Team Chat Logs

2006 4
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 31        

May 13, 2006

[00:23:33] * _idostyle has joined #trac
[00:39:41] * idostyle has quit IRC
[00:39:52] * _idostyle is now known as idostyle
[01:07:23] * Semhirage has quit IRC
[01:19:28] * jjb_lyceum has quit IRC
[01:45:49] * werneck has joined #trac
[02:22:56] * jjb_lyceum has joined #trac
[02:28:38] * perchr has quit IRC
[02:32:21] * MasterC has joined #trac
[02:47:20] * vmx_ has joined #trac
[02:49:48] * vmx has quit IRC
[03:30:14] * tenshiKur0 has joined #trac
[03:39:13] * jborg has joined #trac
[03:51:55] * tuxipuxi has joined #trac
[04:13:51] * jjb_lyceum has quit IRC
[04:30:56] * tenshiKur0 has quit IRC
[05:02:14] * tic_ has joined #trac
[05:09:16] * jtoy has joined #trac
[05:12:04] * tic_ has quit IRC
[05:14:58] * cboos has joined #trac
[05:15:42] <cboos> hi,
[05:16:40] <cboos> have you seen the tickets 3145 and 3146?
[05:16:54] <cboos> they are heavily loaded with spam ...
[05:17:05] <cboos> summary description and all fields...
[05:17:46] <cboos> strange that they were not caught by akismet
[05:24:29] * Riccc`Miam has joined #trac
[05:36:08] * thm has joined #trac
[05:37:20] <thm> hi
[05:40:03] * jtoy has quit IRC
[05:42:09] * meldor has joined #trac
[05:42:09] * Riccc has quit IRC
[05:42:36] <meldor> Hello
[05:43:17] <meldor> anyone here ?
[05:44:01] * imajes has joined #trac
[05:44:43] <meldor> looking for help to install trac, got a trouble with subversion :S
[05:45:12] * tic_ has joined #trac
[05:50:55] * Semhirage has joined #trac
[05:52:47] * meldor has quit IRC
[06:45:23] <tuxipuxi> [Sat May 13 15:57:57 2006] [notice] child pid 29502 exit signal Segmentation fault (11) can i backtrace this somehow?
[06:49:53] <tuxipuxi> hm must be something related to mod_python.. works in tracd
[07:15:28] * eblot has joined #trac
[07:19:27] <eblot> Hi all,
[07:19:46] <eblot> Is there any news about the IUserDirectory interface ?
[07:49:55] <thm> can someone point me to a doc about IConfigurable or how to get configuration options for extensions properly?
[08:04:40] * Riccc has joined #trac
[08:06:34] * eevar2_ has joined #trac
[08:07:38] * eevar2 has quit IRC
[08:18:24] <coderanger_> thm: The only docs for that right now are the code
[08:18:48] <coderanger_> pacopablo: No, the main drive died
[08:19:04] <cboos> thm: + the IConfigurable had a very transient life ;)
[08:19:20] <thm> cboos: yeah, seems so
[08:19:38] <cboos> thm: there's now a different way to do this,
[08:20:04] <cboos> http://projects.edgewall.com/trac/search?q=IConfigurable&changeset=on
[08:23:13] * Riccc`Miam has quit IRC
[08:26:22] <thm> cboos: thx, will look into this
[08:47:05] * cboos has quit IRC
[08:47:19] <eblot> cboos: quick question about UTF-8 support: it seems the UTF-8 source code as shown in [3293] does not look like as expected. The unix "file" tool reports: 'UTF-8 Unicode English text' for trac/ticket/tests/notification.py. Is this an known issue, or did I do something wrong with this file ?
[09:03:08] * MasterC has quit IRC
[09:13:35] * coderanger_ has quit IRC
[09:25:16] * tiagovaz has joined #trac
[10:04:35] * tic_ has quit IRC
[10:30:31] * MasterC has joined #trac
[10:41:08] * macem has joined #trac
[11:11:13] * chandlerc has quit IRC
[11:12:03] * macem has quit IRC
[11:12:17] * chandlerc has joined #trac
[11:14:17] * chandlerc has joined #trac
[11:19:48] <thm> cboos: there's a small bug in the doc for trac.versioncontrol.Changeset.get_changes():
[11:19:48] <thm> base_rev and base_path need to be swapped
[11:22:22] * jjb_lyceum has joined #trac
[11:22:59] * tic_ has joined #trac
[11:26:24] * chandlerc has quit IRC
[11:27:25] * chandlerc has joined #trac
[11:29:04] * tic has quit IRC
[11:35:11] <thm> is there an archive/log of this irc available anywhere?
[11:49:05] * robs0r has joined #trac
[12:04:44] * robs has quit IRC
[12:04:44] * robs0r is now known as robs
[12:09:55] * tag has quit IRC
[12:21:29] * oferw has joined #trac
[12:23:10] <tuxipuxi> anybody an idea how edgewall embeds the trac trac project into projects.edgewall.com?
[12:28:41] * tiagovaz has quit IRC
[12:44:19] * prisoner has left #trac
[12:45:21] * BaconTastesGood has joined #trac
[12:45:40] <BaconTastesGood> anyone know if there are plans for adding mathematical markup to the wiki language in trac?
[12:45:43] <BaconTastesGood> and anyone know of a good trac host?
[12:46:27] <tuxipuxi> i haven't heard of such plans, maybe you should have a look at the open tickets.
[12:46:37] <tuxipuxi> hosted-projects.com is a good host
[12:48:15] <BaconTastesGood> wow, their prices are real good
[12:48:42] <tuxipuxi> yeah and an unlimited number of projects included in all accounts
[12:49:19] <dilinger> BaconTastesGood: heh, is your nick taken from Pulp Fiction?
[12:49:35] <BaconTastesGood> dilinger: pork chops taste good =)
[12:49:45] <dilinger> hehe, cool
[12:50:01] <dilinger> yea, i read your nick and that quote popped in my head
[12:50:12] <dilinger> "sewer rat may taste like pumpkin pie..."
[12:51:19] <BaconTastesGood> heh
[12:51:27] <BaconTastesGood> so anyone know off-hand how up to date hosted-projects keeps up with trac?
[12:54:02] <tuxipuxi> BaconTastesGood: always the latest version
[12:54:19] <BaconTastesGood> cool. Are their servers in the USA or in Germany? and do they support PHp as well?
[12:54:29] <BaconTastesGood> I'd LOVE to convert away from my current provider, who is more expensive and overall much crappier.
[12:54:41] <BaconTastesGood> but since moving is so difficult, I'd like to know I actually chose the right company this time =)
[12:55:23] <tuxipuxi> they are in germany, but no php support. which is your current host?
[12:55:32] <BaconTastesGood> aplus.net
[12:55:36] <BaconTastesGood> they don't offer anything =(
[12:55:38] <BaconTastesGood> no python, nothing
[12:55:40] <BaconTastesGood> crappy control panel
[12:55:47] <BaconTastesGood> $10/month, and they have outages about 3x a year
[12:55:59] * cboos has joined #trac
[12:56:53] <tuxipuxi> ah.. looks like a more general web hosting company
[12:56:59] <BaconTastesGood> yeah
[12:57:05] <BaconTastesGood> I need PHP so I can install phpBB
[12:57:10] <BaconTastesGood> I have an existing phpBB install that needs to migrate
[12:57:56] <tuxipuxi> are you sure that you would like to have your repositories served by the same server as a phpBB installation? :)
[12:58:00] <dilinger> BaconTastesGood: unixshell.com
[12:58:11] <BaconTastesGood> here's the thing -- no repos
[12:58:17] <BaconTastesGood> I'm going to use Trac to replace a blog like thing
[12:58:21] <dilinger> i'm paying $12/mo for a xen instance w/ 4.5gb of space
[12:58:25] <dilinger> i'm quite happy w/ it
[12:58:31] <tuxipuxi> ah i see
[12:58:45] <BaconTastesGood> unixshell.com has suspended ordering
[12:58:49] <dilinger> oh
[12:58:57] <dilinger> oh well :)
[12:59:00] <BaconTastesGood> heh
[13:10:55] * Marcel has quit IRC
[13:11:24] <BaconTastesGood> well, railsplayground.com looks pretty good, ironically enough
[13:24:59] * cboos has quit IRC
[13:27:07] * tiagovaz has joined #trac
[13:51:40] * tiagovaz has quit IRC
[13:53:10] * cboos has joined #trac
[13:56:40] <thm> cboos: there's a small bug in the doc for trac.versioncontrol.Changeset.get_changes() base_rev and base_path need to be swapped
[13:56:56] <cboos> thm: ack, I'll have a look
[13:57:39] <cboos> correct :)
[14:00:43] <thm> cboos: in a versioncontrol plugin, I can't overwrite [revid] wiki-syntax :-(
[14:01:01] <cboos> why/how would you need that?
[14:01:18] <cboos> you can already use hexa ids... do you need more?
[14:02:51] <thm> yeah, but the changeset message is not looked up properly
[14:03:06] * MasterC has quit IRC
[14:03:26] <thm> anyway, the concrete format of a revid should be defined by the vc plugin
[14:03:32] <thm> resp. backend
[14:05:11] <cboos> I had the same problem with the mercurial plugin... my initial approach was to subclass the ChangesetModule but right now, it is generic enough for at least the needs of Mercurial. That's why I was wondering what you would need...
[14:06:00] * tic_ has quit IRC
[14:06:03] <cboos> You also have to be careful to not conflict with the long form of TracLinks: [target label]
[14:06:23] <thm> My mtn plugin indeed has some similarities to your hg plugin ;-)
[14:06:46] <thm> btw, I can't overwrite [changeset: revid] either
[14:08:14] <cboos> yeah, same issue. If you look back in the vc-refactoring branch, you could find that alternative approach. The subclasses ChangesetModule was quite small, and included such things like overriding the changeset: link resolver. ..
[14:08:25] <thm> and the rRevid syntax doesn't support hexa ids, but that's easy
[14:08:52] <cboos> really? what about the race condition? ;)
[14:08:52] * tic has joined #trac
[14:09:06] <thm> which one?
[14:09:13] <cboos> rACE ;)
[14:09:37] <thm> no, only ids of len 40
[14:10:06] <cboos> ah! ok... mercurial supports shortened ids, that's why I didn't add it ...
[14:11:00] * stevegt has quit IRC
[14:11:22] <cboos> but yeah if you go for a long enough id, it should be possible to add it. Even 10 or even 8 would probably be OK.
[14:11:33] <thm> but, mtn has a mechanism called selectors, for selecting revisions via shortened ids, tag names, branch heads, dates and so on, and it would be nice to expose that to users
[14:12:06] <thm> but for that, I would have to catch other forms of revids
[14:13:18] <cboos> I guess you've seen that I added 'branch:' and 'tag:' ... can't you do someting similar, even 'select:' if that term is common for Monotone users?
[14:13:38] <thm> I did something like that already...
[14:14:36] <thm> select: is missing though, good idea
[14:15:31] <thm> I have some more q's, do you have some time?
[14:16:22] <cboos> well, a bit yes,
[14:17:15] <cboos> so please ask
[14:18:28] <thm> Repository.short_rev is only used for the diffs, why?
[14:18:54] <thm> in the hg plugin, you "cheated" somehow by only using short revids everywhere
[14:19:56] <thm> I would prefer to use the complete revids and only display short versions in the ui
[14:21:53] <thm> 40byte are much :)
[14:22:58] <cboos> yes, I understand.
[14:23:26] * ValiSystM has joined #trac
[14:24:31] <cboos> iiuc, you have only the 40bytes ids and a shortrev?
[14:26:03] <cboos> mercurial has the shortrev, shortened ids (12bytes) and the full one (also 40bytes)
[14:26:17] <cboos> so you miss the intermediate form, right?
[14:28:05] <thm> no, I have only 40 byte revs in the first place
[14:28:42] <thm> but, I want to display shorter revs
[14:29:21] <cboos> how? You mean, truncated ids?
[14:29:27] <thm> additionally, yes
[14:29:36] <thm> ah no
[14:30:12] <thm> sorry. I want to display shortened revids for example in the changelog
[14:31:54] <cboos> the ui should ask the repository to give the revision in some format: short/compact/full for example
[14:32:19] <cboos> then each api would know how to produce such formats: for svn, all the 3 are the same.
[14:32:26] <thm> yes, that what I thought short_rev was intended for
[14:33:42] <cboos> in mercurial, short would be the sequence number (and that's the current short_rev), the compact form is the 12bytes format (and that could be compact_rev) and the full is the 40bytes format (that could be full_rev)
[14:34:15] <cboos> and for monotone, you could implement short_rev/compact_rev by truncating to the appropriate length the 40bytes ids.
[14:34:38] <thm> yep
[14:35:33] <cboos> ok, I'm taking note of this ;)
[14:35:34] <thm> but it should not be used for any other purpose than displaying
[14:36:19] <cboos> maybe an api like: Repository.format_rev(SHORT/COMPACT/FULL)
[14:36:57] <cboos> format_rev(rev, format=SHORT|COMPACT|FULL)
[14:37:14] <thm> why not three methods? saves two if's
[14:37:43] <cboos> or maybe create an object to represent revisions... would be handy too :)
[14:38:26] <thm> but there is already the changeset
[14:38:50] <cboos> yes, but the changeset is "heavy" to instantiate... no only a kind of handle to represent revision numbers
[14:39:17] <cboos> you know, we already have this kind of "normalize_rev" thing,
[14:39:36] <cboos> because from the UI a revision is a string, and in svn the revision is expected to be an int
[14:39:44] <thm> of which I didn't the real purpose...
[14:40:31] <thm> (insert get)
[14:41:07] <cboos> "normalize_rev" is meant to take a string representing a revision number from the UI, and convert it to a form understandable by the backend (an "int" in the case of Subversion)
[14:41:56] <thm> does it have to convert None to the youngest_rev or not?
[14:42:23] <cboos> no, it shouldn't.
[14:42:43] <cboos> It used to, but now "None" also has a special meaning in the UI:
[14:43:00] <cboos> there it means "stay on the youngest revision"
[14:43:45] <thm> so for mtn, it could simply pass everything through
[14:44:19] <cboos> maybe... or you could use it to convert it to canonical 40bytes ids
[14:44:49] <thm> i.e. do the select here, that's what I do currently
[14:46:38] <thm> ok. will 3017 be merged to trunk?
[14:47:47] <thm> it seems to have a bug anyway, because jumping to tags didn't work when I tried the vc-ref branch
[14:48:07] <thm> (besides from the other problems caused by the unicode change)
[14:49:40] <cboos> mh, I think it worked for me, but to tell the truth, it's been quite a few weeks I was not able to check that branch ...
[14:51:09] <cboos> I intend to resume work on this asap, to start to implement the new cache, so I'll take a look at that feature before.
[14:51:39] * oferw has quit IRC
[14:51:58] * tiagovaz has joined #trac
[14:52:16] <cboos> But to answer your question, no, I don't think it will make it for trunk for the 0.10 release.
[14:55:05] <thm> Now the biggest problem: implementing Node.get_history...
[14:55:24] <thm> because it assumes a single line of history
[14:56:31] <thm> which in monotone need not be the case
[14:57:18] * eblot has quit IRC
[14:57:34] <cboos> I understand ... but isn't there something that can *fake* this linearity? Like the sequence in time?
[14:57:39] <cboos> What I mean is this:
[14:58:20] <cboos> the whole Log module and the related api (Node.get_history and friends) is simply not adapted to the SCM where the history is a DAG.
[14:58:21] <thm> yeah, but the renaming detection will not work anymore
[14:59:45] <cboos> So the point would be to come with a different API, and a different module for these SCM. Something similar to Gitk (maybe there's such a GUI for Monotone already...)
[15:01:13] <thm> on the other hand it may be worth doing that right once, because linear history is a special case of a dag
[15:01:53] <cboos> Well, sure, but it's not only that.
[15:03:12] <cboos> In Svn, you have the concept of branching by copy, which is why you have this special display for the copy/rename operations, why you have the stop on copy checkbox, .. .Not something that you can transpose to the other SCMs ... or?
[15:04:45] <cboos> I'm don't thing we can at the same time support the "branch by copy" philosophy and the "branch by multiple children revisions" philosophy, in the same UI.
[15:04:53] <cboos> s/I'm/I/
[15:05:44] <cboos> ... but of course I'll be pleased if you come with a better idea :)
[15:07:00] <thm> Well, for a short time before starting with the plugin, I though of (virtually) converting the branches into toplevel dirs
[15:07:44] <thm> but then, a revision can be part of multiple branches...
[15:07:58] <cboos> ... and that would not solve the pb for micro-branches (if I undestood correctly what micro branches are...)
[15:08:14] <cboos> but certainly I didn't :)
[15:09:29] <thm> one could invent sub-branchnames for the micro-branches, but I thought something like that is a really bad idea
[15:10:35] <thm> micro-branch means that there is more than one head revision for a branch
[15:10:55] <thm> i.e. the history for a branch itself is a dag, too
[15:11:16] <cboos> In monotone, a branch is a top-level, named "fork" of the initial(?) changeset, while a micro-branch is an anonymous "fork" of any changeset within a branch... is it like that?
[15:13:04] <thm> the latter seems to be right, while the first is wrong. there might be multiple independent roots in a monotone db
[15:14:46] <cboos> yeah, that's why I was not confident about "initial". So different branches may have nothing in common...
[15:15:08] <thm> yes
[15:16:21] <cboos> Overall, it's not that different from Mercurial. In mercurial, you have a "magic" null revision, which is the common ancestor of any branch ...
[15:16:45] <cboos> any new repository I mean
[15:17:37] <cboos> so in effect you can also have two unrelated branches in the same repository.
[15:18:33] * ged has joined #trac
[15:19:47] <thm> some mtn users complained about my plugin not 'handling' branches
[15:20:24] <thm> whatever handling means. restricting the timeline to certain branches on demand would be a first step
[15:20:32] * ValiSystM has quit IRC
[15:21:07] <cboos> yes, there are already a few tickets about extending the timeline filters...
[15:22:08] <cboos> thm: sorry, but I think I'll have to go now, was a real pleasure talking to you!
[15:22:33] <thm> ok, cu later
[15:22:47] <cboos> In general, don't hesitate to push your questions on Trac-Dev/the wiki, I'm there much more often than on IRC :)
[15:22:55] <cboos> bye
[15:23:07] * cboos has quit IRC
[15:25:15] * chandlerc has quit IRC
[15:33:37] * ValiSystM has joined #trac
[15:38:09] * tiagovaz has quit IRC
[15:46:46] * thm has left #trac
[15:54:32] * jborg has quit IRC
[15:56:13] * fholmes has joined #trac
[15:57:32] <fholmes> I am trying to install trac and following the install guide for Ubuntu and I am having problems getting SVN to work properly
[15:58:03] <fholmes> I don't know if it is a problem with Apache2 or my DNS configuration.
[16:44:56] * imajes has quit IRC
[16:46:04] * tic has quit IRC
[16:51:10] <BaconTastesGood> well...this is weird
[16:51:18] <BaconTastesGood> how do I log out of trac completely?
[16:51:28] <BaconTastesGood> if I logout, then click 'login' it logs me back in immediately
[16:51:30] <BaconTastesGood> no password, username
[16:51:34] <BaconTastesGood> just right back in as who I was
[17:02:07] * fholmes has quit IRC
[17:04:51] <tuxipuxi> long-standing problem
[17:04:56] <tuxipuxi> there's a ticket about it
[17:04:58] <BaconTastesGood> oh
[17:05:01] <BaconTastesGood> well, okay
[17:05:04] * chandlerc has joined #trac
[17:05:04] <BaconTastesGood> that's not good =)
[17:07:11] <BaconTastesGood> is there a workaround that doesn't involve exiting the browser?
[17:07:34] <tuxipuxi> restart the computer
[17:07:36] <tuxipuxi> =)
[17:08:11] <BaconTastesGood> ugh...seriously
[17:08:25] <tuxipuxi> none that i know of
[17:09:35] <tuxipuxi> http://projects.edgewall.com/trac/ticket/791
[17:10:01] <BaconTastesGood> holy crap, that's a major issue
[17:10:06] <tuxipuxi> why?
[17:10:14] <BaconTastesGood> well, multiple users on a single system for one
[17:10:33] <tuxipuxi> true, but that's rather unusual for serious use, no? :)
[17:11:23] <BaconTastesGood> um, not really. public access machines, shared computers, etc. As a general security thing it's considered REAL BAD.
[17:11:36] <BaconTastesGood> Like, hey, if I login to my bank account and can't logout, even if it's on my computer, that's probably not good =)
[17:12:19] <tuxipuxi> public machines.. true.. but i would close the browser there anyway. shared computers with multiple trac users? i'd say that's unusual :)
[17:12:39] <tuxipuxi> ok i'm paranoid, i'd never enter a password on a public machine anyway ;)
[17:21:50] * ValiSystM has quit IRC
[17:23:28] * fholmes has joined #trac
[17:23:43] * tuxipuxi has quit IRC
[17:23:53] <fholmes> I am installing trac on Ubuntu and I keep getting the following error message:
[17:24:07] <fholmes> PythonHandler trac.web.modpython_frontend: ImportError: No module named web
[17:24:12] <fholmes> Anyone have any ideas?
[17:33:12] * fholmes has left #trac
[17:44:03] * Sage has joined #trac
[17:44:36] * Sage has left #trac
[17:58:01] * imajes has joined #trac
[17:59:15] * werneck has left #trac
[18:05:43] * imajes has quit IRC
[18:27:53] * cmlenz has quit IRC
[18:28:16] * stevegt has joined #trac
[18:47:51] * coderanger_ has joined #trac
[19:13:50] * Standfast has joined #trac
[19:15:45] * s0undt3ch has joined #trac
[19:51:20] * Arc has left #Trac
[20:37:33] * whitelynx|firest has quit IRC
[20:52:24] * vmx_ has quit IRC
[20:54:41] * Standfast has quit IRC
[20:58:01] * whitelynx|firest has joined #trac
[21:12:21] * contingencyplan has joined #trac
[21:19:00] * chandlerc has quit IRC
[21:21:08] * whitelynx|laptop has joined #trac
[21:21:19] * whitelynx|firest has quit IRC
[21:27:02] * tic has joined #trac
[21:27:27] * BaconTastesGood has left #trac
[21:51:46] * jjb_lyceum has quit IRC
[21:58:56] * chandlerc has joined #trac
[22:01:27] * stevegt has quit IRC
[22:01:46] * jjb_lyceum has joined #trac
[22:37:23] * stevegt has joined #trac
[22:41:35] * jjb_lyceum has quit IRC
[22:42:46] <pacopablo> evening all
[23:00:40] <coderanger_> yo
[23:59:39] * tic has quit IRC