| [00:21:48] |
* |
tuxipuxi has quit IRC |
| [00:32:17] |
* |
imajes has joined #trac |
| [01:54:50] |
* |
Dnumgis has joined #trac |
| [01:55:13] |
<Dnumgis> |
howdy, anyone here able to tell me how i enable ticket closing by commit message? |
| [02:08:56] |
<athomas> |
trac-post-commit-hook in the contrib directory |
| [02:13:37] |
<Dnumgis> |
thanks |
| [02:16:11] |
* |
stevegt has quit IRC |
| [02:16:29] |
* |
stevegt has joined #trac |
| [02:29:17] |
<Dnumgis> |
now I just need to find a bug, file a ticket for it and commit a fix to see if this works |
| [02:47:32] |
* |
tuxipuxi has joined #trac |
| [02:51:37] |
<Dnumgis> |
juba! worked (after some minor fiddeling) |
| [02:51:44] |
<Dnumgis> |
athomas: thanks a bunch |
| [03:01:25] |
* |
Ryan_ has joined #trac |
| [03:01:36] |
* |
Ryan_ has left #trac |
| [03:04:09] |
<jave> |
I would like to help my brother set up a trac instance to handle writing a oowriter in word format. im unsure about the best way to do it with trac. any hints? |
| [03:04:36] |
<jave> |
like, should the report be handled as a file in subversion, or as attachments? |
| [03:04:45] |
<athomas> |
what report? |
| [03:05:09] |
<jave> |
ah, only a word document, not a trac report |
| [03:05:27] |
<jave> |
some people need to collaborate on writing a word document |
| [03:05:35] |
<athomas> |
ah |
| [03:05:53] |
<jave> |
with some kind of version control, and being able to give out trac tasks |
| [03:06:01] |
<athomas> |
probably svn, as attachments are not versioned |
| [03:06:26] |
<jave> |
ah ok |
| [03:06:51] |
<jave> |
but then next question, is there some kind of web svn client that can be integrated with trac? |
| [03:07:02] |
<jave> |
the inbuilt one just supports browsing, right? |
| [03:07:13] |
<athomas> |
no, not that i'm aware of |
| [03:07:22] |
<athomas> |
there are tickets on PEC requesting that feature |
| [03:07:39] |
<jave> |
ok |
| [03:07:56] |
<jave> |
you dont know any third party software that does it either? |
| [03:08:27] |
<athomas> |
nope |
| [03:09:00] |
<jave> |
my spontaneous reaction was, just use the wiki and format later, but he wanted to be able to draw diagrams and stuff |
| [03:09:26] |
<athomas> |
i'd use the wiki |
| [03:09:29] |
<athomas> |
word sucks :) |
| [03:09:31] |
<jave> |
me too |
| [03:09:37] |
<jave> |
indeed word sucks |
| [03:10:17] |
<jave> |
im having a hard time moving my brother to free software, so ill have to do it step by step |
| [03:10:47] |
<athomas> |
ah, good idea |
| [03:11:04] |
<jave> |
we are using trac to collaborate on music making, which works ok so far, except he doesnt have the linux audio stack on his end |
| [03:17:46] |
* |
tuxipuxi has quit IRC |
| [03:19:32] |
* |
sam`_ has joined #trac |
| [03:24:51] |
<stevegt> |
hey athomas! you still there? |
| [03:25:07] |
<athomas> |
maybe |
| [03:25:21] |
<stevegt> |
heh ;-) |
| [03:25:25] |
<athomas> |
:) |
| [03:26:45] |
* |
sam` has quit IRC |
| [03:27:06] |
<stevegt> |
i'm trying to decide whether to try to build on discussionplugin, start from scratch, or cheat with wiki macros... you done any more with that recently? |
| [03:28:20] |
<athomas> |
not myself, but soundtech and blackhex are collaborating on bringing it up to scratch |
| [03:28:35] |
<athomas> |
you could help them out if you wish to expidite it |
| [03:28:48] |
<stevegt> |
yeah, i saw that too... |
| [03:35:56] |
<stevegt> |
problem is some friends and i are under the gun to set up threaded discussion for a project in days; managed to fight off php, now need to do something quick |
| [03:36:10] |
<stevegt> |
this is for http://www.hobbyspace.com/nucleus/index.php?itemid=1427 in case you're curious |
| [03:40:51] |
<athomas> |
ooh |
| [03:40:55] |
<stevegt> |
yeah |
| [03:41:28] |
<athomas> |
that is very cool |
| [03:42:43] |
<stevegt> |
this is mostly volunteer so far, too -- a bunch of us working out of a shop in Santa Clara, one guy (Dave, unix sysadmin) funding most of it out of his cisco stock options; almost spent all that though ;-) |
| [03:42:47] |
* |
tenshiKur0 has joined #trac |
| [03:43:42] |
* |
neosun has joined #trac |
| [03:44:13] |
<neosun> |
hi all |
| [03:44:19] |
<neosun> |
I've a little question |
| [03:44:29] |
<athomas> |
steve: do you have a working vehicle? |
| [03:44:36] |
<stevegt> |
the guys working on the engines and vehicle have gotten ahead of me, too; now they're getting so much interest in payload flights that we gotta put together some sort of community site for payload users to figure out the elecronics, etc |
| [03:45:13] |
<neosun> |
can I set by milestone for the 'due:' a calendar week? |
| [03:45:19] |
<neosun> |
give's a hack? |
| [03:45:38] |
<stevegt> |
engines work; first vehicle's sitting in the middle of our shop, half the crew's moving down to mojave spaceport in 6 weeks to start flight testing |
| [03:46:18] |
<stevegt> |
neosun: not that i know of |
| [03:46:37] |
* |
psino has joined #trac |
| [03:46:47] |
<neosun> |
:( that's bad for professional use |
| [03:46:57] |
<neosun> |
ok, thx |
| [03:47:14] |
<stevegt> |
neosun: i'm curious, which industry/country? |
| [03:47:15] |
<psino> |
do I have to do anything more than just drop an egg in the plugins/-directory to activate a plugin? |
| [03:47:38] |
<stevegt> |
psino: normally no; which plugin? |
| [03:47:45] |
<psino> |
http://trac-hacks.org/wiki/BuildbotPlugin |
| [03:47:49] |
<neosun> |
stevegt software developing/germany |
| [03:47:58] |
<stevegt> |
ah |
| [03:48:01] |
<athomas> |
@plugins |
| [03:48:01] |
<evil_twin> |
athomas: "plugins" is http://projects.edgewall.com/trac/wiki/TracPlugins |
| [03:48:10] |
<athomas> |
follow the instructions there ^^ |
| [03:48:14] |
<neosun> |
on germany we always use calendar week's for milestones |
| [03:48:58] |
<psino> |
thanks, Ill have a look at it :) |
| [03:50:11] |
<athomas> |
well, that's very cool steve, best of luck |
| [03:50:13] |
<stevegt> |
neosun: makes sense; i was wondering, because US/software at any rate normally is just months or particular dates |
| [03:50:51] |
<neosun> |
:) can I add a hack for doing this? |
| [03:51:09] |
<stevegt> |
athomas: thanks; it's gonna be hectic |
| [03:51:21] |
<stevegt> |
already is getting that way ;-) |
| [03:51:27] |
<athomas> |
heh |
| [03:52:03] |
<athomas> |
which part are you working on? |
| [03:53:59] |
<stevegt> |
neosun: i'd start by looking at the *.cs that renders that page, see what the hdf data item name is, grep for that in the trac code, find where it's set, see if there is any extension point near there that you can intercept and add your own hdf data item next to the original one.... and no, i had no idea hw to do this myself earlier this year -- it's actually not bad ;-) |
| [03:55:22] |
<neosun> |
thx a lot stevegt |
| [03:55:26] |
<stevegt> |
and if there's no easy place to intercept, you need the RequestPostProcessor patch that I'm going to be uploading to trac-hacks later this week; watch for that, because it will let you do arbitrary things to the hdf just before rendering any page |
| [03:55:32] |
<stevegt> |
you're welcome |
| [03:55:42] |
<stevegt> |
s/this week/next week/ |
| [03:57:26] |
<stevegt> |
athomas: this is all going on in office/machine shop/warehouse/data center space that my wife and i originally built out for her dad's business; but i've always been a space geek, invited some friends into the machine shop to work on their own projects, and things kinda got out of hand |
| [03:57:35] |
* |
tenshiKur0 has quit IRC |
| [03:57:43] |
<athomas> |
hehe |
| [03:59:10] |
<stevegt> |
so mainly i'm just the evil landlord, it guy, and source of free advice |
| [03:59:22] |
<stevegt> |
s/it/IT/ |
| [04:01:34] |
<stevegt> |
which is how i wound up saying *noooo!!!* php is a security hole! not on my servers! we're using trac! i think i can get that discussionplugin thing working! i think it's just relying on a patch against an old head! i can fix it, then we're good to go! |
| [04:01:43] |
<stevegt> |
famous last words |
| [04:01:47] |
<athomas> |
hehe |
| [04:01:58] |
<athomas> |
yes, it was missing a lot of functionality |
| [04:02:39] |
* |
neosun has left #trac |
| [04:02:48] |
<stevegt> |
well i didn't actually notice the fact that there are no INSERT or UPDATE statements in there until the next day... ;-) |
| [04:03:10] |
<stevegt> |
doh! |
| [04:05:47] |
<stevegt> |
thing is, since all these guys really need is rudimentary threaded discussion, on a trac site that is only going to be open to payload users, so no spam, low message volume, and so on, i'm thinking that it might be possible to just do this in flat wiki text, adding some sttructure and ui, maybe start with the addcomment macro code |
| [04:05:48] |
* |
sam`_ is now known as sam` |
| [04:06:19] |
<athomas> |
22142 has cloaked infantry |
| [04:06:24] |
<athomas> |
whhops |
| [04:06:39] |
<athomas> |
addcomment is mighty ugly code |
| [04:06:59] |
<athomas> |
and trac doesn't seem to allow POST data through to the macros |
| [04:07:08] |
<athomas> |
so it has to use GET for everything :\ |
| [04:08:04] |
* |
hartym has joined #trac |
| [04:08:20] |
<stevegt> |
i'd write this as a plugin anyway -- i think addcomment's idea of editing the page text with each post is about right for this though |
| [04:10:08] |
<stevegt> |
i realized yesterday that even thread expand/collapse, <next> and <prev> buttons, and so on can even be just selective text hiding on a single large flat page; it's all regexes and a ui |
| [04:10:23] |
<stevegt> |
ugly hack, but it would be quick |
| [04:10:42] |
<athomas> |
cool, sounds good |
| [04:10:58] |
<athomas> |
worth it if you don't have to use php ;) |
| [04:12:02] |
<stevegt> |
i've just spent the last four months writing about 10,000 lines of code for another trac plugin; this one should be *much* more fun |
| [04:12:19] |
<stevegt> |
hmmm |
| [04:12:34] |
<athomas> |
10K lines for a plugin?!?! |
| [04:13:21] |
<stevegt> |
yeah -- once you see it you'll see why |
| [04:14:09] |
<stevegt> |
i did just now do a wc -l though; didn't realize until this moment that trac itself is only 25k lines :-P |
| [04:18:47] |
<stevegt> |
the 10,000 lines is a market plugin; it's an entire double-auction exchange trading system, backend accounting ledger, and hooks into wiki versioning to turn the wiki into a decision market -- have you run across any of the prediction/decision market stuff anywhere yet? |
| [04:19:20] |
<athomas> |
can't say that i have, no |
| [04:19:27] |
* |
dot has joined #trac |
| [04:20:19] |
<stevegt> |
ummm -- look up "google prophit" sometime, for instance |
| [04:20:35] |
<stevegt> |
that's a parallel project |
| [04:22:50] |
<stevegt> |
hmmm |
| [04:23:44] |
<stevegt> |
ah, here: http://googleblog.blogspot.com/2005/09/putting-crowd-wisdom-to-work.html |
| [04:25:20] |
<athomas> |
interesting |
| [04:25:34] |
* |
dot has quit IRC |
| [04:27:07] |
<stevegt> |
you're the first trac person to hear this much detail about this plugin -- but it's why i needed to post the wiki workflow patch a few months ago |
| [04:27:38] |
<athomas> |
aah, it makes sense now :) |
| [04:31:38] |
<stevegt> |
i better go try to get some sleep again -- woke up in the middle of the night; it's 04:30 here |
| [04:31:48] |
<stevegt> |
where are you anyway? |
| [04:33:14] |
* |
cprov has joined #trac |
| [04:34:32] |
<athomas> |
sydney, it's 2130 here |
| [04:35:45] |
<stevegt> |
no wonder i never ran across you in here ;-) |
| [04:35:56] |
<stevegt> |
i think most of you guys are much earlier than me |
| [04:36:31] |
<stevegt> |
at 2130 my time this channel's a ghost town |
| [04:36:56] |
<athomas> |
cmlenz, cboos and eblot are in germany and france |
| [04:37:02] |
<athomas> |
mgood is in the US somewhere though |
| [04:37:34] |
<stevegt> |
i'll have to figure out where/when he is |
| [04:38:42] |
<stevegt> |
anyway, good talking with you finally -- g'night |
| [04:39:06] |
<athomas> |
see ya steve |
| [04:47:53] |
<psino> |
does tracd have to be restarted or something after installing a plugin? |
| [04:49:44] |
<ranq_> |
yeah |
| [04:49:56] |
<ranq_> |
(i guess) |
| [04:51:57] |
<psino> |
damn, I dont have access to the server that trac runs on, only the project-specific filesystem |
| [04:54:15] |
<psino> |
guess I'll have to wait for the admin to wake up |
| [04:55:41] |
* |
Dnumgis has left #trac |
| [05:29:10] |
* |
MasterC has joined #trac |
| [05:32:54] |
* |
cprov is now known as cprov-office |
| [05:35:58] |
* |
TovAre_ has joined #trac |
| [05:36:28] |
* |
_idostyle has joined #trac |
| [05:47:07] |
* |
cprov has joined #trac |
| [05:47:45] |
* |
coderanger_ has joined #trac |
| [05:50:02] |
* |
idostyle has quit IRC |
| [05:50:12] |
* |
_idostyle is now known as idostyle |
| [05:51:05] |
* |
cprov-office has quit IRC |
| [05:55:39] |
* |
hartym has quit IRC |
| [05:58:46] |
* |
stretch has joined #trac |
| [06:15:16] |
* |
mgalvin has joined #trac |
| [06:45:27] |
* |
__doc__ has left #trac |
| [06:52:41] |
* |
mgalvin has quit IRC |
| [06:56:40] |
* |
agile has quit IRC |
| [07:04:31] |
* |
mgalvin has joined #trac |
| [07:19:49] |
* |
danbeck has joined #trac |
| [07:32:29] |
* |
bwmcadams has joined #trac |
| [07:33:49] |
* |
agile has joined #trac |
| [07:42:58] |
* |
whitelynx has quit IRC |
| [07:47:15] |
* |
klasstek has joined #trac |
| [07:48:05] |
* |
lightcap has joined #trac |
| [08:00:51] |
* |
jrsharp has joined #trac |
| [08:17:26] |
* |
sam`_ has joined #trac |
| [08:24:42] |
* |
sam` has quit IRC |
| [08:24:57] |
* |
tuxipuxi has joined #trac |
| [08:29:54] |
* |
halorgium has joined #trac |
| [08:41:27] |
<halorgium> |
yo |
| [09:01:34] |
<coderanger_> |
Alo |
| [09:16:52] |
* |
sam`_ is now known as sam` |
| [09:19:21] |
<halorgium> |
i've got my svn repo setup with multiple projects in a single repo, can trac handle that? |
| [09:19:26] |
<coderanger_> |
Yes |
| [09:19:30] |
<halorgium> |
"separate the projects out"? |
| [09:20:39] |
<coderanger_> |
What do you mean? |
| [09:21:28] |
<halorgium> |
i would like a different wiki/ticketing system for each project |
| [09:22:06] |
<halorgium> |
but i can't seem to find if trac can have a many installs for a single svn repo |
| [09:22:20] |
* |
ChanServ has quit IRC |
| [09:23:06] |
<coderanger_> |
if the repo is at /var/svn/repo and the projects are at /proj1 and /proj2 |
| [09:23:25] |
<coderanger_> |
just have one trac with the svn path of /var/svn/repo/proj1, and another with /var/svn/repo/proj2 |
| [09:23:51] |
<halorgium> |
the trac-admin init env fails though |
| [09:24:01] |
<halorgium> |
something about SVNROOT/format |
| [09:24:06] |
<coderanger_> |
Note that moves or copies between the top-level branches will break things |
| [09:24:15] |
<coderanger_> |
what version of Trac |
| [09:26:41] |
* |
coderanger_ -> Lunch |
| [09:26:57] |
* |
exarkun has joined #trac |
| [09:31:47] |
<exarkun> |
After an update, my rss reader always shows rss items from trac as unread. Does trac generate guids for items improperly or is this because I using a broken rss reader? |
| [09:37:14] |
<idnar> |
what kind of update? |
| [09:38:49] |
* |
ChanServ has joined #trac |
| [09:38:49] |
* |
irc.freenode.net sets mode: +o ChanServ |
| [09:38:56] |
<exarkun> |
an automatic one? |
| [09:39:40] |
<idnar> |
as in your RSS reader fetching the feed again? or something else? |
| [09:40:01] |
<exarkun> |
Yes, that. |
| [09:41:13] |
<idnar> |
hmm |
| [09:41:33] |
<idnar> |
my reader handles it fine |
| [09:42:00] |
<idnar> |
one of the problems is that RSS doesn't really *have* GUIDs |
| [09:42:16] |
<exarkun> |
Er? |
| [09:42:59] |
<exarkun> |
Oh. |
| [09:43:06] |
<exarkun> |
Trac's RSS doesn't have GUIDs. |
| [09:43:12] |
<exarkun> |
Yes, I expect that's a bit of a problem. |
| [09:44:28] |
<exarkun> |
At least that's a nice unambiguous answer. |
| [09:44:50] |
<idnar> |
oh, my bad, RSS 2 does have GUIDs it seems; but yeah, the Trac feed doesn't have 'em |
| [09:46:39] |
<idnar> |
I suppose generating one is a bit tricky |
| [09:48:31] |
<exarkun> |
eh, there is a maximum possible hardness and I think it is pretty low |
| [09:48:36] |
<exarkun> |
anyway, #3167 created |
| [09:48:52] |
<idnar> |
changesets are easy, I guess for wiki changes you can reference the specific version, and ticket changes you can reference the bookmark |
| [09:48:59] |
<idnar> |
but I dunno what you'd do for milestones |
| [09:49:14] |
<idnar> |
maybe just attach the timestamp of the event |
| [09:58:33] |
* |
cendrizzi has joined #trac |
| [10:41:00] |
* |
whitelynx has joined #trac |
| [10:58:58] |
* |
ged has joined #trac |
| [10:59:15] |
* |
exarkun has left #trac |
| [11:21:22] |
* |
Semhirage has quit IRC |
| [11:21:28] |
* |
Semhirage_ has joined #trac |
| [11:27:01] |
<pacopablo> |
morning all |
| [11:30:20] |
<bwmcadams> |
blecch |
| [11:30:45] |
<bwmcadams> |
i just changed our trac instances over to authenticating against activedirectory |
| [11:30:47] |
<bwmcadams> |
<shudder> |
| [11:31:02] |
<pacopablo> |
nasty |
| [11:36:31] |
* |
steveg1 has joined #trac |
| [11:38:16] |
* |
stevegt has quit IRC |
| [11:40:28] |
* |
MasterC has quit IRC |
| [11:41:25] |
* |
MasterC has joined #trac |
| [11:49:56] |
* |
Semhirage_ is now known as Semhirage |
| [11:58:46] |
* |
MasterC has quit IRC |
| [11:59:13] |
* |
MasterC has joined #trac |
| [12:00:39] |
* |
imajes has quit IRC |
| [12:00:45] |
<coderanger_> |
Any JS/CSS gurus around? |
| [12:01:13] |
* |
datafirm has quit IRC |
| [12:15:41] |
* |
imajes has joined #trac |
| [12:48:38] |
<pacopablo> |
what are you trying to do coderanger_ ? |
| [12:49:19] |
<coderanger_> |
http://www.cs.rpi.edu/~kantrn/fader.html |
| [12:49:26] |
<coderanger_> |
Works fine in Firefox |
| [12:49:30] |
<coderanger_> |
Fails miserably in IE |
| [12:49:49] |
<pacopablo> |
good thing I use FF :) |
| [12:52:24] |
<coderanger_> |
It looks like you need the layout set at the point the element is created for filters to work |
| [13:03:15] |
<coderanger_> |
Gah, I dont understand why this doesn't work |
| [13:03:25] |
<pacopablo> |
simple, IE sucks |
| [13:03:56] |
<pacopablo> |
wish I could help more, but I'm no JS nor CSS guru |
| [13:04:02] |
<coderanger_> |
Oh wait, I think I know |
| [13:04:03] |
<coderanger_> |
hmm |
| [13:04:44] |
<pacopablo> |
how does one go about killing a thread in python? |
| [13:13:53] |
* |
chandlerc has quit IRC |
| [13:14:55] |
* |
chandlerc has joined #trac |
| [13:43:33] |
* |
peppelorum has quit IRC |
| [13:46:35] |
* |
coderanger_ has quit IRC |
| [13:52:54] |
* |
robs0r has joined #trac |
| [14:06:12] |
* |
sam`_ has joined #trac |
| [14:09:01] |
* |
robs has quit IRC |
| [14:09:02] |
* |
robs0r is now known as robs |
| [14:11:08] |
* |
mgalvin has quit IRC |
| [14:12:18] |
* |
jrsharp has quit IRC |
| [14:13:24] |
* |
sam` has quit IRC |
| [14:18:58] |
* |
steveg1 has quit IRC |
| [14:20:18] |
* |
TovAre_ has quit IRC |
| [14:30:56] |
<pacopablo> |
@faq tracd |
| [14:30:56] |
<evil_twin> |
pacopablo: "tracd" is http://projects.edgewall.com/trac/wiki/TracStandalone |
| [14:32:05] |
<pacopablo> |
I thought there was a way to daemonize tracd |
| [14:32:16] |
<pacopablo> |
was I dreaming? |
| [14:32:43] |
<pacopablo> |
no, I wasn't |
| [14:32:45] |
<bwmcadams> |
there is. |
| [14:32:51] |
<pacopablo> |
it's just not available in windows |
| [14:32:55] |
<bwmcadams> |
ah |
| [14:33:02] |
<bwmcadams> |
who really uses windows , anyway? |
| [14:33:09] |
<pacopablo> |
i wish I could say I didn't |
| [14:33:13] |
<bwmcadams> |
:) |
| [14:33:15] |
<pacopablo> |
or rather didn't need to |
| [14:33:24] |
<bwmcadams> |
that's ok.. I had to slap kerberos -> activedirectory auth on top of my trac installs today |
| [14:33:30] |
<bwmcadams> |
to reduce duplication of usernames & passwords... |
| [14:33:52] |
<bwmcadams> |
other than that i manage to avoid touching windows |
| [14:34:01] |
<pacopablo> |
how'd you do that? |
| [14:34:04] |
<pacopablo> |
mod_kerb? |
| [14:34:08] |
<pacopablo> |
or something like that? |
| [14:34:08] |
<bwmcadams> |
1 sec |
| [14:34:21] |
<bwmcadams> |
http://support.microsoft.com/?id=555092 |
| [14:35:26] |
<pacopablo> |
ahh, mod_auth_kerb :) |
| [14:35:35] |
<pacopablo> |
that's a nice little article |
| [14:35:41] |
<bwmcadams> |
ya... it worked too |
| [14:35:42] |
<bwmcadams> |
flat otu |
| [14:35:48] |
<bwmcadams> |
i would prefer to do ldpa |
| [14:35:53] |
<bwmcadams> |
but there are no good docs out there for it |
| [14:36:11] |
<bwmcadams> |
(I wanted to hijack ldap groups so we can define who can access what pages in exchange, rather than maintaining a separate authz list) |
| [14:37:30] |
<pacopablo> |
I use mod_ldap for auth |
| [14:37:38] |
<bwmcadams> |
against activedirectory? |
| [14:37:40] |
<pacopablo> |
but that's just against a plain OpenLDAP tree |
| [14:37:42] |
<bwmcadams> |
ah.. |
| [14:37:43] |
<bwmcadams> |
ya |
| [14:38:04] |
<pacopablo> |
can AD do simple binds? |
| [14:38:17] |
<bwmcadams> |
supposedly |
| [14:38:20] |
<bwmcadams> |
but i couldn't get it towork |
| [14:38:24] |
<pacopablo> |
ahh |
| [14:38:27] |
<bwmcadams> |
http://thomas-howard.com/Reference/Articles/Apache+AD/ |
| [14:44:56] |
* |
sam`_ is now known as sam` |
| [14:46:47] |
* |
bwmcadams has left #trac |
| [14:48:34] |
<pacopablo> |
hot damn! |
| [14:48:38] |
<pacopablo> |
tracd as a service :) |
| [14:50:29] |
* |
stevegt has joined #trac |
| [15:15:50] |
<stevegt> |
anyone here who knows anything about why there's no db.commit() before db.close() in the wiki save code path? |
| [15:16:16] |
<pacopablo> |
probably because the close call does a commit |
| [15:17:33] |
* |
Riccc has joined #trac |
| [15:17:40] |
<pacopablo> |
actually, maybe it doesn't |
| [15:17:59] |
<stevegt> |
hey, thanks pacopablo -- i was afraid it was too late in the day; umm, in my case, close() is not committing; i don't see ... okay, you just got the point ;-) |
| [15:18:20] |
<stevegt> |
i can see no reason why it works at all |
| [15:21:44] |
<stevegt> |
and in my case (after applying a patch for something else) new pages aren't getting saved to the db -- i'm sitting here stepping through it in pdb; the INSERT's running, everything's fine until after that close() |
| [15:21:58] |
<pacopablo> |
what patch was it? |
| [15:23:09] |
<stevegt> |
it's my wikiworkflow patch -- i don't know when i broke this, but i also can't see why it worked in the first place |
| [15:23:34] |
<stevegt> |
lemme step through again, this time stepping into that close()... |
| [15:23:52] |
<pacopablo> |
well, I can see why it worked in the first place |
| [15:24:00] |
<pacopablo> |
http://projects.edgewall.com/trac/browser/trunk/trac/wiki/model.py#L131 |
| [15:24:07] |
<pacopablo> |
that's where it does the commit |
| [15:24:22] |
<pacopablo> |
http://projects.edgewall.com/trac/browser/trunk/trac/wiki/web_ui.py#L209 |
| [15:24:30] |
<pacopablo> |
that is where it calls save |
| [15:24:44] |
* |
danbeck has quit IRC |
| [15:25:07] |
<pacopablo> |
granted, if you're not working against trunk, the I'm not looking at the same code |
| [15:29:31] |
<stevegt> |
yeah, i know where the _do_save and all is called; wikiworkflow patch is near there, and in model.py itself, so at least i'm familiar with that area of the code; i've been eyeing that lack of an explicit commit in trac.web.main.dispatch_request() for the last two weeks though, and i still don't get it |
| [15:29:50] |
<stevegt> |
and so now of course that's where i'm hitting a bug ;-) |
| [15:30:25] |
<pacopablo> |
well, I think the thing is that if it hits that close() then everything sould already be committed |
| [15:30:39] |
<pacopablo> |
at least that's what I understand |
| [15:31:03] |
<pacopablo> |
if there is something in the middle of a transaction, we don't want to commit it, because we don't know what state it's in |
| [15:31:18] |
<stevegt> |
by what? there's no commit() called before then, due to those handle_ta checks in model.py....hmmmm |
| [15:32:02] |
<pacopablo> |
but the handle_ta is only set to false (skip commit) is someone passes a db connection to the save function |
| [15:32:12] |
<pacopablo> |
which the main wiki web_ui.py doesn't |
| [15:32:29] |
<stevegt> |
huh? is that where i broke it? (looking...) |
| [15:32:42] |
<stevegt> |
i thought web_ui passes it... |
| [15:32:53] |
<pacopablo> |
so when trac.wiki.web_ui calls trac.wiki.model.WikiPage.save, handle_ta will always be set to true due to the lack of a db connection being passed |
| [15:32:58] |
<pacopablo> |
http://projects.edgewall.com/trac/browser/trunk/trac/wiki/web_ui.py#L209 |
| [15:33:10] |
<pacopablo> |
unless I'm blind, I don't see it passing a db connection there |
| [15:33:31] |
<pacopablo> |
I see authoer, comment, and remote_addr |
| [15:34:02] |
<stevegt> |
HOLY HECK -- you nailed it |
| [15:34:20] |
<stevegt> |
i've been staring at this, including the diff, for two days |
| [15:34:27] |
<pacopablo> |
hehe |
| [15:34:35] |
<pacopablo> |
it's alwyas nice to get a fresh pair of eyes |
| [15:34:51] |
<stevegt> |
i had added a db=db to that .save() call for some reason --- ouch ;-) |
| [15:35:06] |
* |
Riccc`Miam has quit IRC |
| [15:35:19] |
<stevegt> |
fresh pair of eyes works even when you can't see the broken code ;-) |
| [15:35:27] |
<pacopablo> |
hehe :) |
| [15:35:53] |
<pacopablo> |
I've been working on my ESP |
| [15:36:13] |
<stevegt> |
THANK YOU -- glad to have finally run into you in here; i'm west coast U.S., most people are gone by the time I'm here |
| [15:36:22] |
<stevegt> |
where are you anyway? |
| [15:36:36] |
<pacopablo> |
Oregon |
| [15:37:12] |
<stevegt> |
WOW! i was just telling my wife "we need to get more people from the western US to work on trac" ;-) |
| [15:37:29] |
<pacopablo> |
well, I'm a trac adict :) |
| [15:37:31] |
<stevegt> |
we're down in the SF bay area |
| [15:38:02] |
<stevegt> |
trac is soooo nice to work in |
| [15:38:07] |
<pacopablo> |
agreed |
| [15:38:44] |
<pacopablo> |
I'm trying to hack together tracd as a windows service at the moment |
| [15:39:11] |
<pacopablo> |
unfortunately it's doing nothing to erase my hatred for windows |
| [15:39:22] |
<stevegt> |
i'm looking forward to posting the work we've done with it here these last few months -- totals about 10,000 lines of plugin code |
| [15:39:23] |
* |
cendrizzi has quit IRC |
| [15:39:29] |
<stevegt> |
you have my sympathy ;-) |
| [15:39:36] |
<pacopablo> |
that's a lot of code :) |
| [15:39:52] |
<stevegt> |
considering trac itself is only 25,000 lines |
| [15:40:02] |
* |
pacopablo wonders how many lines of code the TracBlogPlugin is right now |
| [15:40:42] |
<stevegt> |
uh oh -- getting an unmaskable interrupt from my wife -- gotta unplug and drive back to our office; should be back in a little while though |
| [15:40:47] |
<stevegt> |
umm... |
| [15:40:52] |
<pacopablo> |
hehe |
| [15:40:53] |
<pacopablo> |
later |
| [15:43:08] |
<pacopablo> |
TracBlogPlugin weighs in at a paltry 1397 lines of code |
| [15:43:19] |
<pacopablo> |
and that's including templates and CSS |
| [15:50:09] |
* |
galis has quit IRC |
| [15:50:23] |
* |
galis has joined #trac |
| [15:52:45] |
* |
Blackhex has joined #trac |
| [15:53:06] |
<Blackhex> |
hello |
| [15:53:58] |
<pacopablo> |
ello |
| [15:58:45] |
* |
jcw9 has quit IRC |
| [16:03:32] |
* |
sam` is now known as sam|dodo |
| [16:03:44] |
* |
sam|dodo is now known as sma` |
| [16:04:10] |
* |
sma` is now known as sam` |
| [16:04:15] |
* |
stevegt has quit IRC |
| [16:04:16] |
* |
cprov has quit IRC |
| [16:15:57] |
* |
galis has quit IRC |
| [16:19:55] |
<pacopablo> |
hehe, windows evilness >:) |
| [16:20:24] |
<pacopablo> |
now, what's the best way to supply parameters to a windows service? |
| [16:31:31] |
* |
agile has quit IRC |
| [16:37:07] |
* |
Blackhex has left #trac |
| [16:54:03] |
* |
klasstek has quit IRC |
| [16:54:42] |
* |
tinjaw has joined #trac |
| [16:54:55] |
<tinjaw> |
hello |
| [16:55:20] |
<tinjaw> |
is there a working WebAdmin for 0.9.5? |
| [ |
