| [00:04:36] |
* |
Guardian has quit IRC |
| [00:17:25] |
* |
flox has joined #trac |
| [00:22:28] |
* |
flox has quit IRC |
| [00:22:47] |
* |
Guardian has joined #trac |
| [00:31:43] |
* |
peppelorum has joined #trac |
| [00:46:32] |
* |
blake_ has joined #trac |
| [00:47:23] |
* |
flox has joined #trac |
| [01:05:04] |
* |
^BuGs^ has quit IRC |
| [01:12:56] |
* |
x^k_ has joined #trac |
| [01:18:17] |
* |
cmlenz has joined #trac |
| [01:23:32] |
* |
x^k has quit IRC |
| [01:25:38] |
* |
x^k_ is now known as x^k |
| [01:29:00] |
* |
shawn has quit IRC |
| [01:49:14] |
* |
cookin_ has joined #trac |
| [02:05:32] |
<cboos> |
cmlenz: hi, just seen you closed #3419 |
| [02:05:43] |
<cmlenz> |
yeah |
| [02:05:52] |
<cboos> |
I think you're actually "completely mistaken" on this one ;) |
| [02:06:16] |
<cboos> |
the ticket was talking about entering straight numbers |
| [02:06:20] |
<cboos> |
like "123" |
| [02:06:34] |
<cboos> |
and then get links to either r123, #123, {123} and what not |
| [02:06:34] |
<cmlenz> |
yes, as is #1268 |
| [02:06:37] |
<cmlenz> |
"e.g. when I input 614 in the search field, the changeset 614 should be among the top search results." |
| [02:07:00] |
<cboos> |
Ok, so I'm mistaken on #1268 ;) |
| [02:07:51] |
<cmlenz> |
maybe you only looked at one aspect raised in the comments |
| [02:08:00] |
<cboos> |
yeah, I've mistaken #1268 for the one which was just about "r123" |
| [02:08:29] |
<cboos> |
ok, so I said nothing |
| [02:08:37] |
<cboos> |
too bad the IRC logs are now archived ;) |
| [02:08:40] |
<cmlenz> |
:-) |
| [02:08:59] |
<cboos> |
on a more serious front, I'm now 95% done with the reports |
| [02:09:12] |
<cboos> |
that was a bit challenging ... got to rewrite a lot |
| [02:09:52] |
<cboos> |
it's still messy, but much less so, I think |
| [02:10:15] |
* |
tuxipuxi has joined #trac |
| [02:10:36] |
<cmlenz> |
I briefly looked at the patch from yesterday |
| [02:11:25] |
<cmlenz> |
one thing I noticed was that report_edit etc are not standalone pages... they appear to be included from some main template based on the action or sth, right? |
| [02:11:37] |
<cboos> |
exact. |
| [02:11:51] |
<cboos> |
This is for trying out a different approach than, say, the wiki module |
| [02:11:53] |
<cmlenz> |
I'd rather see the ReportModule just render the correct template directly |
| [02:12:04] |
<cmlenz> |
ah |
| [02:12:37] |
<cmlenz> |
well, I think I prefer the wiki way ;-) |
| [02:13:01] |
<cboos> |
Well, the point is that the report subpages have all the same ctxt nav and footer, so I was interested to see how it would look like with a "webadmin" like approach... |
| [02:13:59] |
<cboos> |
What I wanted to do is to have that, and then raise the question on the Trac-Dev ML, saying "look, we have those 2 approaches, what's best?" |
| [02:14:45] |
<cboos> |
I see pros & cons for both, so it was only for experimenting and getting feedback |
| [02:15:01] |
<cboos> |
not the "definitive" word on it, of course ;) |
| [02:15:26] |
* |
sgorilla1 has joined #trac |
| [02:17:16] |
<cmlenz> |
understood |
| [02:17:20] |
<sgorilla1> |
hello |
| [02:17:38] |
<cmlenz> |
we'll have to refactor the ctxtnav stuff anyway to make it extensible/cleaner |
| [02:17:54] |
<cmlenz> |
so reusing a common ctxtnav across a set of templates will be less of a problem at that point |
| [02:18:09] |
<sgorilla1> |
i have set up stunnel and got tracd to work standalone |
| [02:18:21] |
<sgorilla1> |
and set up user accounts using basic auth |
| [02:18:47] |
<sgorilla1> |
im not sure how to change css for page |
| [02:19:09] |
<sgorilla1> |
using the site_css.cs |
| [02:19:24] |
<sgorilla1> |
do i need to overide css using the clearsilver template? |
| [02:20:09] |
<sgorilla1> |
clearsilver template commands.. |
| [02:20:34] |
<cboos> |
cmlenz: about the ctxtnav redesign, I know someone you could ask for help (he'll give you some ARCHITECTURE ideas ;) ) |
| [02:20:52] |
<cmlenz> |
heh |
| [02:21:02] |
<cmlenz> |
sgorilla1: you just put your CSS style rules in that file |
| [02:21:15] |
<cmlenz> |
or do an @import |
| [02:21:20] |
<sgorilla1> |
cmlenz: i did not find anything about css style rules in the documenation |
| [02:21:24] |
<sgorilla1> |
I rather do the former |
| [02:21:46] |
<sgorilla1> |
oh nm, the docs on the site are newer than the code |
| [02:22:02] |
* |
cboos thinks the reports are starting to look good... |
| [02:22:16] |
<sgorilla1> |
the reports look awesome =) |
| [02:22:22] |
<cmlenz> |
I'm not sure what you're asking for... look at the CSS files that come with Trac, inspect the markup Trac generates, etc, to find out what styles you want to override |
| [02:22:37] |
<sgorilla1> |
trac can probably be made more general purpose, like a project manager |
| [02:23:06] |
<sgorilla1> |
cmlenz: i just want to change the background color, the simplest way possible |
| [02:24:12] |
<sgorilla1> |
body { background #e1dfce;} |
| [02:24:12] |
<sgorilla1> |
|
| [02:24:25] |
<sgorilla1> |
i did that, but nothing got effected |
| [02:25:09] |
<cmlenz> |
make sure you're outside of the <? ?> comment block |
| [02:25:27] |
<sgorilla1> |
oh im inside of that block |
| [02:26:11] |
<sgorilla1> |
i put it outside no change, do i need to restart trac? |
| [02:27:05] |
<cmlenz> |
look at the page source in the browser to see if what you put in site_css is actually included in the output |
| [02:27:31] |
<sgorilla1> |
its not |
| [02:27:58] |
<cmlenz> |
then you're doing something wrong ;-) |
| [02:28:04] |
<cmlenz> |
you don't need to restart the server |
| [02:28:09] |
<sgorilla1> |
possibly something not set up right in the trac.ini |
| [02:28:21] |
<sgorilla1> |
do i need to set the htdocs_location ? |
| [02:28:32] |
<sgorilla1> |
or base_url ? |
| [02:28:43] |
<cmlenz> |
no |
| [02:29:08] |
* |
Squirk has quit IRC |
| [02:29:37] |
<sgorilla1> |
ok its embedding it |
| [02:29:53] |
<sgorilla1> |
its working i believe, but its not overriding the previous value from trac.css |
| [02:29:56] |
<sgorilla1> |
which has teh background |
| [02:30:45] |
<sgorilla1> |
what is the correct way of overriding the background color from trac.css? |
| [02:31:46] |
<cmlenz> |
body { background: #e1dfce; } |
| [02:31:49] |
<cmlenz> |
(note the colon) |
| [02:32:09] |
<sgorilla1> |
body { background #e1dfce;} |
| [02:32:14] |
<sgorilla1> |
thats what i have' |
| [02:32:22] |
<cmlenz> |
(note the colon) :-P |
| [02:32:29] |
<sgorilla1> |
oh =) |
| [02:33:14] |
<sgorilla1> |
how can you get it to force you to authenicate when you just try to load the page using standalone trac? |
| [02:33:22] |
<sgorilla1> |
or is that even possible in standalone trac? |
| [02:34:30] |
<sgorilla1> |
like make you login |
| [02:34:42] |
<cmlenz> |
that's not possible without some surgery |
| [02:34:52] |
<sgorilla1> |
i don't mind surgery |
| [02:35:02] |
<sgorilla1> |
you think people would want something like that? |
| [02:35:44] |
<cmlenz> |
maybe |
| [02:36:26] |
<sgorilla1> |
what components should i modify to add that? |
| [02:37:16] |
<sgorilla1> |
trac.web.standalone |
| [02:37:47] |
<cmlenz> |
yes |
| [02:37:55] |
<sgorilla1> |
ok i will add that now |
| [02:41:42] |
<sgorilla1> |
i guess easiest way to do it is look to see if the user is logged in, if not ask for authentication |
| [02:42:19] |
<sgorilla1> |
then kick them out to some access denied page |
| [02:42:22] |
<cmlenz> |
there's class called AuthenticationMiddleware in there (IIRC) which you should probably being look at |
| [02:42:43] |
<sgorilla1> |
ok im looking at that now |
| [02:43:07] |
<sgorilla1> |
i guess the object it gets sent to it includes path info |
| [02:46:31] |
<sgorilla1> |
cmlenz: what gets fired each time you load a page, and checks to the path string on the url? |
| [02:47:11] |
<cmlenz> |
? |
| [02:48:03] |
<cmlenz> |
if I understand you correctly, you mean the request dispatcher, but that probably isn't what you want... what's wrong with AuthenticationMiddleware? |
| [02:48:22] |
<sgorilla1> |
does that get fired each time you load a page? |
| [02:49:42] |
<sgorilla1> |
it seems it does |
| [02:54:02] |
<sgorilla1> |
ok i got it to catch if somone is not logged in |
| [02:54:26] |
* |
hpnadig has joined #trac |
| [02:55:55] |
<sgorilla1> |
i guess i cause a redirect to login |
| [02:55:59] |
* |
idnar has quit IRC |
| [02:56:02] |
* |
idnar_ has joined #trac |
| [02:57:16] |
<sgorilla1> |
is it possible to redirect inside of AuthenticationMiddleware? |
| [02:58:48] |
<sgorilla1> |
whats the difference between HTTPAuthentication and BasicAuthentication? |
| [03:01:29] |
<cboos> |
hm, it seems I miss my dear ".flatten" method from Ruby... |
| [03:02:02] |
<cboos> |
what's the common idiom in Python to flatten a list of lists? |
| [03:02:17] |
<cboos> |
e.g. going from [[1,2],[3,4,5]] to [1,2,3,4,5]? |
| [03:02:52] |
<The_Tick> |
http://cheeseshop.python.org/pypi/Flatten/0.2 ? |
| [03:03:37] |
<sgorilla1> |
choos: you can implement that with metaprogramming |
| [03:03:43] |
* |
The_Tick runs off |
| [03:03:47] |
<cmlenz> |
http://www.voidspace.org.uk/python/weblog/arch_d7_2006_08_12.shtml#e437 (+ comments) |
| [03:03:48] |
<cboos> |
hm yes ;) |
| [03:04:06] |
<sgorilla1> |
choos: show me the implementation in ruby |
| [03:04:31] |
<cboos> |
he, in Ruby it's a builtin, you don't care ... |
| [03:04:45] |
<sgorilla1> |
sorry not implementation, usage |
| [03:05:00] |
<cboos> |
[[1,2],[3,4,5]].flatten == [1,2,3,4,5] |
| [03:05:25] |
<cboos> |
cmlenz: hm, I tried something like that, weird, let me try again... |
| [03:06:35] |
<cboos> |
oh! got the order of the embedded 'for' wrong... |
| [03:07:12] |
<sgorilla1> |
choos: http://aspn.activestate.com/ASPN/Mail/Message/python-tutor/2302231 |
| [03:08:10] |
<cboos> |
sgorilla1: thanks, but I'll go for the iterator version, I think |
| [03:09:09] |
<sgorilla1> |
choos: http://rightfootin.blogspot.com/2006/09/more-on-python-flatten.html |
| [03:09:50] |
<cboos> |
ok ok, I'm overwhelmed now ;) |
| [03:10:25] |
<sgorilla1> |
i think there is a bug in standalone trac where it does authentication |
| [03:10:32] |
<sgorilla1> |
it doesn't log you out correctly |
| [03:10:47] |
* |
__ has joined #trac |
| [03:11:34] |
<sgorilla1> |
but maybe its like that because im not using the current svn |
| [03:14:47] |
* |
Synapse_ has joined #trac |
| [03:15:49] |
* |
Synapse has quit IRC |
| [03:16:05] |
* |
Synapse_ is now known as Synapse |
| [03:25:16] |
<s0undt3ch> |
cmlenz: arround? |
| [03:25:44] |
<s0undt3ch> |
I'm having a slight trouble, mind argueing |
| [03:26:06] |
<cboos> |
cmlenz: et voila, the report migration is done, enjoy r3816 ;) |
| [03:26:15] |
* |
Rica has quit IRC |
| [03:26:45] |
<cmlenz> |
cboos: cool |
| [03:26:47] |
<cmlenz> |
s0undt3ch: whatsup? |
| [03:27:03] |
<s0undt3ch> |
i'll explain, I have trac behind pound, an http proxy |
| [03:27:37] |
<s0undt3ch> |
pound rewrites REMOTE_ADDRESS ot it's own, and adds X-Forwarded-For |
| [03:28:08] |
<s0undt3ch> |
I tought about adding support for X-Forwarded-For, ie, if that's found first, use that instead of REMOTE_ADDRESS |
| [03:28:47] |
<s0undt3ch> |
but, that might introduce some security probs, because a bad client could add X-Forwarded-For |
| [03:29:24] |
<s0undt3ch> |
another thing I found out is that trac.web.api, does not use config, for example to switch that behaviour on/off |
| [03:29:30] |
<s0undt3ch> |
cmlenz: what would you sugest? |
| [03:30:14] |
<cmlenz> |
I don't really understand |
| [03:30:44] |
* |
Thanatermesis_ has joined #trac |
| [03:30:45] |
<s0undt3ch> |
cmlenz: my trac envs are behing an http proxy, so far you undertoof right? :) |
| [03:30:56] |
<cmlenz> |
yeah |
| [03:46:13] |
* |
Thanatermesis has quit IRC |
| [03:46:32] |
* |
prologic has joined #trac |
| [03:50:30] |
<cmlenz> |
s0undt3ch: I understand the REMOTE_ADDR problem, but don't understand what you mean with the config switch |
| [03:58:27] |
<s0undt3ch> |
cmlenz: sorry I had to work a bit :) |
| [04:01:32] |
<s0undt3ch> |
cmlenz: the idea is, only use the X-Forwarded-For if the user has enabled that on config, and a user will only enable that if he runs trac behind a proxy service like me |
| [04:02:22] |
<cmlenz> |
ah |
| [04:03:36] |
<s0undt3ch> |
but my fear is how much impact will that cause |
| [04:04:02] |
<s0undt3ch> |
at least a performance impact, since that config check will be made per request right? |
| [04:06:40] |
<cmlenz> |
trac.web.api shouldn't have any direct ties to config |
| [04:07:04] |
<cmlenz> |
what you'd need would be an option for the Request constructor, and pass that option from trac.web.main, depending on some config option |
| [04:07:13] |
<cmlenz> |
the config is only reloaded if it's changed |
| [04:12:51] |
* |
jesterKing is now known as amino |
| [04:13:07] |
<s0undt3ch> |
alect: so, we re-set req.remote_addr on the Request constructor, if that hipotetical option is set? |
| [04:14:03] |
<alect> |
what what |
| [04:14:35] |
<alect> |
misdirected? |
| [04:15:49] |
* |
amino is now known as jesterKing |
| [04:15:54] |
<s0undt3ch> |
alect: huh? |
| [04:16:43] |
<alect> |
you asked me a question, which i assume was meant for cmlenz |
| [04:17:59] |
<s0undt3ch> |
ah, yes misdirected :) |
| [04:18:12] |
<s0undt3ch> |
cmlenz: so, we re-set req.remote_addr on the Request constructor, if that hipotetical option is set? |
| [04:19:07] |
<cboos> |
cmlenz: about your tracd refactoring [3811:3813], should I backport that to 0.10.1 ? |
| [04:20:18] |
<cmlenz> |
nah, it's neither bug fixes nor enhancements... just plain boring refactoring |
| [04:20:27] |
<cmlenz> |
bbiab |
| [04:21:10] |
* |
peppelorum has quit IRC |
| [04:25:21] |
* |
jesterKing is now known as amino |
| [04:30:50] |
<cboos> |
dammit, I don't have the perms on 0.10-stable, cmlenz can you fix that? |
| [04:35:20] |
* |
hlb has quit IRC |
| [04:35:45] |
* |
hlb has joined #trac |
| [04:53:07] |
* |
Epcylon_ has joined #trac |
| [04:54:30] |
* |
flox has quit IRC |
| [04:54:30] |
* |
Epcylon has quit IRC |
| [04:54:32] |
* |
Epcylon_ is now known as Epcylon |
| [05:07:56] |
<andres> |
cboos, how do you thought to implement the solution with a custom collation sequence? Afaik they have to be implemented in c? |
| [05:08:19] |
<cboos> |
no, recent versions of pysqlite allow you to do that in python |
| [05:08:30] |
<cboos> |
but I haven't checked the details yet |
| [05:09:58] |
* |
andres_ has joined #trac |
| [05:10:24] |
<cboos> |
http://docs.python.org/lib/sqlite3-Connection-Objects.html#l2h-2557 |
| [05:10:53] |
<cboos> |
and http://www.sqlite.org/lang_createindex.html |
| [05:11:01] |
<cboos> |
(look for COLLATE) |
| [05:11:13] |
<cboos> |
putting the two together should do the trick |
| [05:13:19] |
* |
andres has quit IRC |
| [05:13:23] |
* |
andres_ is now known as andres |
| [05:14:30] |
<andres> |
cboos, cool. |
| [05:14:48] |
<andres> |
cboos, now the question remains about how efficent generating this index will be... |
| [05:16:11] |
<cboos> |
+ the minor question about how to implement this index in a true svn backend /and/ database backend independent way ;) |
| [05:17:05] |
<cboos> |
so hopefully the performance will be so terrible than we won't have to code it at all ;) |
| [05:17:23] |
<cboos> |
but I doubt, should be a big win |
| [05:18:53] |
<andres> |
cboos, especially as the index generation shouldnt happen that often. |
| [05:19:42] |
<cboos> |
yes, the index creation should be done after a resync, and the index update will happen during incremental resync |
| [05:20:18] |
<cboos> |
(the first "resync" being the one done with trac-admin, or when creating the env) |
| [05:24:13] |
<cboos> |
Actually, there's also a totally different approach that could be taken. We could explicitely track the latest revision, in a separate table. Other version system have the notion of intra-repository branches, each having a "latest" revision (heads in Mercurial terminology). In the SVN case, that would be only the youngest rev, but then, we wouldn't need to scan all the revisions (with or... |
| [05:24:14] |
<cboos> |
...without an index) to find it. |
| [05:27:17] |
* |
boorad has joined #trac |
| [05:39:21] |
<prologic> |
http://prologic.shortcircuit.net.au/ <-- any ideas why the Image macro does this ? |
| [05:39:40] |
<prologic> |
I'm gonna investigate and try to remove this annoying defect (ihmo) but some pointers by nice :) |
| [05:53:24] |
* |
andres has quit IRC |
| [05:54:55] |
* |
flox has joined #trac |
| [05:59:55] |
* |
milk-it has joined #trac |
| [06:02:07] |
* |
andres has joined #trac |
| [06:13:00] |
<prologic> |
Q: Is there any way to set the default edit height for individualy users ? |
| [06:22:21] |
* |
neddy has quit IRC |
| [06:47:14] |
* |
hlb has quit IRC |
| [06:58:32] |
<cboos> |
cmlenz: ping? |
| [06:58:49] |
<cmlenz> |
pong |
| [06:58:53] |
<cboos> |
good ;) |
| [06:59:21] |
<cboos> |
I tried to merge a few things on 0.10-stable, but I have no permissions there... |
| [06:59:25] |
<cboos> |
could you fix that please? |
| [07:00:29] |
<cmlenz> |
done |
| [07:00:33] |
<cboos> |
thanks! |
| [07:00:36] |
<cmlenz> |
(noone had access :-P ) |
| [07:00:43] |
<cboos> |
heh |
| [07:00:50] |
<cboos> |
on a different topic, the move of the wiki_to_html to the templates seems to go rather well |
| [07:01:02] |
<cboos> |
lots of things to change, though, so it's taking a while... |
| [07:01:44] |
<cboos> |
it simplifies a lot of stuff, e.g. no longer the need for (format=='rss') hacks |
| [07:01:56] |
<cmlenz> |
what would you think about merging the genshi branch to trunk rather sooner than later? |
| [07:02:18] |
<cboos> |
actually, I wanted to propose a review of the genshi branch just after that last change ;) |
| [07:02:28] |
<cboos> |
i.e. probably this evening |
| [07:02:46] |
<cmlenz> |
okay, sounds good |
| [07:03:11] |
<cboos> |
I haven't tested with the latest Genshi, though |
| [07:03:31] |
<cmlenz> |
I'd merge before moving wiki_to_html into the templates |
| [07:03:56] |
<cboos> |
oh, that's really not that far away... give me one hour more ;) |
| [07:04:56] |
<cmlenz> |
I just think it'll break stuff |
| [07:05:10] |
<cboos> |
hm, plugins? |
| [07:05:13] |
<cmlenz> |
which I still haven't remembered |
| [07:05:17] |
<cmlenz> |
no, macros |
| [07:05:37] |
<cboos> |
old macros using hdf, for sure, but for the others, seems to be fine |
| [07:07:03] |
<cmlenz> |
okay, I trust we with that ;-) |
| [07:07:09] |
<cmlenz> |
you, even |
| [07:08:15] |
* |
TinoW has joined #trac |
| [07:08:23] |
<TinoW> |
hi |
| [07:10:15] |
* |
TinoW is biten by SubversionException: ("Can't set position pointer in file ' ... project/db/revs/416': Invalid argument", 22) - problem. And only if I go mod_python. It works w/ tracd |
| [07:10:36] |
* |
gakman has quit IRC |
| [07:10:40] |
<cmlenz> |
cboos: now I remember... try the diff processor |
| [07:11:02] |
<cmlenz> |
it's too late to add the stylesheet it needs when wiki_to_html is called in the template |
| [07:11:40] |
<cboos> |
Ok, reproduced... |
| [07:12:36] |
<cboos> |
well, that looks like to be something that should be addressed rather than avoided |
| [07:12:37] |
* |
gakman has joined #trac |
| [07:12:53] |
<cmlenz> |
cboos: any ideas |
| [07:12:53] |
<cmlenz> |
? |
| [07:13:38] |
<cboos> |
well, I certainly will, give me some time to think about it ... |
| [07:16:02] |
<cboos> |
Hm, this looks like a serious issue. I'll ask Ilias. |
| [07:17:03] |
* |
Pyretic has joined #trac |
| [07:17:24] |
<Pyretic> |
hi, just updated to 0.10 and my source browser displays "Access denied" in the source |
| [07:18:07] |
<omry> |
you guys, FireStats 0.9.6 is out, and it supports trac. http://firefang.net/projects/firestats/ |
| [07:19:33] |
<TinoW> |
very strange - all projects work w/ tracd - some work with mod_python, some dont and the ones they work have output on svnadmin list-dblogs |
| [07:20:08] |
<coderanger> |
TinoW: Try a dump/load maybe? |
| [07:21:15] |
* |
danbeck has joined #trac |
| [07:21:47] |
<TinoW> |
coderanger: well, they work as repositories and they work w/ tracd. trashing the repositories isnt really an option since I'm not the only one using it ;) |
| [07:22:14] |
* |
s0undt3ch_ has joined #trac |
| [07:22:29] |
<coderanger> |
You don't trash them, just dump to a file, move them over, make a clean repo, and reload from the dump |
| [07:22:37] |
* |
agile has quit IRC |
| [07:23:32] |
<TinoW> |
coderanger: sure, but then I have an unrelated copy. And I really dont see the difference between tracd and mod_python |
| [07:23:49] |
<coderanger> |
TinoW: You delete the old one afterwards .... |
| [07:24:30] |
<TinoW> |
coderanger: and I'd have to block people from writing to the old during the process ;) I'd rather not do that ;) |
| [07:24:41] |
<TinoW> |
coderanger: beside that I cant see how it would help? |
| [07:24:49] |
<coderanger> |
It takes less than a minute |
| [07:25:08] |
<coderanger> |
And if you are seeing db issues, it will fix them |
| [07:25:25] |
<coderanger> |
Since it has to reform the DB as it does the load |
| [07:25:28] |
<TinoW> |
coderanger: but why dont I see these db issues when I use tracd? |
| [07:25:49] |
<coderanger> |
Not a clue, but its worth a shot |
| [07:27:17] |
* |
_dmax has joined #trac |
| [07:27:37] |
<coderanger> |
Possibly its related to some kind of file locking, but I doubt it |
| [07:27:51] |
<coderanger> |
You probably want to ask the #subversion people |
| [07:28:32] |
* |
ryepup has quit IRC |
| [07:30:19] |
* |
s0undt3ch has quit IRC |
| [07:32:31] |
<TinoW> |
coderange: ok, the copy does not work - same error |
| [07:32:34] |
<TinoW> |
:( |
| [07:32:40] |
<TinoW> |
code: thx for help anyway |
| [07:34:24] |
<Pyretic> |
hmmm |
| [07:34:40] |
* |
s0undt3ch_ is now known as s0undt3ch |
| [07:34:57] |
<coderanger> |
TinoW: DB corruption then |
| [07:37:27] |
<TinoW> |
code? |
| [07:37:35] |
<TinoW> |
coderanger: which db? |
| [07:37:40] |
<coderanger> |
The subversion db |
| [07:37:57] |
<TinoW> |
coderanger: but only corrupt for mod_python? |
| [07:38:10] |
<TinoW> |
coderanger: and also corrupt after reload? |
| [07:38:35] |
<coderanger> |
wait, the action of copying doesnt work, or the end result doen work? |
| [07:38:51] |
<Pyretic> |
hi, just updated to 0.10 and my source browser displays "Access denied" in the source |
| [07:39:15] |
<TinoW> |
coderanger: the end result behaves the same as the original in respect of work/dontwork w tracd/mod_python |
| [07:39:37] |
* |
dmax has quit IRC |
| [07:39:46] |
* |
_dmax is now known as dmax |
| [07:41:19] |
* |
em-dash has joined #trac |
| [08:02:45] |
<TinoW> |
*ping* |
| [08:02:46] |
<TinoW> |
;) |
| [08:03:14] |
<prologic> |
pong |
| [08:05:39] |
* |
klasstek has joined #trac |
| [08:08:25] |
<prologic> |
s0undt3ch, I have another bug for you |
| [08:09:54] |
<misc> |
would it make sense to auto-detect installed component in trac/db_default.py, at least for db backend and vcs backend, instead of using a static list ? |
| [08:10:54] |
<prologic> |
s0undt3ch, when you create a new page for the first-time I keep getting: AttributeError: 'WikiNotifyEmail' object has no attribute 'wikidiff' |
| [08:11:01] |
<coderanger> |
misc: hmm? |
| [08:11:27] |
* |
chandlerc has joined #trac |
| [08:11:58] |
<misc> |
coderanger: i am doing a package of trac, i have splitted mysql, sqlite and postgresql backend in subpackage, to make them requires the proper binding |
| [08:12:24] |
<misc> |
coderanger: but, trac try to load all backend by default |
| [08:12:40] |
<coderanger> |
misc: If you removing parts of Trac, I think thats a risk you take ;-) |
| [08:13:16] |
<coderanger> |
Trac only autoloads components that are actually part of Trac |
| [08:13:20] |
<s0undt3ch> |
prologic: heh, great, another one :) |
| [08:13:41] |
<misc> |
so either i place the file back in the main packages, and let empty subpackages handle the requires on binding, or manage to load them by default |
| [08:14:09] |
<misc> |
coderanger: so you suggest to not place them in subpackages ? |
| [08:14:37] |
<coderanger> |
misc: Trac already handles autodetecting what DB bindings are available |
| [08:14:49] |
<misc> |
coderanger: mhh, ok |
| [08:15:06] |
<misc> |
i do not like empty subpackages, but well |
| [08:15:36] |
<coderanger> |
misc: You may want to do something like Gentoo's USE flags |
| [08:15:45] |
<coderanger> |
Or BSD's flavors |
| [08:16:45] |
<misc> |
in the principe, yes, that is what i want |
| [08:17:00] |
<prologic> |
s0undt3ch, sorry :) |
| [08:18:20] |
<TinoW> |
hm. does tracd use another code path to access svn then mod_python? |
| [08:18:50] |
<coderanger> |
Nope |
| [08:19:04] |
<TinoW> |
so I wonder where the difference comes |
| [08:19:14] |
<coderanger> |
The only difference between the frontends is how the pass the request to the dispatcher and how the show output |
| [08:19:15] |
<s0undt3ch> |
prologic: np, I suspected of it yesterday, but I forgot to take care of it today |
| [08:19:16] |
<TinoW> |
very very strange |
| [08:19:39] |
<coderanger> |
TinoW: Maybe some libarary that compiled into apache is interfering? |
| [08:19:50] |
* |
flox has quit IRC |
| [08:20:12] |
<prologic> |
s0undt3ch, nps, I'll update when 1.1 is out :) |
| [08:20:12] |
<TinoW> |
coderange: but only interferring with some repositories? |
| [08:20:34] |
* |
gmcinnes has joined #trac |
| [08:21:13] |
<s0undt3ch> |
prologic: nope rc4 ;) |
| [08:21:33] |
<LarstiQ> |
prologic: ah, but you do not need the working tree for trac bzr support |
| [08:22:14] |
<prologic> |
you don't ? |
| [08:22:14] |
* |
idnar_ is now known as idnar |
| [08:22:18] |
<prologic> |
please explain :) |
| [08:22:26] |
<prologic> |
s0undt3ch, ahh nps rc4 then :) |
| [08:33:16] |
<TinoW> |
prologic: uh. nps, the cms? |
| [08:37:45] |
<prologic> |
eh ? |
| [08:41:21] |
<LarstiQ> |
prologic: only the .bzr is ever needed |
| [08:41:48] |
<LarstiQ> |
prologic: the working tree is for when you want to edit it (or, deployment) |
| [08:42:20] |
<LarstiQ> |
prologic: but for publishing branches, working with trac, there is no need for a working tree |
| [08:43:12] |
<cboos> |
cmlenz: what about keeping the genshi branch for experimenting with the wiki_to_html refactoring? |
| [08:45:36] |
<cmlenz> |
dunno, does that really require a whole branch? |
| [08:46:03] |
<prologic> |
LarstiQ, ahh thanks for that info |
| [08:47:38] |
<LarstiQ> |
prologic: np, it did take some tracking you down to provide that info though ;P |
| [08:47:41] |
<s0undt3ch> |
how do I get every header a req has? I need to log that |
| [08:48:41] |
* |
gmcinnes has quit IRC |
| [08:48:43] |
<cboos> |
well, fixing the add_stylesheet and co might require some experiments... But if we don't mind doing that on trunk, well, I'm ok |
| [08:49:08] |
<cboos> |
cmlenz: want to have a look at the patch? |
| [08:49:20] |
<cmlenz> |
sure |
| [08:49:46] |
<s0undt3ch> |
cmlenz: any idea how I can achieve that? |
| [08:50:12] |
<cmlenz> |
s0undt3ch: you'd need to look into req.environ |
| [08:50:22] |
<s0undt3ch> |
k, thanks |
| [08:50:25] |
<cmlenz> |
which is a WSGI environ, with values very similar to that of a CGI script |
| [08:50:40] |
<cboos> |
patch is quite big, "pasting" that on t.e.o: http://trac.edgewall.org/attachment/wiki/ChristianBoos/wiki2x_in_templates.diff |
| [08:50:59] |
<prologic> |
LarstiQ, haha I appreciate it :) I wasn't really worried either way |
| [08:51:01] |
<prologic> |
I mean it works :) |
| [08:52:21] |
<cmlenz> |
cboos: nice... no solution yet for the stylesheet problem, right? |
| [08:52:35] |
<cboos> |
yeah, still brainstorming... |
| [08:52:46] |
<cboos> |
(and Ilias hasn't answered ;) ) |
| [08:53:01] |
<cmlenz> |
I think the only thing that may work would be to put an <style> block directly above the macro output... but that's not valid HTML |
| [08:53:27] |
<cboos> |
but when you said you wanted to refactor the context nav stuff, |
| [08:53:48] |
<cboos> |
wasn't that meant to allow for more flexible manipulation of the context nav data, |
| [08:53:52] |
<cboos> |
e.g. by macros? |
| [08:54:03] |
<cboos> |
Then the add_stylesheet might be solved the same way... |
| [08:54:15] |
<cmlenz> |
not really by macros... by other kinds of plugins (INavigationContributor, I guess) |
| [08:54:36] |
<cboos> |
I mean, we have to find a way to do some postprocessing anyway, |
| [08:54:48] |
<LarstiQ> |
prologic: you're using trac-bzr? |
| [08:54:58] |
<cmlenz> |
I don't see a way to add something to the <HEAD> of a doc by some function that's called later in the template rendering process without doing multiple passes over the output somehow |
| [08:54:59] |
<prologic> |
*nods* |
| [08:55:06] |
<prologic> |
LarstiQ, *nods* |
| [08:55:18] |
<prologic> |
LarstiQ, on http://prologic.shortcircuit.net.au/ |
| [08:55:37] |
<cmlenz> |
cboos: indeed, a stream filter might wo |
