Team Chat Logs

July 31, 2009

2009 6
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    
<< <   > >>

[01:03:23]<evil_twin>t.e.o: Ticket #8536 (Don't display Default Repository on main Browse page) updated - <http://trac.edgewall.org/ticket/8536#comment:3> - rblank
[01:06:45]<sanmarcos>how can I change the footer?
[01:06:49]<sanmarcos>where the Trac Image is?
[01:36:18]<thm>can I "lock" a ticket somehow? i.e. tell people: please don't change something, I'm working on it. or is that what "accepted" is for?
[01:36:37]<sanmarcos>i guess
[01:36:45]<sanmarcos>probably write a plugin
[01:37:18]<thm>standard answer.
[01:43:26]<evil_twin>t.e.o: Ticket #8536 (Don't display Default Repository on main Browse page) updated - <http://trac.edgewall.org/ticket/8536#comment:4> - cboos
[01:48:25]<evil_twin>t.e.o: Ticket #8536 (Don't display Default Repository on main Browse page) updated - <http://trac.edgewall.org/ticket/8536#comment:5> - cboos
[02:28:25]<evil_twin>t.e.o: TracUsers edited - <http://trac.edgewall.org/wiki/TracUsers?version=858> - vasili@…
[02:33:26]<evil_twin>t.e.o: Ticket #8536 (Don't display Default Repository on main Browse page) updated - <http://trac.edgewall.org/ticket/8536#comment:6> - rblank
[02:43:26]<evil_twin>t.e.o: Ticket #8536 (Don't display Default Repository on main Browse page) updated - <http://trac.edgewall.org/ticket/8536#comment:7> - cboos
[02:43:31]<evil_twin>t.e.o: Ticket #2471 (The logo in the left corner should not be named "banner") closed - <http://trac.edgewall.org/ticket/2471#comment:6> - cboos
[03:03:29]<evil_twin>t.e.o: Ticket #454 ([patch]Edit ticket comments) updated - <http://trac.edgewall.org/ticket/454#comment:161> - rupert.thurner
[03:03:34]<evil_twin>t.e.o: Changeset [8429]: 0.12dev: Update the wiki page text in the model ''before'' validation, so ... - <http://trac.edgewall.org/changeset/8429> - rblank
[03:03:39]<evil_twin>t.e.o: Ticket #7731 (IWikiPageManipulator.validate_wiki_page not called with the correct data) updated - <http://trac.edgewall.org/ticket/7731#comment:10> - rblank
[03:03:44]<evil_twin>t.e.o: InterMapTxt edited - <http://trac.edgewall.org/wiki/InterMapTxt?version=27> - cboos
[03:08:27]<evil_twin>t.e.o: setuptools edited - <http://trac.edgewall.org/wiki/setuptools?version=2> - cboos
[03:13:26]<evil_twin>t.e.o: TracDev/ApiChanges/0.12 edited - <http://trac.edgewall.org/wiki/TracDev/ApiChanges/0.12?version=4> - rblank
[03:13:31]<evil_twin>t.e.o: Ticket #7731 (IWikiPageManipulator.validate_wiki_page not called with the correct data) closed - <http://trac.edgewall.org/ticket/7731#comment:11> - rblank
[03:13:36]<evil_twin>t.e.o: InterMapTxt edited - <http://trac.edgewall.org/wiki/InterMapTxt?version=28> - cboos
[03:18:27]<evil_twin>t.e.o: Ticket #8459 (svn:mergeinfo rendering is far too slow) updated - <http://trac.edgewall.org/ticket/8459#comment:20> - rupert.thurner@…
[03:33:27]<evil_twin>t.e.o: Ticket #8459 (svn:mergeinfo rendering is far too slow) updated - <http://trac.edgewall.org/ticket/8459#comment:21> - eblot
[03:43:28]<evil_twin>t.e.o: Ticket #8536 (Don't display Default Repository on main Browse page) updated - <http://trac.edgewall.org/ticket/8536#comment:8> - rblank
[03:53:28]<evil_twin>t.e.o: Changeset [8430]: 0.12dev: Fixed a regression on Python 2.4 introduced in [8421] (missing ... - <http://trac.edgewall.org/changeset/8430> - rblank
[04:03:29]<evil_twin>t.e.o: Changeset [8431]: MultiRepos: Add the repository URL to the repository metadata and display ... - <http://trac.edgewall.org/changeset/8431> - rblank
[04:03:34]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) closed - <http://trac.edgewall.org/ticket/1781#comment:17> - rblank
[04:23:29]<evil_twin>t.e.o: Ticket #8536 (Don't display Default Repository on main Browse page) updated - <http://trac.edgewall.org/ticket/8536#comment:9> - cboos
[04:23:34]<evil_twin>t.e.o: Ticket #8459 (svn:mergeinfo rendering is far too slow) updated - <http://trac.edgewall.org/ticket/8459#comment:22> - cboos
[05:03:29]<evil_twin>t.e.o: Ticket #8483 (UndefinedError: "dateinfo_delta" not defined) closed - <http://trac.edgewall.org/ticket/8483#comment:2> - cboos
[05:03:34]<evil_twin>t.e.o: InterMapTxt edited - <http://trac.edgewall.org/wiki/InterMapTxt?version=29> - cboos
[05:03:39]<evil_twin>t.e.o: Ticket #8483 (UndefinedError: "dateinfo_delta" not defined) updated - <http://trac.edgewall.org/ticket/8483#comment:3> - cboos
[05:03:44]<evil_twin>t.e.o: Ticket #8483 (UndefinedError: "dateinfo_delta" not defined) updated - <http://trac.edgewall.org/ticket/8483#comment:4> - cboos
[05:13:29]<evil_twin>t.e.o: Ticket #8459 (svn:mergeinfo rendering is far too slow) updated - <http://trac.edgewall.org/ticket/8459#comment:23> - cboos
[05:33:29]<evil_twin>t.e.o: setuptools edited - <http://trac.edgewall.org/wiki/setuptools?version=3> - cboos
[05:43:29]<evil_twin>t.e.o: Ticket #8536 (Don't display Default Repository on main Browse page) updated - <http://trac.edgewall.org/ticket/8536#comment:10> - rblank
[06:52:48]<gpannwitz>I want to configure ( per component ) a default user to be assigned, and also others to be cced
[06:59:10]*retracile growls something incontrovertible about mornings.
[07:13:31]<evil_twin>t.e.o: TracUsers edited - <http://trac.edgewall.org/wiki/TracUsers?version=859> - daniel@…
[07:41:17]<retracile>pacopablo: glad to see logs up again :)
[07:42:16]<retracile>pacopablo: however, the link in the topic needs to have a /trac added. And whatever you're using to display the logs needs to provide a list of channels logged instead of an error on that page....
[07:58:01]<doki_pen>pacopablo: what's wrong?
[07:59:05]<doki_pen>is it working?
[07:59:52]<doki_pen>pacopablo: cool, but looks like you need to set the charset?
[08:00:26]<doki_pen>and you still have to import the old logs in the db, right?
[08:37:28]<pacopablo>doki_pen: yeah, still need to import
[08:37:42]<pacopablo>how do I set the charset? channel.foo.charset = UTF8 ?
[08:38:26]<pacopablo>doki_pen: and I've gota couple issues
[08:38:45]<pacopablo>first, there should be a listing page so that one can see all the channels logges
[08:38:53]<pacopablo>s/logges/logged
[08:38:53]<evil_twin>pacopablo meant: first, there should be a listing page so that one can see all the channels logged
[08:39:18]<pacopablo>second, if no default channel is set, then you get the error that is currently up at http://pacoapblo.com/irclogs
[08:39:24]<pacopablo>KeyError: 'channel'
[08:40:33]<pacopablo>third, if you have a default channel set, as well as a multichannel, then it shows two links for the same channel
[08:40:56]<pacopablo>for the third, I guess that one need consider the default channel as one of the multichannels
[08:41:25]<pacopablo>but since the default is inherited into the channel.* section, then I think a more explicit handling is appropriate.
[08:41:51]<pacopablo>retracile: yeah, I'm going to try to get the trac logs back at /irclogs
[08:42:41]<retracile>pacopablo: cool
[08:42:59]<retracile>pacopablo: My logs are somewhat spotty over the past few months, but if you need 'em, let me know.
[08:43:13]<pacopablo>well, I should have most of the last bit
[08:43:18]<pacopablo>it's been logging, but to file
[08:43:26]<pacopablo>and so I have to import them into the db still
[08:43:56]<doki_pen>pacopablo: yes, as you say is how you set charset
[08:44:15]<doki_pen>pacopablo: yeah, i'm thinking about that
[08:44:22]<doki_pen>pacopablo: the problem is, it shouldn't be /irclogs
[08:44:30]<doki_pen>because that is for default channel
[08:44:42]<doki_pen>(for ppl who don't want to use the multi-channel feature)
[08:45:07]<doki_pen>pacopablo: ah, error checking, ok
[08:46:00]<doki_pen>about the 3rd issue, I think I made it work the way I wanted
[08:46:08]<pacopablo>ok, so setting charset to UTF8 doesnt' work
[08:46:21]<pacopablo>complains about not being able to decode unicode
[08:46:23]<doki_pen>the only reason you should set channel = blah is because you want the default channel
[08:46:37]<doki_pen>pacopablo: really? it should be errors=ignore
[08:46:40]<pacopablo>doki_pen: you could have /irclogs/channels
[08:46:46]<doki_pen>this must be specific to the D
[08:46:48]<doki_pen>B
[08:47:07]<doki_pen>I'd rather have /irclogs-channels
[08:47:15]<pacopablo>that's fine too
[08:47:36]<doki_pen>the trouble with irc is, ppl can send in any charset they want
[08:47:39]<pacopablo>so, to me, a default channel comes into play in two instances: 1: that's the only channel in use
[08:47:48]<doki_pen>so the logs will contain strings in different charsets
[08:47:51]<pacopablo>2: you're selecting between one of the many channels you have configured
[08:48:29]<doki_pen>so maybe default channel should be shut off if there are any multichannels setup
[08:48:29]<pacopablo>in case 2: where you have channel.* set, the default channel should then be one of those entries
[08:48:34]<doki_pen>I could life with that
[08:48:38]<doki_pen>live
[08:48:50]<pacopablo>but if you specify it that way, then you get duplicate channels
[08:49:03]<pacopablo>don't get me wrong, I do like the default channel
[08:49:17]<pacopablo>and especially for #trac, it makes it nice
[08:49:25]<doki_pen>what is nice about it?
[08:49:26]<pacopablo>since I've got a history of #trac logs at /irclogs
[08:49:37]<doki_pen>it's just a url though, how does it help?
[08:49:56]<pacopablo>anyone who's linked to me already
[08:50:01]<doki_pen>ah..
[08:50:03]<pacopablo>changing the URL breaks that
[08:50:03]<doki_pen>yes, good point
[08:50:16]<pacopablo>and google has indexed the logs fairly well
[08:50:26]<retracile>pacopablo: of course, the url has been broken for a while now... ;)
[08:50:32]<pacopablo>though since they've been down for a bit, I don't know how that's affected their chache
[08:50:48]<pacopablo>retracile: details....
[08:50:54]<doki_pen>so did you get the charset issue worked out?
[08:50:56]<pacopablo>retracile: if I can fix it, then that's a good thing ;)
[08:51:05]*retracile snipes from the shadows for cheap amusement
[08:51:06]<doki_pen>I'm thinking it may be a problem when you store it and not when you display it
[08:51:08]<pacopablo>doki_pen: no, I don't know what charset to use
[08:51:24]<pacopablo>well, the DB is UTF8
[08:51:27]<doki_pen>let me check how it looks from bip logs
[08:51:29]<retracile>pacopablo: oh, definitely. Glad to see you're getting to it. :)
[08:51:31]<pacopablo>so that's how it's giong into the db
[08:51:58]<doki_pen>pacopablo: if you look at the actual bytes, is it def. utf-8?
[08:52:14]<doki_pen>maybe I neglected to put the charset conversion in the db-backend
[08:52:21]<pacopablo>huh?
[08:52:32]<pacopablo>how do I look at the acutal bytes in the DB?
[08:52:53]<pacopablo>oh, btw, I've got a patch for the chatlog to covert it ot use SQLAlchemy instead of the old gozerbot DB backend
[08:53:09]<pacopablo>cus I've always had issues with the old DB backend and PostgreSQL
[08:53:22]<pacopablo>I'll post the patches to dev.gozerbot.org today
[08:53:41]<doki_pen>ok, great
[08:53:50]<doki_pen>if you want to send it to me I can apply to my branch
[08:53:53]<pacopablo>it'll do the table creation also
[08:53:56]<doki_pen>or whatever is faster
[08:53:59]<doki_pen>awesome
[08:54:11]<pacopablo>which was bit of a sore point when trying to configure it
[08:54:22]<doki_pen>yeah, creating the table?
[08:54:39]<doki_pen>I wonder if anyone else will find DB backend useful
[08:54:39]<pacopablo>you fail to mention the field types on the wiki page, and you also fail to mention that you're using an id field as the PK
[08:54:54]<pacopablo>if you build it, they will come :)
[08:54:59]<doki_pen>haha
[08:55:00]<pacopablo>I'm sure they will
[08:55:26]<doki_pen>field types == I tested with sqlite
[08:55:27]<doki_pen>lol
[08:55:56]<doki_pen>I mainly was avoiding making suggestions about field sizes vs. blobs etc.
[08:56:07]<doki_pen>but feel free to add more docs, I just didn't know
[08:56:57]<pacopablo>will do.
[08:57:18]<pacopablo>have you looked at the IRC RFC ?
[08:57:21]<doki_pen>look at provider/db.py
[08:57:30]<pacopablo>yeah, have done
[08:57:30]<doki_pen>line 69-71
[08:57:33]<doki_pen>that should do the conversion
[08:57:41]<pacopablo>k
[08:57:57]<doki_pen>no, I haven't look at the RFC, but figured it would be needed to make useful suggestions
[08:57:58]<pacopablo>do you know what the max lengh for nick and channel are?
[08:58:05]<pacopablo>k
[08:58:14]<pacopablo>also, how does one define the "network" ?
[08:58:22]<doki_pen>no, but I'm guessing that the different server software doesn't stick to the RFC
[08:58:31]<doki_pen>I wouldn't go strictly by the RFC, but pad it a little
[08:58:47]<doki_pen>network is more like a namespace then anything else
[08:58:52]<pacopablo>yeah, right now I have ll the fields, except message, set to VARCHAR(256)
[08:58:57]<pacopablo>well, time is a timestamp
[08:59:07]<doki_pen>but really, the only time you really need it is when you are using multi-networks
[08:59:16]<pacopablo>k
[08:59:39]<doki_pen>because there could be #dunkbots on two different networks, and you could be /join 'd to both
[08:59:45]<doki_pen>or something like that
[09:00:00]<pacopablo>true
[09:00:15]<doki_pen>have you looked at my code at all?
[09:00:24]<doki_pen>I'm not much of a python dev, does it look ok?
[09:00:50]<pacopablo>so far it looks fine.
[09:01:16]<doki_pen>one thing that would make performance much better is genshi streaming
[09:01:43]<doki_pen>There should be some way in trac to skip the content-length header and stream the output to the browser
[09:01:44]<pacopablo>yeah, but that's a bit outside the scope of the plugin
[09:01:54]<doki_pen>yeah, it would be a nice feature of trac though
[09:02:08]<pacopablo>well, you can handle the request yourself
[09:02:23]<pacopablo>you jsut have to do all the HTML generation, etc.
[09:02:34]<doki_pen>but that's just calling genshi, right?
[09:02:39]<pacopablo>no
[09:02:54]<pacopablo>if you don't return anything from process_request
[09:03:00]<pacopablo>it'll skip the genshi rendering parts
[09:03:04]<retracile>Hm. SimpleTicketPlugin has a warning that it doesn't work with 0.11dev, and hasn't been updated since '08. Is that or similar functionality available for 0.11-stable somewhere?
[09:03:04]<doki_pen>pacopablo: did you see line 97-117 in web_ui?
[09:03:16]<pacopablo>the problem is that you have to send the headers and make sure you close it properly, etc.
[09:03:58]<doki_pen>I need to add a comment to that BS
[09:04:35]<pacopablo>you mean in _render_line ?
[09:04:39]<doki_pen>yeah
[09:06:22]<somethingelse>hello. i am running trac and svn. when i commit changes in svn, some directories have the trac ticket box in the commit screen and some don't. i was expecting all the directories in the repo to have it. how can i fix this?
[09:06:58]<somethingelse>i'm using tortoisesvn client, working with a single svn repo
[09:07:49]<doki_pen>trac has tortoise integration?
[09:08:07]<doki_pen>12:00 < pacopablo> the problem is that you have to send the headers and make sure you close it properly, etc.
[09:08:11]<doki_pen>12:00 < pacopablo> the problem is that you have to send the headers and make sure you close it properly, etc.
[09:08:15]<doki_pen>poop, sorry
[09:08:33]<evil_twin>t.e.o: Changeset [8432]: MultiRepos: A few small fixes in repository management: * Fixed an issue ... - <http://trac.edgewall.org/changeset/8432> - rblank
[09:08:54]<somethingelse>doki_pen, yes, it does :)
[09:08:54]<pacopablo>doki_pen: re: charset, in provider/db.py, it's barfing because you're trying to convert it to unicode
[09:09:13]<pacopablo>doki_pen: but with postgresql, the data being returned from the db is already unicode
[09:09:58]<doki_pen>pacopablo: what is the solution?
[09:10:08]<pacopablo>um, not sure yet
[09:10:08]<doki_pen>pacopablo: will trac's to_unicode fix it?
[09:10:33]<pacopablo>so, theoretically, all text in trac *should* be unicode already
[09:10:49]<pacopablo>unless it's possibly coming in from the request
[09:11:17]<pacopablo>I'll try to play around with it a bit
[09:11:21]<doki_pen>pacopablo: but unicoding a unicode str, shouldn't be an issue
[09:12:10]<doki_pen>ah, it is when you set the charset
[09:12:53]<doki_pen>it should be if not isinstance(message, unicode)
[09:13:23]<doki_pen>or.. if isinstance(message, str):
[09:14:57]<doki_pen>but the problem remains
[09:15:11]<doki_pen>you remove charset, because it doesn't need to be converted, it's already unicode
[09:15:17]<doki_pen>and now you get funny symbols..
[09:15:28]<doki_pen>ah, you know why?
[09:15:47]<doki_pen>it's the bold irc commands that it's having trouble with
[09:16:38]<doki_pen>but it's something to do with postgres.. because my text logs don't have that problem
[09:16:52]<doki_pen>maybe you need to specify the charset in the postgres table?
[09:17:14]<pacopablo>it's set to UTF-8
[09:17:26]<pacopablo>the encoding, at least
[09:17:54]<pacopablo>how do your text logs render the bold ?
[09:18:18]<doki_pen>so my file parser appears to be ignoring unknown chars
[09:18:30]<doki_pen>it doesn't render it at all
[09:18:32]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:18> - cboos
[09:18:36]<pacopablo>ahh
[09:18:45]<pacopablo>I'm fine with the blocks
[09:18:55]<doki_pen>let me hexedit the log file, maybe bip strips it out
[09:20:10]<pacopablo>the other option may be to use .encode()
[09:20:15]<pacopablo>instead of unicode()
[09:20:19]<doki_pen>I think chatlog should strip out control chars
[09:20:32]<doki_pen>i think the problem is with chatlog
[09:21:33]<pacopablo>it'd be really cool if the logs could render the bold
[09:22:01]<doki_pen>yeah it would
[09:22:25]<doki_pen>bip strips that out
[09:22:44]<doki_pen>ok, so leave the unicode blocks in for now, and I'll put it on the wishlist
[09:22:59]<doki_pen>convert them to <b></b>
[09:23:29]<doki_pen>and also interpret the color codes
[09:24:12]<pacopablo>yeah ;)
[09:24:32]<doki_pen>I think I need pagination first, it's already pretty slow
[09:26:03]<doki_pen>also, if I moved the caching up to the UI layer, it would have less impact
[09:27:50]<pacopablo>yeah
[09:28:33]<evil_twin>t.e.o: Changeset [8433]: MultiRepos: Avoid issues when copy/pasting repository paths in the ... - <http://trac.edgewall.org/changeset/8433> - rblank
[09:29:05]<doki_pen>pacopablo: so do you need help setting up the default channel?
[09:29:13]<doki_pen>paste your config
[09:29:14]<pacopablo>hmm, so, your charset code actually works with SQLite?
[09:29:24]<pacopablo>or just with file based?
[09:29:28]<doki_pen>pacopablo: who knows, maybe I didn't test it right
[09:29:38]<pacopablo>so, i'm playing with .ecode()
[09:29:45]<evil_twin>pacopablo meant: so, i'm playing with .encode()
[09:29:45]<pacopablo>s/ecode/encode
[09:30:14]<pacopablo>problem is that trying to assign it to the result returned from the cursor isn't possible due to that being a tuple
[09:30:18]<pacopablo>and tuple are immutable
[09:30:32]<doki_pen>at some point it's a dict
[09:30:48]<pacopablo>doki_pen: I got the default channel setup, my complaint was more in how it happens
[09:31:00]<pacopablo>doki_pen: yeah, right after you convert it you assign it to a dict
[09:31:51]<doki_pen>so instead of returning the dict, mess with it, then return
[09:32:14]<doki_pen>oh, the l[i] = blah is a bug
[09:32:21]<doki_pen>right?
[09:32:43]<doki_pen>I guess I didn't test it then
[09:33:33]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:19> - cboos
[09:36:51]<pacopablo>doki_pen: i'm playing with the code now
[09:37:01]<pacopablo>I'll submit a patch once I get it working
[09:40:35]<doki_pen>ok
[09:44:52]<pacopablo>mm, slightly bigger problem :(
[09:45:11]<pacopablo>genshi.escape() doesn't handle charsets
[09:45:56]<pacopablo>so even though the I can get the text encoded via .encode() genshi then barfs on teh \x02 chars
[09:46:11]<pacopablo>I wonder why it works if I don't specify a charset
[09:51:10]<doki_pen>it's ascii?
[09:54:32]<pacopablo>or so it claims ;)
[09:54:58]<pacopablo>I'm trying to verify how the data is coming out of the DB
[09:57:05]<somethingelse>ok, i found solution to my problem. i needed to add SVN properties for trac integration to each folder
[09:59:04]<doki_pen>I should make a backend in chatlog that logs in <td> lines
[10:01:28]<doki_pen>paco, I also have #trac logs since june 22nd
[10:01:34]<doki_pen>if you are missing any
[10:01:39]<pacopablo>cool
[10:01:44]<pacopablo>I have to see what I'm missing
[10:03:33]<evil_twin>t.e.o: Changeset [8434]: MultiRepos: Removed `[trac] repository_url` again (introduced in [8431]), ... - <http://trac.edgewall.org/changeset/8434> - rblank
[10:03:38]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:20> - rblank
[10:05:08]<doki_pen>pacopablo: so after I get your changes and I fix search, I'm going to announce a release candidate
[10:06:07]<pacopablo>sounds wonderful
[10:08:07]<pacopablo>grr, the data is defintely coming back as unicode
[10:09:11]<pacopablo>ahh, the .encode() turns it back to a str
[10:09:39]<pacopablo>so that's not waht I want
[10:09:53]*pacopablo hates unicode and ascii :)
[10:11:33]<doki_pen>pacopablo: I don't understand what you are trying to do
[10:11:42]<doki_pen>I thought we figured it out
[10:12:06]<doki_pen>it's already utf-8, it's just the irc control characters that look funny, because they are not valid utf-8
[10:12:31]<doki_pen>the only reason I'm not seeing it, is because I'm using bip, and bip strips out the control characters before logging
[10:13:09]<doki_pen>if the logs are already unicode, and there is a charset specified, then we should through a warning and ignore charset
[10:13:22]<doki_pen>s/through/throw/
[10:13:22]<evil_twin>doki_pen meant: if the logs are already unicode, and there is a charset specified, then we should throw a warning and ignore charset
[10:14:19]<pacopablo>I was trying to see if there was a generic way of doing that without testing to see what type the string was
[10:16:27]<pacopablo>though it doesn't look like it
[10:16:48]*pacopablo falls back to testing the type of data
[10:22:52]<pacopablo>ok, now to generate some patches
[10:23:33]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:21> - cboos
[10:23:38]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:22> - cboos
[10:50:58]<pacopablo>doki_pen: #5598 on t-h.o
[10:53:33]<evil_twin>t.e.o: Changeset [8435]: TracMercurial for 0.12 (or MultiRepos, rather): add `get_path_url` in ... - <http://trac.edgewall.org/changeset/8435> - cboos
[10:53:38]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:23> - cboos
[11:16:21]<pacopablo>doki_pen: you have a login @ dev.gozerbot.org ?
[11:17:08]<pacopablo>indeed you do :)
[11:30:48]<pacopablo>doki_pen: http://dev.gozerbot.org/trac/ticket/126
[11:43:34]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:24> - rblank
[11:48:35]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:25> - rblank
[12:03:35]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:26> - rblank
[12:45:56]<helo>i can't figure out where to put my logo... the path is configured as "site/logo.png", but the html refers to /tracproj/chrome/site/logo.png
[12:46:20]<helo>there is no directory named "chrome" anywhere in my trac setup
[12:53:06]<helo>oh, i was using the wrong htdocs
[12:58:41]<evil_twin>t.e.o: Changeset [8436]: TracMercurial for 0.12: tweaks for `MercurialRepository.get_path_url` - ... - <http://trac.edgewall.org/changeset/8436> - cboos
[12:58:46]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:27> - cboos
[13:03:36]<evil_twin>t.e.o: Changeset [8437]: Remove a log statement from r8436 - <http://trac.edgewall.org/changeset/8437> - cboos
[13:23:36]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:28> - rblank
[13:37:58]<pacopablo>helo: just put it in the htdocs folder of the trac environment
[13:46:35]<jhammel>hmmm, there is no 'or's for filters :(
[13:48:40]<evil_twin>t.e.o: Ticket #1781 (Link from 'Browser Source' to Subversion Repository URL) updated - <http://trac.edgewall.org/ticket/1781#comment:29> - cboos
[13:50:20]<pacopablo>there is a ticket someplace on t.e.o that provides the functionality
[13:51:10]*jhammel still needs to write the PatchManagerPlugin so he can start using patches
[13:51:46]<rblank>jhammel: http://trac.edgewall.org/ticket/2647
[13:52:25]<jhammel>rblank: thanks
[13:53:36]<evil_twin>t.e.o: Changeset [8438]: MultiRepos: Fixed title of source browser and log pages. - <http://trac.edgewall.org/changeset/8438> - rblank
[13:53:41]<evil_twin>t.e.o: TracMercurial edited - <http://trac.edgewall.org/wiki/TracMercurial?version=55> - cboos
[14:07:31]<retracile>rblank: I really like comment #2 on that ticket :)
[14:08:37]<evil_twin>t.e.o: Changeset [8439]: MultiRepos: Merged from 0.12dev[8429-8430/trunk]. - <http://trac.edgewall.org/changeset/8439> - rblank
[14:13:18]<doki_pen>pacopablo: thanks
[14:15:45]<rblank>retracile: It did take *some* time, but he was proven right ;-)
[14:16:07]<retracile>rblank: :D
[14:19:32]<helo>pacopablo: thanks :)
[14:33:37]<evil_twin>t.e.o: Changeset [8440]: MultiRepos: Added missing `svn:eol-style` properties. - <http://trac.edgewall.org/changeset/8440> - rblank
[14:43:38]<evil_twin>t.e.o: 8536-hide-repos-r8439.patch attached to Ticket #8536 - <http://trac.edgewall.org/attachment/ticket/8536/8536-hide-repos-r8439.patch> - rblank
[14:46:11]<rblank>pacopablo: Want to help testing? http://trac.edgewall.org/ticket/8536#comment:11
[14:46:24]<pacopablo>will do
[14:46:32]<rblank>pacopablo: You'll need the current HEAD of multirepos, though.
[14:46:38]<pacopablo>k
[14:48:37]<evil_twin>t.e.o: Ticket #8536 (Don't display Default Repository on main Browse page) updated - <http://trac.edgewall.org/ticket/8536#comment:11> - rblank
[14:51:12]<pacopablo>rblank: so, if I already have a repos defined in my trac.ini, I need to remove it and add it via the web interface if I don't want it to be the default?
[14:54:39]<pacopablo>rblank: much nicer ;)
[15:01:40]<rblank>pacopablo: You can also define it in the [repositories] section of trac.ini
[15:03:38]<evil_twin>t.e.o: Ticket #8536 (Don't display Default Repository on main Browse page) updated - <http://trac.edgewall.org/ticket/8536#comment:12> - jhampton
[15:04:23]<pacopablo>rblank: so, if it's in the [repositories] section it won't be considered default? just if it's in the [trac] section?
[15:33:38]<evil_twin>t.e.o: install created - <http://trac.edgewall.org/wiki/install?version=1> - anonymous
[15:37:19]<sanmarcos>how do I change the footer where the trac logo is?
[15:39:19]<evil_twin>customization is http://projects.edgewall.com/trac/wiki/TracInterfaceCustomization
[15:39:19]<pacopablo>!faq customization
[15:39:28]<pacopablo>I beleive the answer is theri
[15:39:34]<sanmarcos>yeah it doesn't say how to customize it, it only changes the userfooter not the real footer
[15:39:35]<evil_twin>pacopablo meant: I beleive the answer is there
[15:39:35]<pacopablo>s/theri/there
[15:39:41]<sanmarcos>and genshi has no docs apparently
[15:40:15]<pacopablo>er, so you create a template that has a py:match block that matches the footer
[15:40:27]<pacopablo>I'll admit I'm not too good at the whole interface customization thing
[16:13:19]<nphase>is there any way to speed up trac performance?
[16:13:28]<nphase>it um... seems to take forever to load some pages
[16:13:34]<nphase>and its a really small install (< 250 revs, 5 tickets. new install)
[16:15:26]<doki_pen>pacopablo: disable the irclogs search module
[16:15:32]<doki_pen>for now
[16:16:17]<pacopablo>k
[16:16:27]<doki_pen>sanmarcos: can't you just copy the template to your projects template folder and edit it?
[16:16:49]<pacopablo>nphase: how are you serving trac? tracd/fastcgi/mod_python/cgi ?
[16:17:07]<nphase>mod_python
[16:17:16]<pacopablo>hmm
[16:17:24]<pacopablo>what plugins are you using?
[16:17:33]<doki_pen>nphase: first thing is to serve static content directly (not through trac)
[16:18:02]<nphase>doki_pen: static content seems fine. but when i do the view tickets page, it takes like 10 seconds to load
[16:18:04]<sanmarcos>doki_pen: can I ?
[16:18:08]<nphase>pacopablo: probably none? i don't know, how would i tell
[16:18:12]<doki_pen>sanmarcos: I'm pretty sure
[16:18:18]<sanmarcos>doki_pen: we are suppose to use this genshi thing that only the developers know how to use
[16:18:41]<doki_pen>sanmarcos: they are genshi templates
[16:18:51]<sanmarcos>doki_pen: the problem with that is that with updates you have to edit the file again
[16:18:57]<doki_pen>sanmarcos: I would advice against transformation, it's slow
[16:19:13]<doki_pen>sanmarcos: then you could use a transformation
[16:19:15]<sanmarcos>its the default on trac doki_pen
[16:19:22]<nphase>attachments, browser, changeset, header_logo, inherit, logging, ilestone, mimeviewer, notification, project, query, report, revisionlog, roadmap, search, svn, ticket, ticket-workflow, timeline, trac, wiki, svn-post-commit-hook, mainnav
[16:19:24]<sanmarcos>right I want to do a tansformation of the trac footer, but I cant
[16:19:26]<pacopablo>nphase: then not sure, shouldn't be that bad.
[16:19:34]<doki_pen>sanmarcos: how are you doing it?
[16:19:35]<nphase>are the headers in trac.ini
[16:20:05]<sanmarcos>doki_pen: I am not doing it because I don't know genshi, and there are no docs
[16:20:20]<doki_pen>sanmarcos: there are docs, what are you having trouble with?
[16:20:51]<sanmarcos>i want to change the footer where the powered by trac image is, nothing more.
[16:21:04]<doki_pen>sanmarcos: http://genshi.edgewall.org/
[16:21:27]<doki_pen>do you mean you don't know how to setup a transformation in trac?
[16:21:39]<sanmarcos>yes, how would I change the footer?
[16:21:48]<sanmarcos>with Genshi
[16:23:15]<doki_pen>sanmarcos: I think you would make a theme.html genshi template
[16:23:24]<sanmarcos>yes of course, but how do you change the footer?
[16:23:35]<sanmarcos>what is the syntax of genshi
[16:23:42]<sanmarcos>to change the footer WHERE the trac powered image is
[16:24:20]<doki_pen>one sec, I've never done it, only in code have I done it
[16:24:23]<doki_pen>give me a minute
[16:24:36]<sanmarcos>thanks, you see how hard it is for us end users
[16:24:50]<doki_pen>sanmarcos: I'm just reading the docs on edgewall.org :P
[16:25:38]<doki_pen>sanmarcos: http://genshi.edgewall.org/wiki/Documentation/xml-templates.html#id8
[16:26:39]<sanmarcos>how would you use that to change this <div> div id="footer" lang="en" xml:lang="en">
[16:26:57]<sanmarcos>as I see it that replaces python variables
[16:28:28]<doki_pen>I don't think that's right
[16:28:34]<doki_pen>let me look a little longer
[16:28:57]<sanmarcos>trac/templates/footer.cs
[16:29:58]<doki_pen>sanmarcos: look at this: http://genshi.edgewall.org/wiki/GenshiTutorial
[16:30:08]<doki_pen>sanmarcos: ah, css is probably more performant, good thinking
[16:30:30]<doki_pen>sanmarcos: go to the section "Adding a Layout Template"
[16:32:19]<sanmarcos>nothing
[16:32:31]<sanmarcos><py:match path="footer> does nothing
[16:32:51]<sanmarcos>Genshi is wonderful I am sure, but it is so damn complex that it is near impossible for end users to do basic things
[16:35:18]<doki_pen>sanmarcos: http://genshi.edgewall.org/wiki/Documentation/xml-templates.html#id5
[16:35:20]<doki_pen>it should work
[16:35:35]<doki_pen>you need xpath in path, me thinks
[16:35:43]<sanmarcos>what is an xpath?
[16:35:55]<doki_pen>sanmarcos: it's a query language for xml
[16:36:02]<doki_pen>sanmarcos: I'll make one for you, one sec
[16:36:13]<doki_pen>sanmarcos: it's pretty standard, you can find instructions anywhere
[16:36:56]<doki_pen>a[@id="tracpowered"]/img
[16:37:32]<doki_pen>try <py:match path='a[@id="tracpowered"]/img'></py:match>
[16:38:36]<doki_pen>if you want to remove the entire link, delete the trailing /img
[16:39:01]<sanmarcos><py:match path='div[@id="footer"]'>
[16:39:10]<sanmarcos>that should match a DIV with ID "footer" right?
[16:39:14]<doki_pen>yes
[16:41:50]<sanmarcos>I cant put HTML inside the <py:match> ?
[16:42:16]<sanmarcos>sorrry I meant, call <?cs ?
[16:44:20]<doki_pen>I can't parse that.
[16:45:08]<sanmarcos>in the trac templates they call variables with <?cs?>
[16:45:18]<sanmarcos>example <?cs var:htdocs_location ?>
[16:45:23]<sanmarcos>how can I do it from my site.html?
[16:45:35]<sanmarcos>(if I put that, trac crashes)
[16:45:58]<doki_pen>I think that is really old
[16:46:13]<sanmarcos>that is trac 0.11.5 in templates/footer.cs
[16:46:24]<doki_pen>something is wrong
[16:46:34]<sanmarcos>?
[16:46:48]<doki_pen>all templates should end with .html
[16:46:56]<doki_pen>that is the old clearsilver templating system
[16:47:04]<sanmarcos>still there
[16:47:22]<doki_pen>you did an upgrade?
[16:47:26]<sanmarcos>no
[16:47:27]<sanmarcos>clean install
[16:47:30]<sanmarcos>from source
[16:47:58]<doki_pen>ha.. I see
[16:48:27]<sanmarcos>got it
[16:48:31]<sanmarcos>${chrome.htdocs_location}
[16:48:35]<doki_pen>yeah
[16:48:46]<doki_pen>but why do the old templates still get deployed?
[16:48:51]<doki_pen>I'm pretty sure that's a bug
[16:49:03]<sanmarcos>check in your own installation
[16:49:11]<sanmarcos>it still uses CS
[16:49:11]<doki_pen>sanmarcos: I see them
[16:49:45]<doki_pen>I don't think so, in the egg it's genshi
[16:50:01]<sanmarcos>huh?
[16:50:10]<doki_pen>the trac egg has templates in it
[16:50:55]<doki_pen>and footer is defined in theme.html
[16:51:20]<doki_pen>maybe there is some reason that the cs stuff is still in the tree
[16:51:52]<sanmarcos>oh you are right is in theme.html
[16:52:01]<sanmarcos>and it has the syntax I want!
[16:52:08]<sanmarcos>why is that .cs crap stil in there?
[16:52:17]<doki_pen>no idea, very confusing though
[16:53:03]<sanmarcos>the thing is that by doing genshi transformations, I can safely update trac with easy_install and not have to worry about editing their template files again
[16:53:28]<doki_pen>sure, but I think the it's safe to ignore changes in theme.html
[16:53:30]<sanmarcos>and also with this new .egg update method it is so much better than having trac from your distro which is outdated and most distros dont update often
[16:53:36]<doki_pen>that's the whole point of putting things in there
[16:54:20]<doki_pen>but doing it in site.html is fine, it will just be a little slower
[16:54:35]<doki_pen>I'm going to see the impact now
[16:55:45]<doki_pen>about 15 milis
[16:55:52]<doki_pen>on my 2Ghz system
[16:56:11]<doki_pen>15-25
[16:56:42]<sanmarcos>how do i reference my project's htdocs?
[16:57:31]<sanmarcos>{chrome.htdocs_location} is for trac's own chrome
[16:57:47]<doki_pen>point (/[^/]+)?/chrome/common at trac/htdocs
[16:57:59]<doki_pen>oh
[16:58:11]<doki_pen>one sec
[16:58:17]<sanmarcos>right, I want to access /var/trac/myproject/htdocs/image.png from my template
[16:58:48]<doki_pen>what we did is put /static to ours
[16:58:56]<doki_pen>and /static/chrome/common to tracs
[16:58:59]<sanmarcos>but there must be a dynamic
[16:59:24]<doki_pen>the base url
[16:59:54]<sanmarcos>where can I see all the variables I can use in my template?
[17:00:25]<doki_pen>ah, I guess we put /static/blah into our templates
[17:01:05]<doki_pen>yeah, our custom templates are all src="/static/blah blah"
[17:01:13]<doki_pen>why not do that?
[17:01:20]<sanmarcos>${href.chrome()}) apparently too
[17:02:02]<sanmarcos>got it
[17:02:09]<doki_pen>I guess it would be a little painful if we moved static stuff to a different host, to go through the templates and change them
[17:02:12]<sanmarcos>${href.chrome('site/image.png'} did it
[17:02:18]<doki_pen>ah, cool
[17:02:27]<doki_pen>what does the url look like?
[17:02:42]<sanmarcos> /chrome/site/python_logo_mini.png
[17:02:53]<sanmarcos>which references /var/trac/project/htdocs/python_logo_mini.png
[17:05:58]<sanmarcos>right, because you have /chrome/common and /chrome/site
[17:09:22]<sanmarcos>i've got it! thanks for your help doki_pen!
[17:16:00]<doki_pen>no problem, sorry i'm not better at this stuff..
[18:46:00]<pacopablo>coderanger__: any reason you don't handle HEAD request in the HTTPAuthPlugin ?
[18:46:45]<pacopablo>nm I take that back
[18:47:00]<pacopablo>you do handle them you just don't write anything back on the request
[18:47:06]<pacopablo>which is obviously fine