Team Chat Logs

February 23, 2010

2010 1
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
<< <   > >>

[00:29:02.933786]<cboos>t.e.o going down for ugprade... expect 10 minutes offline
[00:29:25.214220]<otaku42>moin
[00:29:33.192212]<cboos>moinmoin!
[00:30:38.776954]<cboos>otaku42: you know that now that 0.12dev is finalizing and that the 0.11 version of TracHacks is not yet final, I'll now bug you to upgrade directly to 0.12 ;-)
[00:31:30.537613]<otaku42>cboos: yes :) i'm currently gathering my first experiences with 0.12dev in a non-public, production site i'm setting up for the office
[00:31:41.685047]<thijstriemstra>test
[00:31:43.741473]<cboos>great!
[00:32:04.004796]<thijstriemstra>finally, seems I wasn't able to talk here without registering my nick at freenode
[00:32:11.439603]<otaku42>cboos: and the more i play with it, the more i love it and the more i'm tempted to do a direct transition of t-h.o to 0.12dev :)
[00:32:27.299358]<cboos>nice ;-)
[00:32:57.108365]<cboos>but what would really rock would be support for per-hack repository
[00:33:03.901508]<cboos>svn or hg
[00:33:49.870230]<otaku42>cboos: yeah, indeed, but that'd require a lot of additional work to the trachacks plugin
[00:34:01.244413]<cboos>that's not a prereq for a 0.12 upgrade of course, only one (added) possibility
[00:35:48.616753]*cboos => coffee
[00:36:03.993588]<cboos>hopefully the upgrade will be finished when I come back ;-)
[00:36:09.857127]*otaku42 .oO( good idea, getting a coffee, too )
[00:45:13.570982]<cboos>I should really learn how to add a "site being upgraded static page" instead of bringing the whole thing down...
[00:52:15.094739]<cboos>nice, cmlenz and jborg at once :-) hello guys
[00:52:31.883500]<cmlenz>hey it's like in the old days
[00:52:40.664385]<jborg>heh, I guess that doesnät happen too othen these days :)
[00:52:57.382683]<TinoW>hi
[00:53:00.931061]<jborg>so what just happened to t.e.o?
[00:53:09.474656]<cboos>I'm in the middle of an upgrade
[00:53:17.472244]<cboos>so there's a db backup going on
[00:53:33.149318]<TinoW>I get a key error 'id' upon rendering ticket - this is trac0.12 (trunk) - is this a known issue?
[00:53:59.800518]<cboos>currently at 247M on an expected total of 254M ... so it's nearly over
[00:54:26.649823]<cboos>TinoW: no, please post the full traceback somewhere
[00:54:36.042087]<cboos>e.g. http://paste.lisp.org/new/trac
[00:58:33.242560]<Sacho>Is there a way to turn off deprecation warnings when I use trac-admin?
[00:58:45.654607]<Sacho>/var/lib/python-support/python2.6/svn/fs.py:27: DeprecationWarning: The popen2 module is deprecated. Use the subprocess module.
[00:59:45.683303]<cboos>ok, upgrade finished - I'll try to improve the procedure next time with a nice "site being upgraded" page...
[01:08:42.601142]<TinoW>cboos: hm... don't see the announce, but here is the link: http://paste.lisp.org/+21O3
[01:08:43.242817]<gozerbot>trac: Ticket #8046 (Sort browser entries by author) closed - <http://trac.edgewall.org/ticket/8046#comment:12> || Changeset [9245]: Follow-up to r9240: also shorten author name for the repository index and ... - <http://trac.edgewall.org/changeset/9245>
[01:16:36.339613]<cboos>TinoW: are you using the fi translation?
[01:17:57.628842]<cboos>mh, no not the fi one...
[01:18:21.978625]<TinoW>cboos: no, german but it does not work with en_GB either
[01:18:27.143301]<cboos>but you must be using one which doesn't have the proper message keys
[01:19:14.333935]<cboos>well the german one looks correct, but maybe you're using an outdated one
[01:19:41.609362]<cboos>can you do "python setup.py compile_catalog -l de -f" then reinstall?
[01:20:27.671141]<cboos>(anyway, we should trap those errors in a better way)
[01:22:46.919152]<TinoW>cboos: this means at least 2 of them?
[01:23:13.616972]<TinoW>cboos: trunk from weekend - was there an update?
[01:29:50.958853]<cboos>jborg: now that I have you at hand (and if you happen to have 2 minutes), could you look at http://trac.edgewall.org/attachment/ticket/8443/t8443-deactivate-r9234.patch and tell me what you think before I mess up your nice db pool?
[01:34:17.785922]<TinoW>eek. on upgrade: pkg_resources.DistributionNotFound: Trac==0.12dev-r9208
[01:38:52.273593]<jborg>cboos: sure... as far as I can tell it looks good
[01:39:28.970530]<cboos>jborg: thanks!
[01:39:41.980364]<cboos>TinoW: re-deploy ?
[01:39:59.049668]<cboos>TinoW: I'm still trying to understand the error ...
[01:40:19.644776]<cboos>(a Genshi i18n expert would be welcomed ...)
[01:41:52.302068]<cboos>there's also the "child::*|child::text()" part, which doesn't sound familiar - maybe you have a plugin that does some xpath manipulations and this triggers the bug
[01:42:07.253754]<cboos>(or a match template in your site.html)
[01:44:37.492645]<cboos>are you using the "theme engine" stuff? if yes, try to disable it
[02:20:22.556807]<thijstriemstra>the roadmap says 0.12 is 3 months late, but i read earlier its getting close now. can someone update the roadmap, and update the milestone date, even if its a year from now .. :)
[02:20:50.724294]<cboos>well, no need to pretend we're not late ... ;-)
[02:21:24.533421]<cboos>and frankly, I can't make a new estimate either (one that is even remotely realistic)
[02:21:35.256015]<thijstriemstra>id like to switch to 0.12 but wonder if it makes sense to start already
[02:21:40.178490]<thijstriemstra>alright cool
[02:21:58.308331]<cboos>it does ... we're quite near a beta, I'd say
[02:22:51.744477]<cboos>but as lots of changes are happening these days, you might want to wait to see "cool it down" a bit first, unless you're willing to help test the stuff, which would be a good thing of course
[02:23:21.665682]<thijstriemstra>i hate doing css updates and if there are a lot of changes to be merged in the next weeks then id hold back on it
[02:23:28.982991]<thijstriemstra>ok
[02:24:27.763701]<cboos>I was more thinking in terms of the occasional bug inadvertently introduced, I don't expect big changes in the CSS
[02:24:38.183825]<thijstriemstra>yea forgot its sprint time this week hehe, good stuff
[02:26:03.665989]<thijstriemstra>looking at 0.12/TracUpgrade, is it possible to setup a test instance and keep running 0.11 stable for the production sites.. i guess a second .wsgi file pointing to the 0.12 checkout would do it
[02:26:46.838101]<cboos>sure, that's always possible - you may want to install the other version of Trac in a virtualenv, though
[02:27:13.254273]<thijstriemstra>alright, thx
[02:27:24.590501]<cboos>the TracDev/DevelopmentEnvironmentSetup should have some adequate info
[02:28:35.191296]<thijstriemstra>trunk of genshi. i guess a new version will be released before 0.12
[02:29:35.561330]<cboos>yep, seems 0.6 is on the way
[02:38:44.450249]<gozerbot>trac: Changeset [9248]: TracMercurial: add `MercurialChangeset.get_branches()` introduced in ... - <http://trac.edgewall.org/changeset/9248> || Changeset [9247]: TracTimeline: display branch information for changesets, when available. ... - <http://trac.edgewall.org/changeset/9247> || TracDev/DevelopmentEnvironmentSetup edited - <http://trac.edgewall.org/wiki/TracDev/DevelopmentEnvironmentSetup?version=25>
[02:38:49.571419]<gozerbot> || Changeset [9246]: Follow-up to r9224: changeset search results as retrieved from the ... - <http://trac.edgewall.org/changeset/9246>
[02:54:54.597145]<TinoW>cboos: thx, this was a strange error during upgrade. Interactively with trac-admin worked
[02:56:19.138593]<TinoW>now pre-commit-hook fails mysteriously: trac.core.TracError: Unsupported database type "postgres"
[02:56:44.839995]<TinoW>while with the same env I can use trac-admin successfully
[02:57:07.026260]<TinoW>(from another user that is - both www and testuser are in the same group as svn/trac)
[02:57:42.415766]<cboos>TinoW: svn hooks are run by svn with a very minimal environment, be sure to start them the way shown in the svnbook for troubleshooting
[02:58:13.264259]<TinoW>cboos: yes I did - I also had export >&2 in the script to debug so I see the environment is indeed set
[02:58:37.589524]<cboos>brb
[03:08:45.281923]<gozerbot>trac: Ticket #8843 (Display branch name in timeline) closed - <http://trac.edgewall.org/ticket/8843#comment:8>
[03:08:54.381919]<tyll>Hi, I wrote a plugin that shows the contents of a directory inside of trac. But the return value of the plugin is not searchable via tracsearch. Is it possible to make it searchable or do I have to use an external script instead that stores the directory listing in a wiki page directly?
[03:11:09.290492]<cboos>tyll: you could also implement ISearchSource interface in your plugin
[03:16:38.721731]<tyll>cboos: do you know some plugin that does this but is a wiki macro?
[03:17:03.684812]<cboos>wiki macros are plugins like any other ...
[03:18:23.658966]<cboos>ok, if you're inheriting from WikiMacroBase, some of the plugin specific stuff is hidden, but it's nevertheless there, you can implement additional interfaces
[03:20:15.965619]<cboos>(If you add "implements( ... )" in your WikiMacroBase subclass, I'm not sure if you need to list the IWikiMacroProvider again or not, you'd have to try)
[03:51:18.991844]<tyll>cboos: my current plugin just inherits Component and has implements(IWikiMacroProvider)
[03:53:24.599690]<tyll>cboos: what I am more interested in, is how to display the results or to see how other plugins integrate this into the search results
[05:03:27.120114]<tyll>How can I create a new hash that has another hash as a value? Or can I dump a hash using YAML-Syck in only one yaml stream?
[05:09:36.897087]<tyll>argh, sorry, wrong channel
[05:28:38.751078]<scfe>tyll: So you're looking for another search implementation?
[05:34:42.178371]<tyll>scfe: Kind of, especially one that is used to enhance the results for the wiki, e.g. thinking of the trac tags plugin, the search should also search the tags and display this together with the wiki page results
[05:36:12.372852]<scfe>tyll: IIRC this won't work. You can only add additional search results
[05:39:06.437257]<tyll>tyll: this might be good enough, too, as long as I can return wiki pages as a search result.
[05:39:28.341691]<tyll>args
[05:39:36.350335]<tyll>scfe: this might be good enough, too, as long as I can return wiki pages as a search result.
[05:40:20.502409]<tyll>scfe: or is there a way to make the search also expand macros in wiki pages? This would help, too. (not for the tags problem, but for another)
[05:41:09.207718]<scfe>tyll: We're doing something similar in our help page search for agilo
[05:41:19.931626]<scfe>you can check our demo server: http://agilo.agile42.com
[05:41:32.218813]<scfe>login data is written in the middle on the first page
[05:41:58.366216]<scfe>after logging in you can try how another search provider looks like: https://agilo.agile42.com/search?q=foo&noquickjump=1&agilo-help=on (very much like the wiki search provider)
[05:42:19.641519]<scfe>You can download the source ("agilo source code") on www.agile42.com
[05:42:40.890095]<scfe>Basically you can do everything if you do it yourself :-)
[05:45:06.203757]<tyll>scfe: thx, I will check it out later
[06:50:06.710222]*retracile mumbles something sardonic about mornings.
[07:05:40.222301]*pacopablo is pumped up from EXERCIZE!!!!!!!!!!!!!!!!!!!!!!!
[07:05:50.423686]<pacopablo>now to eat some nice fatty foods to slow me down
[07:05:51.405944]<rapha>pacopablo: good! :-)
[07:06:22.249890]<rapha>pacopablo: still didn't get that encoding problem sorted out :-/ ... i was hoping you might be able to take a look at it?
[07:06:27.834047]<pacopablo>so, rapha, can you check the DB and see what character it is that is causing the encoding problem?
[07:06:57.087716]<rapha>pacopablo: not sure where to look ... the test ticket i'm using doesn't have any umlauts or the like in it
[07:07:06.588518]<rapha>pacopablo: there exist some components with umlauts though
[07:07:18.685689]<pacopablo>so, even a test ticket causes the borkage?
[07:07:36.355691]<pacopablo>and what db backend?
[07:07:41.628059]<rapha>sqlite
[07:08:25.410612]<pacopablo>'bout time you got here rblank!
[07:08:46.552578]<rblank>Was here before you. And had breakfast already :)
[07:09:00.122509]<pacopablo>IRC is the ONLY place that matters ;)
[07:09:29.532991]<rapha>pacopablo: e.g.: sqlite> select * from component;
[07:09:32.937674]<rapha>(and so on)
[07:09:35.579900]<rapha>Tätigkeiten|rapha|
[07:10:07.949373]<pacopablo>what version of trac are you using?
[07:10:22.487437]<rapha>0.12-dev ... i wanted l10n
[07:10:29.531459]<pacopablo>k
[07:10:35.910086]<rapha>(and the new git plugin :-) )
[07:10:40.325118]<pacopablo>yay
[07:10:47.740974]<pacopablo>well, which new git plugin?
[07:10:57.897133]<rapha>the one with multi-repo support
[07:11:03.528618]<rapha>http://github.com/lentil/trac-git-plugin
[07:11:19.211890]<pacopablo>yeah, I'll be merging that back into the git plugin soon
[07:11:29.638294]<rapha>ah okay
[07:12:11.769993]<rapha>well, anyway, i already got my fingers dirty at trac/ticket/notification.py with unicode() but even that didn't help
[07:13:55.899967]<pacopablo>and specifying base64 for the mime_encoding didn't work?
[07:14:01.797033]<rapha>nope
[07:14:09.363143]<pacopablo>can you turn on logging and then pastebin the full traceback?
[07:14:29.008024]<rapha>sure
[07:16:49.238775]<rapha>pacopablo: http://gaia.2laborate.com/~rapha/trac.log that's one tracd startup + posting a test ticket
[07:23:53.725464]<pacopablo>rapha: I dont' see any traceback
[07:26:07.846667]<rapha>pacopablo: sorry, then i don't understand what it is that you want and how i can get it?
[07:28:18.008102]<pacopablo>well, you're saying that the notification is saying: ERROR: Failure sending notification on change to ticket #27: UnicodeEncodeError: 'ascii' codec can' ...
[07:28:30.553108]<pacopablo>if it is, then there should be a full traceback along with it
[07:29:04.276718]<pacopablo>such that it shows line numbers an a bit of context, etc. it helps with the debugging
[07:31:16.178367]<rapha>pacopablo: there's no such thing anywere. That ERROR can be seen both in the log and the UI though.
[07:34:41.796669]<rapha>pacopablo: actually, that error you just mentioned looks like one of yesterday's ... the one I'm getting with a fresh ticket is ofc a bit different than when I try to update one, but it's essentially the same and without any more info than is in the log I gave you the URL to.
[07:35:08.499177]<rapha>("Warnung: The ticket has been created, but an error occurred while sending notifications: 'ascii' codec can't encode character u'\xe9' in position 22: ordinal not in range(128)" is what it looks like in the UI, but in the log it's mentioned as an "ERROR")
[07:35:12.704043]<pacopablo>ahh, i see it now in the log
[07:35:16.995451]<rapha>k :)
[07:35:20.682889]<pacopablo>too bad the traceback wasn't printed.
[07:35:28.038994]<pacopablo>give me a second
[07:35:28.884460]<rapha>can I enable that somehow?
[07:43:33.223534]<pacopablo>rapha: on line 1168 of trac/ticket/web_ui.py, add exc_info=true as the last parameter
[07:43:49.112551]<pacopablo>the line should look like:
[07:43:49.437532]<pacopablo> "ticket #%s: %s", ticket.id, exception_to_unicode(e), exc_info=True)
[07:43:58.370312]<pacopablo>then run it again
[07:48:03.082153]<rapha>pacopablo: http://paste.lisp.org/+21OI
[07:52:10.743474]<pacopablo>rapha: does one of your recipient email addresses have unicode characters in it?
[07:52:26.576827]<rapha>definitely not.
[07:53:03.675822]<rapha>the only one right now should be raphael.schmid@gmail.com anyway (that was a paste right out of my prefs)
[07:53:09.075010]<pacopablo>how about your project url or the ticket reporter?
[07:53:54.726637]<rapha>oh
[07:53:59.674181]<rapha>my project url does indeed!
[07:54:04.524962]<rapha>http://iglbde.kalabanté.de/trac/
[07:54:20.225244]<rapha>that's xn--kalabant-a1i.de
[07:54:44.344444]<rapha>and the ticket reporter is me for these test tickets
[07:55:10.243785]<rapha>('rapha', no unicode chars)
[07:55:34.810424]<pacopablo>http://paste.lisp.org/+21OI/1
[07:55:47.715334]<pacopablo>try that patch and see if it fixes the issue
[07:55:58.674421]<rapha>k
[07:56:19.667962]<pacopablo>and then if it does, will you pleaes create a ticket on t.e.o (if you haven't already) and tell me the ticket number?
[07:57:05.904077]<rapha>sure
[07:59:16.633961]<rapha>YIPPEEEEE!
[07:59:21.100380]<rapha>pacopablo: it works! :-)
[08:05:42.744811]<rapha> pacopablo: http://trac.edgewall.org/ticket/9074
[08:08:48.381178]<gozerbot>trac: Ticket #9074 (Notification system shows warning/error when the project domain contains ...) created - <http://trac.edgewall.org/ticket/9074>
[08:11:19.156343]<rapha>pacopablo: thanks a lot! :-)
[08:13:06.280326]<pacopablo>np, I'll commit the fix shortly
[08:38:49.568316]<gozerbot>trac: Changeset [9250]: db: Follow-up to [9210], also use a microsecond timestamp for generating ... - <http://trac.edgewall.org/changeset/9250> || Ticket #9023 ("trac-admin upgrade" to upgrade README and kill wiki-macros/) closed - <http://trac.edgewall.org/ticket/9023#comment:2> || Changeset [9249]: admin: Remove `wiki-macros` on upgrade (or warn if it is non-empty).
[08:38:54.683288]<gozerbot> ... - <http://trac.edgewall.org/changeset/9249>
[08:45:23.613860]<drd20>Is it possible to embed an applet tag within a trac wiki page. Placing it within "{{{#!html " just producing white space within the formated page?
[08:46:45.731408]<cboos>drd20: if this is not yet a FAQ, it should probably become one...
[08:47:04.085784]<cboos>http://www.google.com/search?rlz=1C1CHMG_enDE291DE304&sourceid=chrome&ie=UTF-8&q=trac+embed+applet
[08:47:15.070197]<cboos>see the reply to second result
[08:48:52.576534]<drd20>hmm, don't have admin access to the box, can you create and load macros just within a page?
[08:49:07.768067]<cboos>this is not XWiki, no
[08:49:45.146129]<drd20>can't say I'd ever heard of xwiki
[08:50:04.063836]<cboos>(though you probably could create a macro that would allow you to create other macros ... :-) )
[08:51:05.496296]<drd20>:), but you can incoporate JavaScript? So could that not just create the tag
[08:52:30.113772]<cboos>you'd also need the "render_unsafe_content = true" for that, at which point you could just insert the <applet> tag directly
[08:53:57.860411]<drd20>Is there a security concern about allowing the applet tag?
[08:58:05.779868]<drd20>Are iframes also considered unsafe?
[08:58:48.506699]<cboos>a lot of stuff is considered unsafe, yes, even some styling things commonly used for phishing
[09:01:05.971927]<cboos>if those concerns are too drastic for your context (e.g. a small intranet) you can just set "render_unsafe_content = true", but otherwise it's better to stick with the default
[09:02:22.727335]<Morbus>anyone have any advice on how i could add table sorting capabilities to trac?
[09:02:57.502077]<cboos>see how MediaWiki is doing it, steal it, propose a patch ;-)
[09:03:08.614450]<Morbus>heh, heh.
[09:03:17.178861]<cboos>http://en.wikipedia.org/wiki/Help:Sorting
[09:03:25.318880]<Morbus>well, MW doesnt' have jQuery, so i'd end up using one of the jQuery plugins in trac.
[09:03:34.627317]<Morbus>but trac's table support is just garbage in general.
[09:03:42.499553]<Morbus>no way to define headers, etc.,
[09:03:57.377901]<cboos>... which means you haven't tried 0.12 yet ;-)
[09:04:03.911735]<Morbus>no, i'm still on 11.6
[09:04:07.598689]<cboos>too bad
[09:04:15.701188]<cboos>table support has ... much improved
[09:04:33.696862]<Morbus>what's the status of 12?
[09:04:50.365056]<drd20>Adding the "render_unsafe_contetn=true" hasn't allowed for the applet to load. Any ideas?
[09:04:54.846030]<cboos>Morbus: http://trac.edgewall.org/wiki/0.12/WikiFormatting#Tables
[09:05:11.788394]<cboos>Morbus: we're getting there, a handful of tickets left
[09:05:29.817215]<Morbus>cboos: well, ok. this is certainly bette.r
[09:05:42.460355]<Morbus>but if we were to rip MW off, then the || would have to be split syntactically.
[09:06:00.213593]<Morbus>well, no, that woudlnt' even work, i suppose.
[09:06:14.790428]<cboos>for more complex tables, you have dedicated wiki processors, see
[09:06:15.549219]<Morbus>cos MW's handling of sorts is on a table element, and we have no concept of that in trac - just a row.
[09:06:16.359885]<cboos>http://trac.edgewall.org/wiki/0.12/WikiFormatting#Processors-example-tables
[09:06:50.678619]<Morbus>that's still row level though.
[09:07:16.206215]<cboos>no, there's a #!table processor as well for adding styling or classes at the table level
[09:07:32.280182]<cboos>so basically adding sortable support for table would be to wrap the table with a {{{#!table class=sortable
[09:07:33.568047]<Morbus>could one do {#!table class="ahem"
[09:07:38.428262]<Morbus>||a||b||
[09:07:40.478579]<Morbus>!}
[09:07:40.794089]<Morbus>?
[09:07:44.243167]<cboos>yes
[09:07:50.107329]<Morbus>then that would work, yeah.
[09:08:24.857773]<Morbus>how is 12 with plugins? i use a crapload of plugins. would i have to wait for 'em to upgrade?
[09:09:46.486753]<cboos>as usual, plugins need to be checked and ported, we tried to document every API changes, so this should be possible. Of course, that's up to each plugin's author.
[09:10:04.457901]<Morbus>sure.
[09:10:10.869920]<cboos>and there have not been so many changes
[09:10:18.668562]<cboos>so many incompatible changes I mean
[09:12:28.770643]*Morbus nods.
[10:08:50.088726]<gozerbot>trac: Changeset [9251]: chrome: Increased the size of the Genshi template cache to 128 templates. ... - <http://trac.edgewall.org/changeset/9251>
[10:17:59.179294]<cboos>thumbs-up for r9251 !!!
[10:19:22.794949]<cboos>rblank: now if you (or other sprinters) could play a bit with #9025, we can finalize the last ticket scheduled for 0.11.7 and I'll roll out a rc1 ;-)
[10:19:51.674058]<rblank>cboos: Anything special you want to see checked?
[10:20:08.281069]<cboos>you mean for #9025?
[10:20:10.579145]<rblank>cboos: Performance? How do you measure?
[10:20:11.727041]<rblank>Yes
[10:20:15.352327]<cboos>speed, yes
[10:20:26.078478]<cboos>I did a crude
[10:20:41.271592]<cboos>time PYTHONPATH=. python trac/tests/allwiki.py
[10:20:48.374595]<cboos>that should be fine
[10:21:13.414470]<rblank>cboos: Oh, good idea. I'll quickly finish documenting ISystemInfoProvider, then I'll check #9025.
[10:21:21.246233]<cboos>(the challenge would be to reach the original speed or even beat it)
[10:21:34.002681]<rblank>cboos: Is that a contest?
[10:21:36.365014]<rblank>:)
[10:21:59.927701]<cboos>Yes, I guess that's the more effective way to approach this kind of bug ;-)
[10:22:22.252452]<cboos>each participant tries to grab some more milliseconds ;-)
[10:27:11.386801]<cboos>(btw, I already have a slightly better version, with a small twist to the WikiPageNames syntax - though the tests still pass, so ...)
[10:29:30.382975]*cboos is heading home
[10:38:51.235997]<gozerbot>trac: TracDev/ApiChanges/0.12 edited - <http://trac.edgewall.org/wiki/TracDev/ApiChanges/0.12?version=29>
[11:50:47.666862]<rblank>cboos: The regexp in #9025 is too complicated for my simple mind :)
[11:51:22.779216]<cboos>the new version is way simpler!
[11:51:42.323534]<rblank>cboos: Good news!
[11:52:00.457456]<cboos>one thing I'd like to do is to convert all of these regexps to re.VERBOSE so we can make them multiline, with comments
[11:52:29.848347]<rblank>cboos: Anyway, if the performance is within 5-10% of the old version, I'd say you're good to go.
[11:52:51.345953]<rblank>cboos: Nice to have, indeed.
[11:53:06.605369]<cboos>have you seen the latest version of the patch I just posted?
[11:53:36.860135]<rblank>Not yet. Downloading and testing...
[11:54:24.194571]<cboos>Ok, fine. If it's faster and if you're OK with the small change to WikiPageNames this introduces, I'll take this version.
[11:55:29.089676]<rblank>Slightly faster, 1.23s compared to 1.27 for the other version. What's the change?
[11:55:40.836978]<cboos>less characters ;-)
[11:55:59.755139]<cboos>you mean, to WikiPageNames or to the regexp?
[11:56:32.511865]<rblank>To the WikiPageNames
[11:56:39.448086]<cboos>less characters in the regexp, that's why it's faster, and for the WikiPageNames things like Sub/Page are now valid
[11:56:54.578834]<cboos>previously you needed WikiName/Page
[11:56:56.448379]<rblank>Well that's actually an advantage
[11:57:13.229233]<rblank>Oh, I see. Not a huge difference IMO
[11:57:22.444902]<cboos>right, this also didn't hurt me as being wrong ;-)
[11:57:56.651854]<cboos>I'll add some tests and commit it
[11:57:59.792970]<rblank>One question: shouldn't the (?=:) part be *before* the optional version and anchor?
[11:59:09.327460]<cboos>good point, perhaps, I'll make some tests
[12:00:19.699080]<rblank>cboos: All tests still pass if you put that line right after the "# wiki words" line
[12:00:34.323521]<rblank>But maybe we don't have tests for the edge cases.
[12:00:36.841159]<cboos>same speed?
[12:01:23.664698]<rblank>1.22 instead of 1.23, i.e. not significant.
[12:01:29.074342]<cboos>I'm wondering about the termination of anchors... though XML_NAME is restrictive enough
[12:01:47.907568]<cboos>hey but it's BETTER! ;-)
[12:02:22.068049]<rblank>If you consider a measurement uncertainty of about 5%, not really...
[12:02:35.465366]<cboos>run it a few times, keep the lower
[12:02:47.258489]<rblank>That's what I did. The higher ones were around 1.47.
[12:04:18.554463]<rblank>On a side note: 16 tickets to go for 0.12! Do we still intend to fix them all? I.e. #2299 I'm not sure we have an actual solution.
[12:06:20.822684]<cboos>if mgood is around...
[12:06:25.705588]<rblank>Nope.
[12:06:51.937858]<cboos>I mean #2299 is typically the stuff I would not touch... at least not close to a release
[12:06:57.573311]<rblank>Not here at PyCon, at least. Haven't seen him on IRC lately either.
[12:07:05.373487]<rblank>Me neither.
[12:07:20.163411]<cboos>ok, so -> next-major
[12:07:24.208021]<rblank>Yep
[12:07:29.001424]<misc>how long last pycon ? I have been the impression that you are there since 1 week :)
[12:09:38.896445]<cboos>rblank: did you test 0.11-stable recently?
[12:09:56.610563]<rblank>cboos: No, not recently.
[12:10:15.336597]<rblank>misc: Sprints until Thursday
[12:10:36.571143]<cboos>after my commit for #9025, could you run some tests? If all looks fine for me & you, I'll prepare a rc1
[12:10:44.700471]<rblank>cboos: Ok
[12:23:34.984281]<kcunning>I have what may be a very silly question.
[12:23:40.697905]<kcunning>How do I overwrite index.html? You know, the page that shows available projects.
[12:23:47.108050]<kcunning>I can't seem to overwrite it in my instance, even though I can override any other template.
[12:24:06.992769]<coderanger>kcunning: You need to do silly things, one sec
[12:24:25.043588]<kcunning>Thank you, coderanger!
[12:24:27.046366]*coderanger does that yay-im-still-at-pycon-and-katie-isn't dance :-)
[12:24:35.771087]*kcunning cries
[12:24:44.211846]<kcunning>At least you'll be able to say you contributed to a NASA project!
[12:24:58.951726]<coderanger>kcunning: http://trac.edgewall.org/wiki/TracInterfaceCustomization#ProjectList
[12:25:12.435012]<coderanger>you make a new template, and then set an env variable pointing at it
[12:25:21.696594]<kcunning>Awesome! Thank you!
[12:26:39.836653]<coderanger>kcunning: Do you know when danny is leaving? I need to get some stuff back I left in his room
[12:26:44.046012]<cboos>oops I was about to commit the patch as is on 0.11-stable, remembered at the last minute we still support Python 2.3 there :-(
[12:27:09.532004]<kcunning>Danny is staying for all the sprint days, as far as I know
[12:27:12.445000]<kcunning>I can call him
[12:27:17.470526]<coderanger>eeexcellent
[12:27:20.846274]<coderanger>I'll find him tonight
[12:27:29.441891]<coderanger>thanky :)
[12:29:00.445061]<kcunning>Texted him
[12:29:13.473172]<kcunning>If he doesn't ping me, I'll give him a call in a sec
[12:30:13.897857]<kcunning>And, erm, I will likely need your full name
[12:30:17.824033]<coderanger>kcunning: Noah
[12:30:25.318316]<kcunning>As I know everyone by IRC handles and twitter names
[12:30:32.770737]<coderanger>kcunning: He of the big fuzzy beard
[12:30:47.124329]<kcunning>\o/
[12:31:02.874184]<coderanger>Its much easier to just use that to describe myself
[12:32:08.387852]<kcunning>Heh
[12:36:08.866463]<cboos>rblank: #9025 is closed, 0.11.7dev ready to be tested.
[12:37:11.120260]<cboos>on my side the tests were fine for python23/sqlite, I'll now test the other db backends
[12:38:15.127669]<|rt|>Hey guys I'm having an issue setting the logo in trac...I'm using mod_wsgi mounting at the root of the site
[12:38:31.236501]<cboos>hey osimons, at the right time ;-)
[12:38:43.477940]<osimons>oh. what's up?
[12:38:52.454776]<gozerbot>trac: Ticket #9025 (Trac hangs with certain text) closed - <http://trac.edgewall.org/ticket/9025#comment:9> || Changeset [9252]: 0.11.7dev: more straightforward regular expression for detecting ... - <http://trac.edgewall.org/changeset/9252>
[12:38:56.334088]<cboos>osimons: we're about to release 0.11.7rc1, we'd like to test it a bit
[12:38:57.787834]<|rt|>looking at the source the path it wants to use for the css and images and such are in the python directory under chrome
[12:39:09.016073]<cboos>osimons: would you be ok if I upgrade Bitten's Trac instance?
[12:39:16.668366]<|rt|>how do I get it to use the htdocs file in my trac project instance
[12:39:39.310948]<coderanger>|rt|: src = site/foo.png
[12:39:55.096228]<|rt|>yeah and then it looks in /chrome/site/foo.png
[12:40:01.018101]<|rt|>which is in the python folder somewhere
[12:40:28.663510]<coderanger>No, that it $ENV/htdocs/foo.png
[12:40:50.297515]<|rt|>ah ok...let me have a look at my wsgi script to see if that's set right
[12:41:02.989288]<coderanger>hmm?
[12:41:31.776582]<|rt|>seems to have the correct TRAC_ENV set
[12:42:56.379551]<osimons>oh, cboos. holiday in the mountains with family... i can try against my local development stuff as a quick check, but i won't be touching production via iphone tethering...
[12:43:00.405679]<|rt|>where else would I check the $ENV?
[12:43:19.644668]<coderanger>|rt|: I have no idea what you are talking about
[12:43:26.331691]<coderanger>Just put the file in your trac env
[12:43:30.032395]<coderanger>inside the htdocs folder
[12:43:36.759512]<|rt|>yeah did that
[12:43:45.363791]<coderanger>and set [header-logo].src to = site/filename.png
[12:43:49.264585]<|rt|>or tried that I shold say
[12:44:15.404117]<coderanger>Did you get an error?
[12:44:25.974397]<coderanger>you may ned to chmod the file so apache can read it
[12:44:47.299998]<|rt|>well it just doesn't show up...and the source says it's looking in /chrome/sites/logo.png
[12:44:59.400038]<coderanger>And if you open the URL?
[12:45:14.957985]<|rt|>which is in the site packages folder
[12:45:20.631311]<coderanger>No, it isn't
[12:45:49.476023]<coderanger>chrome/site is mapped inside the environment folder
[12:45:49.862301]<|rt|>c:\python26\lib\site-packages\trac\htdocs
[12:46:03.862096]<coderanger>No, that folder is mapped as chrome/common
[12:46:05.519982]<|rt|>i'm looking at it right now..and that's the only location of the trac.css file on my system
[12:46:27.161364]<coderanger>yes, and you will see the URL is chrome/common/trac.css
[12:46:28.352123]<|rt|>so that's the only possibility in this case....this may be due to a misconfiguration but that's where it is looking currently
[12:46:45.718779]<coderanger>Where did you put your file that you think is $env/htdocs?
[12:46:48.401725]<coderanger>full path
[12:46:51.709377]<pacopablo>|rt|: install the siteuploadplugin form trac-hacks.org
[12:46:58.313009]<pacopablo>us that to uplod your image
[12:47:06.490785]<pacopablo>s/uplod/upload
[12:47:06.500025]<evil_twin>pacopablo meant: us that to upload your image
[12:47:19.158330]<pacopablo>s/us/use
[12:47:19.167559]<evil_twin>pacopablo meant: use that to uplod your image
[12:47:28.707739]<pacopablo>stupid evil_twin
[12:49:10.542927]<|rt|>file is in C:/Trac/htdocs right now and the logo src in the ini is just set to the file name of the image
[12:49:26.641799]<|rt|>source of the html says it's looking here src="/chrome/common/nexus_logo.png"
[12:49:46.911059]<|rt|>so it has the right file name...but it's looking in common for it
[12:50:06.115116]<coderanger>|rt|: set it to "src = site/nexus_logo.png"
[12:50:54.254685]<|rt|>hey that worked....i guess trac looks for that site keyword
[12:51:05.720634]<|rt|>thanks
[12:51:06.066478]<coderanger>sort of, its a folder mapping
[12:53:11.425323]<|rt|>coderanger: I'm glad that worked I was starting to think I was running into some sort of issue running the wsgi script at the document root
[13:11:23.985723]<coderanger>@seen jborg
[13:11:23.995551]<evil_twin>jborg was last seen on irc.freenode.net at Tue, 23 Feb 2010 06:17:27 +0100, parting #trac: #trac
[13:11:25.428795]<gozerbot>jborg was last seen on irc.freenode.net at Tue, 23 Feb 2010 15:18:47 +0100, parting #trac: #trac
[13:59:57.258286]<cboos>rblank: tests for pg and mysql fine as well
[14:00:07.557502]<rblank>cboos: Same here.
[14:00:25.947370]<cboos>osimons: so OK for an upgrade of Bitten's trac? (I'll upgrade demo-0.11)
[14:00:57.977789]<rblank>cboos: I guess this should be enough for creating a rc1 tarball. Next step is "user testing".
[14:01:16.434238]<cboos>yep
[14:01:27.976558]<osimons>cboos: seems like it - i'll just double-check bitten unittests. 20 seconds.
[14:03:04.597007]<osimons>yay. Ran 249 tests in 8.697s
[14:03:11.283060]<osimons>OK
[14:05:35.121557]<cboos>oh, I didn't try 0.11.7 with Genshi 0.5.x, but OTOH I see no reasons why it shouldn't work anymore ;-)
[14:06:25.661700]<cboos>correction: the Python 2.3.5 tests were using Genshi 0.5.1, so all is fine
[14:08:53.607227]<gozerbot>trac: Changeset [9253]: 0.12dev: ported r9252, CamelCase regexp reworked (#9025). - <http://trac.edgewall.org/changeset/9253> || Ticket #9076 (AttributeError: 'module' object has no attribute 'dgettext') created - <http://trac.edgewall.org/ticket/9076>
[14:38:54.405143]<gozerbot>trac: Ticket #9076 (AttributeError: 'module' object has no attribute 'dgettext') closed - <http://trac.edgewall.org/ticket/9076#comment:1> || Changeset [9254]: Get rid of the Python 2.3 specific syntax of r9253 - <http://trac.edgewall.org/changeset/9254>
[14:51:06.178343]<osimons>rblank: pycon treating you ok? i suppose you are there now?
[14:51:25.209777]<rblank>osimons: Yes, it's great. Lots of great minds, all hacking away :)
[14:51:40.435867]<rblank>osimons: How about you? Having nice holidays?
[14:52:20.452005]<osimons>quite nice - but have spent the last hour fixing RPC plugin following the micro-seconds the other day...
[14:52:51.896504]<rblank>You have a separate branch for 0.12, I hope?
[14:53:00.017384]<osimons>RPC plugin has a few instances of database access, and suddenly it started returning way more hits than expected...
[14:53:11.174058]<osimons>nope - one version to rule them all
[14:53:25.763089]<rblank>That's asking for trouble, if you ask me...
[14:53:44.131586]<rblank>Besides, you shouldn't be accessing the DB directly. That's what the model is for.
[14:53:47.314515]<osimons>got it now - made my own compat from and to_utimestamp that either uses 0.12 or uses the old to_timestamp and to_datetime
[14:54:13.634581]<osimons>rblank: i don't - matt and alec did when it was written many years ago...
[14:54:29.331010]<rblank>osimons: Right. Cookie for you, then ;)
[14:55:14.317338]<osimons>anyway, the model classes are somewhat tuned to trac's own usage. if RPC needs an efficient getRecentChanges, Trac is not too helpful
[14:55:45.102177]<rblank>osimons: While you're at it, you could read the top of macros.html and adapt the FullBlogPlugin accordingly. I'm in the process of removing macros.html altogether :)
[14:56:36.996467]<osimons>oh. darn. and the alterantive is?
[14:56:56.443280]<rblank>Every function is becoming a separate template, to be included with <xi:include>.
[14:57:09.499724]<rblank>But I think we'll leave macros.html in for some more time.
[14:57:17.643789]<rblank>Even if we don't use it anymore.
[14:57:20.989582]<osimons>hmm. don't actually know if i use it
[14:57:25.341000]<osimons>heh.
[14:57:26.835040]<rblank>Yes, you do :)
[14:57:34.139395]<osimons>ah. thanks :-)
[14:57:47.809990]<osimons>again, 0.11 + 0.12 compat will be painful
[14:57:51.625555]<rblank>Anyway, nothing urgent. Just a heads-up.
[14:58:15.923660]<rblank>That's the cost of progress :-/
[14:58:36.798620]<rblank>TBH, I think having separate branches is probably the simplest solution.
[14:59:14.681078]<osimons>yeah, but again not really... not really so many things that differ from 0.11 to 0.12
[15:00:03.350188]<osimons>and, plugins may branch and maintain release branches for other purposes than trac compatibility. it can be messy.
[15:00:56.290990]<osimons>like bitten for instance, 0.6 vs 0.7 will be because of internal bitten changes - add 0.11 and 0.12 to the mix and it becomes difficult for users
[15:04:37.955001]<rblank>osimons: Oh, I hadn't thought of that. Synchronized releases would simplify this quite a lot, then.
[15:08:52.780175]<cboos>rblank: the 0.11.7rc1 files under http://ftp.edgewall.org/pub/trac/
[15:08:54.607112]<gozerbot>trac: Changeset [9259]: Tagging trac-0.11.7rc1 - <http://trac.edgewall.org/changeset/9259> || Changeset [9258]: Damn... - <http://trac.edgewall.org/changeset/9258> || Changeset [9257]: Put a better exception on loading a bad Environment. - <http://trac.edgewall.org/changeset/9257> || Changeset [9256]: Tagging trac-0.11.7rc1 - <http://trac.edgewall.org/changeset/9256>
[15:08:59.611255]<gozerbot> || Changeset [9255]: 0.11.7dev: preparing rc1 - <http://trac.edgewall.org/changeset/9255>
[15:09:15.298128]<rblank>cboos: Cool. I'm almost done removing macros.html.
[15:09:19.592654]<cboos>rblank: I just don't have time to update TracDownload
[15:09:23.286324]<cboos>bye!
[15:10:51.201073]<osimons>rblank: will you make macros.html to instead xi:include the various functions? so things won't break on release?
[15:12:12.940683]<rblank>osimons: Ooh, that's a good idea!
[15:12:28.671797]*osimons needs a compat.html template...
[15:12:37.690136]<rblank>I intended to just leave it as-is, but that would be code duplication.
[15:37:07.347050]<rblank>osimons: Still around?
[15:37:20.876332]<osimons>yup, remy
[15:38:02.577040]<rblank>cboos put 0.11.7rc1 packages on ftp.edgewall.com, but four files have bad permissions, i.e. are not accessible. Would you mind adjusting the permissions?
[15:38:28.540530]<rblank>osimons: The ones that are not accessible are the win32 and zip variants.
[15:39:49.597592]<osimons>let's see what i find....
[15:40:00.519392]*osimons looks for the operations manual again...
[15:45:49.723215]<osimons>rblank: think i got it.
[15:46:43.076160]<rblank>osimons: Works great, thanks a bunch
[16:08:55.957539]<gozerbot>trac: TracDownload edited - <http://trac.edgewall.org/wiki/TracDownload?version=113>
[16:17:04.952901]<pacopablo>rapha: you still around?
[17:08:57.407521]<gozerbot>trac: Ticket #9077 (Release Trac 0.11.7) created - <http://trac.edgewall.org/ticket/9077> || Ticket #2901 (Apache crashes when used with modpython and LDAP authentication) closed - <http://trac.edgewall.org/ticket/2901#comment:6>
[17:38:59.197522]<gozerbot>trac: Ticket #7824 (Wiki history diff changes revisions after pressing Update button) closed - <http://trac.edgewall.org/ticket/7824#comment:4> || Changeset [9261]: wiki: Fixed a bug in the wiki diff view where changing the diff parameters ... - <http://trac.edgewall.org/changeset/9261> || Ticket #6374 (get rid of macros.html in favor of xi:include'd snippets) closed
[17:39:04.258221]<gozerbot> - <http://trac.edgewall.org/ticket/6374#comment:10> || Changeset [9260]: rendering: Removed all uses of the macros in `macros.html` and ... - <http://trac.edgewall.org/changeset/9260>
[18:09:02.145390]<gozerbot>trac: Ticket #6337 (Timeline "days back" value is not propagated to build the RSS format link.) closed - <http://trac.edgewall.org/ticket/6337#comment:6>
[20:39:01.111952]<gozerbot>trac: Ticket #9078 (todo) created - <http://trac.edgewall.org/ticket/9078>
[21:10:12.910458]<nombrecinq>is there a job board for trac?
[21:10:30.802906]<nombrecinq>i'd like to hire a consultant to get email -> ticket functionality working
[21:10:36.013026]<nombrecinq>like pop3
[21:28:03.066307]<coderanger>nombrecinq: Not directly, there is http://trac.edgewall.org/wiki/CommercialServices
[21:43:36.381300]<nombrecinq>thanks coderanger
[22:09:03.125447]<gozerbot>trac: Changeset [9262]: Support for session management in `trac-admin`. See #1347 - <http://trac.edgewall.org/changeset/9262>
[22:39:03.834855]<gozerbot>trac: Ticket #1347 (remove old authenticated user sessions) closed - <http://trac.edgewall.org/ticket/1347#comment:26>