:09.253992]<JosefAssad>erm. As I read this http://osdir.com/ml/version-control.subversion.trac.general/2005-08/msg00165.html the component owner should be the default value for assign to?
[08:23:01.423325]<retracile>JosefAssad: this is one of those confusing warts left-over from 0.10. A ticket has an owner. There is also a workflow state called 'assigned'. A user can own a ticket, but it not be in the 'assigned' state. Confusingly, it still gets called 'assign to <the new owner>'
[08:24:09.817956]<JosefAssad>oh, ok
[08:24:29.266012]<JosefAssad>Surprisingly, I kind of understand that. Might be because I built my first workflow today
[08:24:45.473822]<retracile>techtonik: the workflow does not have a hook for newly created tickets. it's a known deficiency. Right now, tickets are _always_ created in the 'new' state. There is a hook to get notified of the ticket being created, but it wasn't intended to be used for automatically actioning a ticket to another state. It may be possible to abuse it for that purpose, however.
[08:25:25.774578]<retracile>JosefAssad: the simplest thing there is to not have an 'assigned' state, but pick some other name for it.
[08:26:04.175005]<JosefAssad>retracile: do you mean in the atual workflow, to not have a state explicitly called "assigned"?
[08:26:37.701281]<JosefAssad>I actually don't have any state called that, but if I understood you right that _is_ a funny wart
[08:26:50.140197]<retracile>JosefAssad: right. it avoids some of that confusion. But it may be the best name for what you're trying to do, in which case... go with it. :)
[08:28:04.366526]<retracile>JosefAssad: http://trac.edgewall.org/wiki/TracWorkflow#Environmentsupgradedfrom0.10
[08:28:37.067436]<retracile>JosefAssad: another wart: 'assign to someone else', resulting state? "new", of course. why would you expect that to be "assigned"? *cough*
[08:29:51.339892]<JosefAssad>retracile: Look sir, I just got done building my (relatively complicated) workflow today and became an even bigger trac fan, so I must ask you to refrain from ruining my image of trac for me!
[08:29:55.326294]<JosefAssad>heh
[08:30:30.622504]<JosefAssad>I actually don't have any "assign" at all language-wise in my workflow, that must be why it all seems so logical to me
[08:30:49.916026]<retracile>JosefAssad: hehe :) Yeah, well, I'm the one that added the configurable workflow in 0.11, so I'm rather familiar with all those warts... and embarrased that there are so many remaininng.
[08:31:13.749330]<techtonik>Software is like humans - the more it grows, the more warts it gets.
[08:31:30.003921]<techtonik>@spelchek
[08:32:24.658729]<JosefAssad>retracile: quit being embarassed and start being proud. I really like it and a lot of people I know do too. And the BMC Remedy people are starting to get really defensive where I work which is a good sign
[08:32:59.238712]<techtonik>Is it possible to remove all the warts in 0.12? Is there a wart-list or something?
[08:33:19.234131]<retracile>JosefAssad: hehe :) It's always good to hear. I'm not familiar with BMC Remedy; what is it?
[08:33:26.630376]<evil_twin>t.e.o: Ticket #8694 (Fehler im Paket »update-manager«) created - <http://trac.edgewall.org/ticket/8694> - micha52@…
[08:33:37.218249]<JosefAssad>wart list: http://trac.edgewall.org/query?status=assigned&status=new&status=reopened&group=status&milestone=0.12
[08:34:12.581619]<retracile>techtonik: I don't currently have time available to start cutting on those warts, and probably won't between now and 0.12 (unless it takes as long as 0.10-0.11 did, which I _really_ hope it doesn't)
[08:34:22.451177]<JosefAssad>retracile: it's a proprietary behemoth which companies invest in when they acquire too many people with ITIL certs
[08:34:33.262259]<retracile>techtonik: life has gotten in the way, and stayed in the way for a while, sadly.
[08:34:53.321850]<retracile>JosefAssad: heh :) What does it claim to do? bug tracking or?
[08:34:55.762003]<JosefAssad>retracile: a REAL behemoth. There's a ganular license for everything. You can search in this list but you are missing a license to search in that one
[08:34:59.066920]<jhammel>JosefAssad: hah! i bet its one of the softwares that tells you how to work and not the other way?
[08:35:13.610625]<jhammel>retracile: http://www.bmc.com/products/offering/bmc-remedy-it-service-management-suite.html
[08:35:24.863376]<jhammel>retracile: though honestly i can't tell what it claims to do
[08:35:33.366114]<jhammel>maybe you install it and...you're done!
[08:35:40.379808]<JosefAssad>jhammel: I suppose that's how the BMC marketroids would describe it. From my observations, it mostly tells you which add-on licenses you need
[08:35:53.169800]<jhammel>JosefAssad: hah!
[08:36:03.438588]<JosefAssad>worst web UI I've ever seen, bar none. And I have seen some REALLY bad ones.
[08:36:18.557324]<retracile>JosefAssad: ever dealt with "Dimensions"?
[08:36:39.466722]<JosefAssad>retracile: nope. I'm not even in the service desk/support/operations business... I'm actually a suit. :)
[08:36:45.187578]<JosefAssad>what's dimensions?
[08:37:28.386259]<retracile>JosefAssad: it's supposed to be bug tracking, requirements tracking, source control, project management, all-in-one-million-dollar-thing
[08:37:55.055368]<jhammel>retracile: well, unless it can tell me the status of the coffee pot i see no reason for me to upgrade
[08:38:25.287073]<JosefAssad>retracile: oh that sounds lovely
[08:38:35.310095]<retracile>JosefAssad: really bad web UI, I'm not sure if it had a Linux client, and when the team I was on tried to do the demo of the source control side of things the ~10 simultanious checkouts of the source tree as step one failed
[08:38:49.117735]<JosefAssad>retracile: do you know any marketing lackeys I can throw 10 million dollars at? You've sold me on it
[08:38:50.510795]<jhammel>retracile: hah!
[08:39:05.471428]<retracile>JosefAssad: At which point, the team lead threw a fit, and we stuck with CVS.
[08:39:28.142118]<retracile>DESPITE the _rest_ of the company going to it, as I understand it.
[08:39:35.534195]<jhammel>retracile: that's what happened to IBM when they were demoing some proprietary (ironically) software monitoring software
[08:39:45.950738]<JosefAssad>retracile: linux client? You kid! That is really sophisticated. I'd show you a view source of even the simplest page in Remedy and you'd get carted off to hospital. 99% javascript. Probably > 99%
[08:40:53.054706]<JosefAssad>nothing wrong with CVS. Horses for courses
[08:41:03.857883]*JosefAssad likes low tech
[08:41:13.740139]<JosefAssad>Low tech doesn't break.
[08:41:27.088482]<retracile>I was trying to get the team to migrate to svn when I left -- had a nightly conversion of all CVS history -> subversion setup and working. Still didn't happen.
[08:42:01.422929]<JosefAssad>retracile: sometimes the only way to get new tools adopted is to hook yourself into another project in the planning phase
[08:42:15.465333]<retracile>JosefAssad: CVS was a solid workhorse; but it has a lot of warts. Subversion fixed most of them. It has its own, not-as-bad warts too of course.
[08:42:32.973808]*JosefAssad goes a-looking for a plugin that will tell you how long tickets have sat assigned to whom
[08:43:10.585516]<jhammel>retracile: not as bad but pretty bad; i use hg not bc i care about DVCS, but bc it fixes the svn warts. what i really want is svn that works well
[08:43:17.709437]<retracile>JosefAssad: I know a lot of people talk about the timingandestimation plugin, but I haven't used it.
[08:43:29.355693]<dunk_>!hg++
[08:43:32.211299]<dunk_>;]
[08:43:46.134602]<jhammel>dunk_: i can like, serve it with python and everything
[08:43:46.441705]<retracile>jhammel: yeah, I understand. But today, subversion is the "safe choice" in a corporate environment.
[08:44:00.574001]<JosefAssad>retracile: I looked at it but that isn't what it does. I just need to show that most of the time we're taking is happening in specific peoples' part of the workflow. :)
[08:44:04.543227]<dunk_>my bot uses hg so sync code
[08:44:10.609444]<jhammel>retracile: sure; we still have all of our stuff in a giant monolith of an svn repo
[08:44:20.201715]<JosefAssad>retracile: I'll probably just end up writing a query against sqlite
[08:44:20.581698]<retracile>JosefAssad: heh. :)
[08:44:47.986774]<jhammel>retracile: but for my own code i use hg, mainly because it was the only VCS i could figure out how to serve via python paste
[08:44:48.899084]<retracile>JosefAssad: yeah, probably.
[08:44:59.878249]<JosefAssad>the most dangerous kind of manager: the one who knows just enough to set up cool tools and use them to hold people accountable. :)
[08:45:08.329734]<jhammel>JosefAssad: heh
[08:45:09.411187]<retracile>hahaha! :)
[08:45:16.429545]*JosefAssad is notorious
[08:45:23.781808]<jhammel>"how's that git repository coming?"
[08:49:00.375514]<JosefAssad>judging from the crickets, I don't think it's coming along very well jhammel
[08:52:10.863579]<techtonik>@hg
[08:52:26.634776]<techtonik>@ewil_twin, where are you?
[08:52:26.652380]<evil_twin>nothing known about ewil_twin, where are you
[08:52:44.482537]<retracile>techtonik: evil_twin's not too bright
[08:52:47.726171]<retracile>evil_twin?
[08:53:00.537546]<dunk_>we should enable markov ;]
[08:53:01.899575]*retracile scowls at evil_twin
[08:53:13.838399]<retracile>dunk_: heh :)
[08:53:26.963140]<evil_twin>t.e.o: Ticket #8694 (Fehler im Paket »update-manager«) updated - <http://trac.edgewall.org/ticket/8694#comment:1> - rblank
[08:53:31.962283]<evil_twin>t.e.o: Ticket #8694 (Fehler im Paket »update-manager«) closed - <http://trac.edgewall.org/ticket/8694#comment:2> - rblank
[09:13:28.671094]<techtonik>Is it o.k. for a filter plugin to catch all exceptions? http://trac-hacks.org/browser/noanonymousplugin/0.11/noanonymous/filter.py
[09:17:40.594628]<jhammel>techtonik: it looks bad....i don't honestly know in that case
[09:22:07.715192]<ismael_>I cannot make a commit from the repository
[09:22:18.389711]<ismael_>svn co http://svn.edgewall.org/repos/trac/trunk trac
[09:22:18.690363]<ismael_>svn: Unrecognized URL scheme for 'http://svn.edgewall.org/repos/trac/trunk'
[09:22:35.234261]<ismael_>anyone knows why it can be?
[09:23:18.005218]<techtonik>Check that your svn works with other http sites. Can be proxy/filter/firewall.
[09:25:06.401469]<retracile>ismael_: unrecognized url scheme? did you build svn yourself or something?
[09:26:16.054809]<ismael_>ahh yes
[09:26:51.469786]<ismael_>i did it
[09:27:20.993108]<ismael_>I built it
[09:28:34.223764]<ismael_>I did it to build with the python bindings and integrate with subversion
[09:28:53.643709]<retracile>ismael_: what OS/distribution are you on?
[09:29:00.585804]<ismael_>ubuntu
[09:29:30.919065]<ismael_>9.04
[09:29:49.169822]<retracile>you should not need to build it yourself. apt-get install subversion will get you the client, etc. For the bindings, look for a package like 'python-subversion' or 'subversion-python'.
[09:30:49.081907]<ismael_>well, I wanted to use the same versions (of python, svn and swig) that the server has for which I am writting plugins
[09:31:02.418323]<jhammel>ismael_: you can always use gentoo ;)
[09:31:12.310088]<ismael_>why?
[09:31:16.378845]<ismael_>:D
[09:31:44.780280]<ismael_>I dont know exactly whats going on, and what can I do to download it
[09:31:48.506119]<retracile>ismael_: well, you built subversion wrong
[09:31:50.762009]<jhammel>ismael_: if you're that particular about having things hand compiled. i don't recommend it, but i do recommend it over building site packages yourself
[09:32:03.306119]<jhammel>(for reasons that now must seem obvious)
[09:32:04.117523]<ismael_>ahha
[09:32:53.246569]<ismael_>well, my subversion is working fine... it looks, I make commit, and I see it trough trac
[09:33:38.061503]<retracile>ismael_: but you can't checkout from any http:// subversion repository, can you?
[09:33:54.121179]<ismael_>let me check
[09:34:01.117386]<docdawning>I really could use some help please - I've done something that causes me to get the message "Invalid action "view"" when I try to modify a ticket. I definitely have permissions (at least via trac-admin) for that.
[09:34:24.657669]<retracile>docdawning: are you modifying your workflow?
[09:34:28.661123]<ismael_>no
[09:34:30.517788]<ismael_>I cant
[09:34:41.271773]<retracile>ismael_: that's because you built it wrong
[09:34:45.285118]<ismael_>ok
[09:35:00.509184]<ismael_>I didn't give me any error at time of building it
[09:35:00.625649]<docdawning>retracile: I don't know how to, but I had installed a bunch of plugins and since I've noticed this issue
[09:35:24.301787]<docdawning>retracile: I tried disabling all my plugins, but the issue remains. It feels like it's in the DB, but I really don't know.
[09:35:26.129814]<retracile>ismael_: it's at this point that I would recommend creating a VM with the same OS and everything as your target server for testing purposes.
[09:35:45.237799]<ismael_>aha
[09:35:56.302262]<docdawning>retracile: Also, the "Action" list when viewing a ticket (where you can change the status) only shows an empty box now.
[09:36:00.039036]<retracile>ismael_: (I'm not going to help you through building subversion etc.)
[09:36:12.321119]<ismael_>yeah, of course
[09:36:21.966891]<ismael_>dont feel bad ;DD
[09:36:39.887382]<ismael_>thank you very much
[09:37:37.278078]<retracile>docdawning: are you logged in as someone with TICKET_MODIFY permissions?
[09:38:08.086761]<docdawning>retracile: Yup
[09:38:13.605291]<docdawning>retracile: I'
[09:38:50.646839]<docdawning>retracile: I've since added all TICKET_ perms to this user in trying to figure it out. I've also tried removing all perms and adding them back one at a time and re-trying
[09:39:14.329109]<docdawning>The user also has "TRAC_ADMIN"
[09:39:17.493219]<retracile>docdawning: Is the action box empty when you login as a user with TRAC_ADMIN perms?
[09:39:22.217697]<retracile>heh. :)
[09:39:25.807986]<docdawning>lol
[09:39:39.250289]<retracile>Ok, TRAC_ADMIN is a super-set of all other perms, so if you have that one, you have them all.
[09:39:48.183321]<docdawning>retracile: It's a small trac deployment, but I'd really rather not remake it
[09:39:55.941223]<retracile>docdawning: that really sounds like you've got a messed up workflow config.
[09:40:13.133867]<docdawning>retracile: Well, I expected TRAC_ADMIN would imply everything, but I was getting desperate
[09:40:30.778509]<docdawning>retracile: Where/What do I look at to see my workflow config?
[09:40:41.218666]<jhammel>docdawning: trac.ini
[09:40:41.922563]<retracile>docdawning: understood. if you haven't done a backup of the environment recently, do one (standard advice ;) )
[09:40:46.990129]<jhammel>whereever the project is
[09:40:55.251231]<jhammel>(well, conf/trac.ini)
[09:41:10.530295]*jhammel wonders if there is a graphviz macro for displaying workflow
[09:41:17.346211]<retracile>docdawning: look through the trac.ini for an "sensitive" information, then pastebin it so we can look over it.
[09:41:27.482062]<docdawning>jhammel: If I post my trac.ini, would a more experienced trac admin be able to see the problem pretty easily?
[09:41:43.406119]<docdawning>retracile: You read my mind.. brb
[09:43:28.270176]<evil_twin>t.e.o: Ticket #8390 (Display "you may use wiki formatting" when creating a ticket) updated - <http://trac.edgewall.org/ticket/8390#comment:2> - idefix
[09:45:44.917747]<docdawning>retracile: Here we go: http://dpaste.com/97711/
[09:48:15.706722]<retracile>in [ticket-workflow] there are a few lines without anything after the '->'; those should have either '*' or a specific state
[09:48:54.500066]<retracile>(the '*' means leave it in the same state it was in; a state name means to transition the ticket to the specified state)
[09:49:32.565968]<retracile>docdawning: I think there's a tool for graphing the workflow in the contrib subdirectory of the trac sources
[09:50:08.157243]<techtonik>@req
[09:50:12.766823]<techtonik>@man req
[09:50:15.274342]<retracile>(note also that the various entries (such as 'resolve') are actions, and not ticket states.)
[09:51:14.082330]<docdawning>retracile: Thanks a ton, I'll peck at that a bit..
[09:51:16.162800]<techtonik>Where is API documentation for properties of `req` that is mentioned in API changes?
[09:51:58.767333]<techtonik>I mean is there any human readable and web-accessible docs?
[09:52:39.434804]<techtonik>@API req
[09:53:05.920886]<dunk_>@docs?
[09:53:07.386021]<evil_twin>nothing known about docs
[09:53:08.509274]<dunk_>hmm
[09:53:22.756636]<dunk_>@sd http
[09:54:08.801885]<dunk_>@wiki
[09:54:08.811260]<evil_twin>wiki <item>
[09:54:12.364632]<dunk_>hmm
[09:54:18.846129]<docdawning>retracile: An initial finding, I just blindly added '*' after each "->" in the ticket-workflow section, doesn't seem to have had an impact
[09:54:43.889442]<retracile>docdawning: did you restart the webserver? (just to be sure)
[09:55:01.386532]<docdawning>retracile: Thanks for checking, indeed, I did restart it.
[09:56:05.202392]<docdawning>retracile: Well, I restarted apache2. That should do it right? (That is what I'm running trac on).
[09:56:32.618172]<retracile>yeah, that should do it
[09:56:46.883409]<docdawning>Wsgi doesn't live outside of Apache I'd think.. In this context.. Total speculation on my part, but my XP seems to indicate that
[09:58:34.349429]<retracile>docdawning: that [components] section still shows a lot of plugins enabled. Strip that down to just account manager for your auth stuff and see if the behavior changes.
[10:00:37.320601]<retracile>docdawning: oh, and just for the record: what version of python, trac, and OS/distro?
[10:09:13.791041]<docdawning>retracile: Thanks for the help, I'll strip down those plugins further.. I'll get ya that other info in a sec
[10:10:17.426955]<docdawning>retracile: It's running 32-bit Ubuntu 9.04, Python 2.6.2, Trac 0.11.5
[10:18:29.037606]<evil_twin>t.e.o: Ticket #8695 (generate API documentation) created - <http://trac.edgewall.org/ticket/8695> - anatoly techtonik <techtonik@…>
[10:23:51.442916]<techtonik>Where can I see the definition of req.href.login()?
[10:25:08.083187]<techtonik>@href
[11:13:12.449608]<workPragmatist>Anyone around to help get post-commit-hook working...
[11:43:30.306495]<evil_twin>t.e.o: Ticket #1106 (Add the ability to rename wiki page.) updated - <http://trac.edgewall.org/ticket/1106#comment:112> - dispressa@…
[11:43:35.305614]<evil_twin>t.e.o: CookBook/About edited - <http://trac.edgewall.org/wiki/CookBook/About?version=7> - anatoly techtonik <techtonik@…>
[11:43:43.222231]<michalind>hello
[11:48:32.846792]<evil_twin>t.e.o: SeaChange/DocumentationRequests edited - <http://trac.edgewall.org/wiki/SeaChange/DocumentationRequests?version=4> - anatoly techtonik <techtonik@…>
[11:53:30.433774]<evil_twin>t.e.o: Ticket #1106 (Add the ability to rename wiki page.) updated - <http://trac.edgewall.org/ticket/1106#comment:113> - cboos
[11:56:05.111513]<workPragmatist>Can someone confirm two things for me... in order to get trac-post-commit-hook going I have two files... /var/svn/myproj/hooks/post-commit (rwx) ... and /var/trac/myproj/hooks/trac-post-commit-hook (rwx)
[11:57:22.601512]<workPragmatist>first: http://pastie.org/629372 second: http://trac.edgewall.org/browser/trunk/contrib/trac-post-commit-hook
[11:57:30.585309]<workPragmatist>is that all that needs to happen?
[11:57:43.795191]<workPragmatist>I confirm as /usr/bin/python exists
[11:58:30.633355]<evil_twin>t.e.o: SeaChange/DocumentationRequests edited - <http://trac.edgewall.org/wiki/SeaChange/DocumentationRequests?version=5> - anatoly techtonik <techtonik@…>
[12:03:22.944569]<workPragmatist>ideas?
[12:07:28.742188]<retracile>!inform workPragmatist svnbook?
[12:07:28.751778]<evil_twin>workPragmatist: http://svnbook.red-bean.com/nightly/en/index.html <-- The Subversion book
[12:07:47.363104]<retracile>workPragmatist: the post-commit script would need to call the trac-post-commit-hook
[12:07:58.158551]<workPragmatist>right
[12:08:03.915154]<retracile>workPragmatist: keep in mind that the environment variables are going to be pretty minimal when those run
[12:08:16.189430]<retracile>workPragmatist: things like $PATH may not have what you are used to
[12:10:54.246868]<workPragmatist>is maybe $TRAC_ENV not set?
[12:11:07.881823]<retracile>workPragmatist: if you haven't set it...
[12:11:13.229109]<workPragmatist>oh hrm
[12:19:03.234332]<workPragmatist>OH WAIT ... does the hook for svn have to have a #!bash\sh or whatever it is
[12:19:14.593942]<workPragmatist>sorry i'm not a script kiddie heh
[12:19:19.318500]<retracile>heh
[12:19:33.053138]<workPragmatist>#!/bin/sh
[12:20:00.342799]<retracile>the post-commit script must be executable, and it must be a valid script or whatever. So, yes, you'll need the #!/bin/bash or whatever
[12:20:06.906930]<workPragmatist>ya gotcah
[12:20:43.726775]<workPragmatist>why isn't REPOS used for http://trac.edgewall.org/browser/trunk/contrib/trac-post-commit-hook
[12:21:18.222344]<workPragmatist>or it is just setting it to be used in the next script?
[12:21:31.852240]<workPragmatist>instead of passing it (seems odd)
[12:22:33.357446]<retracile>the trac-post-commit-hook is given the path to the trac environment which knows where the svn repo is
[12:23:05.045289]<workPragmatist>so REPOS="$1" is not necessary
[12:25:15.403320]<retracile>strictly speaking, no. But it is there more to document the expected arguments being passed to the commit-hook.
[12:25:28.297564]<workPragmatist>I see
[12:26:01.074570]<workPragmatist>well i think i got it
[12:26:03.334317]<workPragmatist>lets see
[12:28:31.213814]<evil_twin>t.e.o: Ticket #8696 (diagrams for Trac request processing) created - <http://trac.edgewall.org/ticket/8696> - anatoly techtonik <techtonik@…>
[12:32:37.110603]<workPragmatist>retracile: yay! thank you
[12:32:48.774041]<retracile>workPragmatist: got it working?
[12:33:21.293708]<workPragmatist>yep!
[12:33:31.477428]<workPragmatist>no more jacking with tickets while working
[12:33:32.933887]<retracile>glad to hear it :)
[12:33:37.166484]<retracile>heh
[12:34:33.989988]<jhammel>now you just need to write the plugin that when closing a ticket commits the appropriate code to svn ;)
[12:34:50.714251]<jhammel>it would save a lot of time
[12:35:27.809215]<retracile>jhammel: have I described mergebot to you before? :)
[12:35:38.693176]<jhammel>retracile: no, does it solve that problem?
[12:37:16.419010]<retracile>jhammel: I do branch-based development... that is, each ticket has its own branch in svn. that adds some overhead, but mergebot takes care of it for you. When I have some time for it, I'm going to add the ability to have a ticket action called 'merge and close' that would merge the ticket's changes upstream, and close the ticket.
[12:38:31.493781]<evil_twin>t.e.o: Ticket #8697 (hg commit comments with text 'rDDD' where 'D' is a digit cause ...) created - <http://trac.edgewall.org/ticket/8697> - anonymous
[12:39:00.810394]<jhammel>retracile: but does mergebot generate the code from the ticket? or is that phase 2?
[12:43:31.678343]<evil_twin>t.e.o: SeaChange/DocumentationRequests edited - <http://trac.edgewall.org/wiki/SeaChange/DocumentationRequests?version=6> - anatoly techtonik <techtonik@…>
[12:44:21.671869]<retracile>jhammel: Oh, that'll have to be phase 2 ;)
[13:18:15.850834]<ronny>hi
[13:18:30.365347]<ronny>is there any way to add a ticket to more than one milestone?
[13:18:32.419536]<evil_twin>t.e.o: Changeset [8616]: #2647: replace the ''-- OR --'' text with a ''Or'' vertically centered in ... - <http://trac.edgewall.org/changeset/8616> - cboos
[13:19:10.126645]<retracile>ronny: nope. unless there is a plugin out there for it, but I don't know of one./
[13:19:32.450931]<jhammel>i would really like different "types" of milestones, where tickets could be added to one of each type
[13:19:44.445811]<jhammel>but...yeah, not a priority so i ain't gonna work on it
[13:19:49.558030]<retracile>jhammel: usecase?
[13:20:33.222399]<jhammel>retracile: my big one is to have, well, version-based milestones and then lets call the strategic milestones
[13:20:43.303651]<jhammel>and of course you could have, say, sprint milestones, etc
[13:20:50.393161]<jhammel>and a ticket could be part of all three of those
[13:21:14.699184]<jhammel>retracile: but yeah, its one of the most common requests from my users here
[13:22:14.865466]<techtonik>jhammel: Mergebot Phase 2 will emerge from rietveld after it gains some data statistically like Google translation did.
[13:22:58.475084]<techtonik>jhammel: tags for "milestone" types?
[13:22:59.998377]<ronny>retracile: well, we have versions, strategy targets and sprints
[13:23:16.962140]<retracile>techtonik: heh :)
[13:23:32.458062]<evil_twin>t.e.o: Ticket #2647 (Support AND/OR queries) updated - <http://trac.edgewall.org/ticket/2647#comment:30> - cboos
[13:23:47.702456]<ronny>retracile: and moving bugs back and forth isnt fun
[13:24:00.158688]<retracile>ronny: jhammel: sounds like Someone(TM) needs to write a plugin...
[13:24:00.987235]<ronny>jhammel: your idea is quite nice
[13:24:12.554096]<techtonik>That I really miss is +1/-1 digg-like comment filtering on especially long discussions.
[13:24:36.697663]<ronny>hmm, is there any good documentation on how to write trac plugins?
[13:25:27.074225]<techtonik>And ability to post filtered comment tread into another bug, or would it be called wave then?
[13:30:15.814515]<techtonik>@plugins
[13:30:54.430159]<techtonik>@TracPlugins
[13:31:19.813297]*techtonik feels evil_twin ignores him
[13:31:23.378256]<techtonik>http://trac.edgewall.org/wiki/TracPlugins
[13:32:28.862135]<retracile>!wiki TracPlugins
[13:32:28.877485]<evil_twin>http://trac.edgewall.org/wiki/TracPlugins
[13:32:34.383556]<retracile>techtonik: try that
[13:37:19.951629]<techtonik>!TracDev/Plugins
[13:37:46.767890]<techtonik>!wiki TracDev/API
[13:37:46.772835]<evil_twin>http://trac.edgewall.org/wiki/TracDev/API
[13:38:32.617852]<evil_twin>t.e.o: Ticket #8695 (generate API documentation) updated - <http://trac.edgewall.org/ticket/8695#comment:1> - cboos
[13:43:32.823111]<evil_twin>t.e.o: Ticket #8695 (generate API documentation) updated - <http://trac.edgewall.org/ticket/8695#comment:2> - anatoly techtonik <techtonik@…>
[13:43:37.826185]<evil_twin>t.e.o: Changeset [8617]: mimeview: associate `application/javascript` MIME-Type to `*.js` files, as ... - <http://trac.edgewall.org/changeset/8617> - cboos
[13:47:58.318647]<ronny>techtonik: so its practically underdocumented
[13:48:35.255776]<evil_twin>t.e.o: Ticket #8680 (trac uses uncommon mimetype for javascript) closed - <http://trac.edgewall.org/ticket/8680#comment:5> - cboos
[13:52:58.182327]<techtonik>ronny: right, most documentation is Trac code itself
[13:53:32.961661]<evil_twin>t.e.o: Changeset [8618]: Follow-up to r8577, don't output the tag for specifying the ... - <http://trac.edgewall.org/changeset/8618> - cboos
[13:53:37.961693]<evil_twin>t.e.o: Ticket #8644 (Missing character encoding in <html> tag) closed - <http://trac.edgewall.org/ticket/8644#comment:11> - cboos
[13:53:40.247639]<techtonik>it is also the easiest way to pick some plugin and start writing yours by example
[13:54:30.457321]<techtonik>but the other side is that even available docs are not concise enough to quickly put accents on some 'issues' plugin writers should be aware of
[13:55:51.567190]<techtonik>but this is again - the lack of some specific docs for plugin writers
[13:58:33.013441]<evil_twin>t.e.o: Ticket #2647 (Support AND/OR queries) updated - <http://trac.edgewall.org/ticket/2647#comment:31> - rblank
[14:00:58.771271]<workPragmatist>Is it possible to maybe put a timeline on a wiki page for a component?
[14:01:11.619351]<workPragmatist>of tickets closed / opened
[14:01:18.513762]<workPragmatist>similar to what is on the milestone page
[14:01:35.938847]<techtonik>A list of tickets, not sure about a timelien.
[14:01:50.094609]<techtonik>@TicketQuery
[14:01:55.986716]<workPragmatist>okay that works
[14:02:22.327926]<techtonik>!wiki TicketQuery
[14:02:22.339527]<evil_twin>http://trac.edgewall.org/wiki/TicketQuery
[14:13:13.505726]<robgleeson>Hi
[14:13:31.910960]<robgleeson>is it possible to remove the wiki altogether, and have markdown parsed into HTML?
[14:14:17.521850]<workPragmatist>techtonik: do I need to quote milestones that have spaces?
[14:14:29.186056]<workPragmatist>I have [[TicketQuery(milestone=Milestone: 2.0.0 (Start)]] that returns No Results
[14:14:43.881713]<workPragmatist>oh wait i'm on crack lol
[14:15:04.169918]<workPragmatist>well still... do i need to quote it?
[14:15:21.003545]<workPragmatist>or escape maybe
[14:17:12.476494]<workPragmatist>maybe i need to use a saved query instead?
[14:29:02.859030]<techtonik>workPragmatist: you need to escape it. Probably like in URL. It can help building a custom report query and look at URL.
[14:29:15.734007]<workPragmatist>i got it
[15:58:35.033790]<evil_twin>t.e.o: Ticket #8698 (OperationalError: database is locked) created - <http://trac.edgewall.org/ticket/8698> - mvannorden
[16:03:35.037874]<evil_twin>t.e.o: Ticket #8698 (OperationalError: database is locked) closed - <http://trac.edgewall.org/ticket/8698#comment:1> - eblot
[17:36:48.810878]<doki_pen>how do I make a fake request?
[17:37:00.977296]<doki_pen>fake request object, that is
[19:16:50.795130]<YGingras>can I set ticket dependency in 0.11?
[19:17:18.595176]<YGingras>0.11.5 as a matter of fact
[19:31:27.099576]<jlf>hi folks, i just installed trac and did "trac-admin /Users/jlf/trac-project initenv", got through the prompts and saw "NameError: global name 'sqlite' is not defined". sqlite --version reports 3.4.0 and i downloaded and installed (with sudo python setup.py install) pysqlite 2.5.5.
[19:31:31.370625]<jlf>anyone have a hint?
[19:35:20.547171]<YGingras>don't you need sqlite3 for trac?
[19:35:36.831092]<jlf>i have 3.4.0
[19:35:37.847382]<YGingras>oh no, I have 2.4.1
[19:35:49.161124]<YGingras>I mean pysqlite
[19:36:34.557730]<jlf>hmm, >>> import sqlite --> ImportError: No module named sqlite
[19:36:54.914325]<jlf>so evidently my sudo command above installed it someplace python isn't looking
[19:39:11.255902]<YGingras>jlf: you should "import sqlite3"
[19:41:20.162687]<jlf>i did "import sqlite" based on http://trac.edgewall.org/wiki/PySqlite#Troubleshooting
[19:42:46.026439]<YGingras>jlf: I don't know what I'm talking about, feel free to ignore me
[19:42:49.756792]<jlf>but import sqlite3 works, and >>> sqlite3.version ---> '2.3.2'
[19:43:08.922090]<jlf>no, you're right.. looks like a typo or stale info on that pag
[19:43:12.329854]<jlf>s/pag/page/
[19:43:12.343337]<evil_twin>jlf meant: no, you're right.. looks like a typo or stale info on that page
[19:45:34.291226]<jlf>i don't know where that 2.3.2 comes from, since sqlite is 3.4.0 and pysqlite is 2.5.5
[19:46:14.855145]<YGingras>some fuck up in your path
[19:46:37.238302]<YGingras>try sqlite3.__file__
[19:47:49.261438]<jlf>'/System/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/sqlite3/__init__.py'
[19:51:23.242344]<YGingras>do you know where you're manually installed version ended up?
[19:51:37.238865]<lisppaste5>jlf pasted "apparent sqlite version mismatch" at http://paste.lisp.org/display/87689
[19:51:51.858090]<YGingras>it's probably shadowed by this one. I'm not too familiar with the Mac enviroment
[19:52:35.191277]<jlf>it seems to be in /Library/Python/2.5/site-packages/pysqlite2
[19:52:50.345154]<jlf>not sure how to set the python library path though
[19:55:13.279131]<YGingras>sys.path
[19:55:26.549806]<YGingras>prepend before your import to change the order
[19:55:59.998875]<YGingras>or the $PYTHON_PATH shell variable
[19:56:13.477606]<YGingras>same syntax as $PATH
[20:03:36.585839]<jlf>ah, that did it - thanks!
[20:36:34.363669]<jlf>ok, another question: i'd like to require authentication before users are allowed to see _anything_ (bugs, wiki entries, etc.). is this possible with the built-in server or do i need to control access through apache?
[20:47:25.810920]<YGingras>jlf: you can remove all the *_VIEW permissions in the admin panel
[20:47:43.208223]<YGingras>but it's probably easier to do it with apache
[20:47:55.273742]<YGingras>I mean, it's less clicks
[20:55:27.438457]<jlf>YGingras: thanks
[22:38:41.301785]<evil_twin>t.e.o: Ticket #8698 (OperationalError: database is locked) updated - <http://trac.edgewall.org/ticket/8698#comment:2> - cboos
[22:43:41.310324]<evil_twin>t.e.o: MostFrequentDuplicates edited - <http://trac.edgewall.org/wiki/MostFrequentDuplicates?version=8> - cboos
[22:43:46.309724]<evil_twin>t.e.o: Ticket #2647 (Support AND/OR queries) updated - <http://trac.edgewall.org/ticket/2647#comment:32> - cboos