| [00:04:13] |
* |
tuxipuxi has joined #trac |
| [00:22:16] |
* |
tuxipuxi has quit IRC |
| [00:33:49] |
* |
cmlenz has quit IRC |
| [00:51:26] |
* |
cmlenz has joined #trac |
| [01:21:38] |
<alect> |
sqlalchemy seems cool' |
| [01:28:39] |
<cmlenz> |
yeah, very |
| [01:36:34] |
<alect> |
it "just worked" on the trac db, very impressive |
| [01:37:07] |
* |
klando has joined #trac |
| [01:43:39] |
* |
ValiSystM has joined #trac |
| [02:07:41] |
* |
idnar has quit IRC |
| [02:07:45] |
* |
idnar_ has joined #trac |
| [02:32:25] |
* |
tuxipuxi has joined #trac |
| [02:36:08] |
* |
imajes has joined #trac |
| [02:48:23] |
* |
jacobat has quit IRC |
| [02:49:12] |
* |
jacobat has joined #trac |
| [03:04:49] |
* |
jacobat has quit IRC |
| [03:09:54] |
* |
jacobat has joined #trac |
| [03:42:56] |
* |
_idostyle has joined #trac |
| [03:47:41] |
* |
Riccc has joined #trac |
| [03:58:51] |
* |
idnar_ is now known as idnar |
| [03:59:18] |
* |
idostyle has quit IRC |
| [03:59:26] |
* |
_idostyle is now known as idostyle |
| [04:05:44] |
* |
tuxipuxi has quit IRC |
| [04:05:48] |
* |
Ricc has quit IRC |
| [04:10:02] |
* |
matt_good has quit IRC |
| [04:10:04] |
* |
matt_good_ has joined #trac |
| [04:16:03] |
* |
whitelynx has quit IRC |
| [04:26:57] |
* |
honda_S2000 has joined #trac |
| [04:27:09] |
<honda_S2000> |
ji |
| [04:27:37] |
<honda_S2000> |
is it possible to populate a ticket custom fiel with the values from a report or query to trac.db? |
| [04:31:29] |
* |
prisoner has quit IRC |
| [04:31:45] |
* |
prisoner has joined #trac |
| [04:31:58] |
<honda_S2000> |
can anyone answer to that please?? |
| [04:34:38] |
<alect> |
if you're using the workflow sandbox, yes, otherwise no |
| [04:36:43] |
<honda_S2000> |
sorry i'm a newbie |
| [04:36:47] |
<honda_S2000> |
explain that to me |
| [04:37:14] |
<honda_S2000> |
how do i know if i am using that and if i don't use it yet i can i use it |
| [04:37:17] |
<alect> |
effectively, no is the answer |
| [04:37:52] |
<alect> |
there is a branch called workflow (search for WorkFlow on PEC) which would let you do that via a plugin |
| [04:37:56] |
<alect> |
but you'd have to write a plugin to do it |
| [04:38:15] |
<honda_S2000> |
ohh i was hopefull on that to be easu |
| [04:38:17] |
<honda_S2000> |
*easy |
| [04:52:20] |
* |
otaku42_away is now known as otaku42 |
| [04:52:40] |
* |
oferw has joined #trac |
| [05:10:30] |
* |
oferw has quit IRC |
| [05:15:50] |
* |
ValiSystM has quit IRC |
| [05:20:11] |
* |
matt_good_ has quit IRC |
| [05:20:40] |
* |
matt_good_ has joined #trac |
| [05:48:25] |
* |
bwmcadams has joined #trac |
| [05:53:42] |
* |
ValiSystM has joined #trac |
| [05:55:40] |
<cboos> |
hi all, |
| [05:55:53] |
<cboos> |
I noticed that external links looks a bit better with this change: |
| [05:56:01] |
<cboos> |
a.ext-link .icon { |
| [05:56:03] |
<cboos> |
- background: url(../extlink.gif) left center no-repeat; |
| [05:56:04] |
<cboos> |
+ background: url(../extlink.gif) center no-repeat; |
| [05:56:37] |
* |
cprov has joined #trac |
| [05:58:46] |
<cboos> |
especially when there's a wrap due to a long link. With "left center" instead of "center", the icon touches the last word on the line before the link, and that's quite ugly. |
| [05:59:24] |
<cboos> |
Of course, a better fix would be a way to force the icon and the link to always stay on the same line... |
| [05:59:49] |
<cboos> |
Any CSS expert listening my rant :) |
| [05:59:50] |
<cboos> |
? |
| [06:00:01] |
<bwmcadams> |
its 9am. go back to bed. |
| [06:06:28] |
* |
sam`_ has joined #trac |
| [06:09:55] |
* |
Hydra_ has quit IRC |
| [06:10:07] |
<cboos> |
eh, it's funny how asking a question often leads to discovering the answer... |
| [06:10:24] |
<cboos> |
sorry for the noise ;) |
| [06:12:37] |
* |
sam` has quit IRC |
| [06:18:53] |
* |
tuxipuxi has joined #trac |
| [06:29:11] |
* |
idostyle has quit IRC |
| [06:35:37] |
<boorad> |
cmlenz: you around? |
| [06:35:59] |
<cmlenz> |
yeah |
| [06:36:08] |
<BaconTastesGood> |
how do I disable anonymous browsing of the top level project list? |
| [06:36:19] |
<boorad> |
cmlenz: I'll try to be brief: any reason I'd be seeing this: http://trac.pastebin.com/707368 |
| [06:36:42] |
<boorad> |
cmlenz: seems the plugins are loading twice, something is sticking in memory? |
| [06:36:51] |
* |
honda_S2000 has quit IRC |
| [06:37:21] |
<cmlenz> |
or a dupe log appender |
| [06:37:47] |
<cmlenz> |
hmm, yeah... I've seen that before |
| [06:38:03] |
<cmlenz> |
the logging would keep getting a new handler for every request |
| [06:38:09] |
<boorad> |
ok, I'm still having issues with PG, so anything that looked out of whack, or might hold weak references to objects open, I wanna chase it |
| [06:38:27] |
<boorad> |
sounds unrelated, though |
| [06:38:36] |
<cmlenz> |
not sure |
| [06:39:03] |
<cmlenz> |
might be related |
| [06:39:39] |
<boorad> |
cmlenz: I posted this a few days ago - it's pretty sad ... http://trac.pastebin.com/697270 |
| [06:40:02] |
<cmlenz> |
heh, yeah |
| [06:40:17] |
<cmlenz> |
you should try the sqlalchemy branch some time |
| [06:40:27] |
<boorad> |
cronjob every hour didn't work, had to go to minute b/c of a bot scanning the site |
| [06:40:46] |
<boorad> |
sqlalchemy - I saw you guys talking about that earlier - what's the status of the branch? |
| [06:41:21] |
* |
Hydra_ has joined #trac |
| [06:41:22] |
<cmlenz> |
not sure... apparently it worked well for matt... didn't try it myself |
| [06:41:38] |
<cmlenz> |
I think you may be able to take the diff of the branch and apply that to trunk with only a few mods |
| [06:41:53] |
<cmlenz> |
I imagine it has much more solid connection management |
| [06:41:58] |
<boorad> |
I'll give it a whirl |
| [06:42:06] |
<boorad> |
any thoughts on the appender thing? |
| [06:42:20] |
<cmlenz> |
trying to remember what caused that the last time around :-P |
| [06:42:25] |
* |
matt_good_ has quit IRC |
| [06:42:40] |
<boorad> |
I'm current on /trunk as of a couple of days ago |
| [06:43:14] |
<boorad> |
cmlenz: btw, did something stop allowing me to set req.hdf values in match_request() ?? |
| [06:43:43] |
<cmlenz> |
not sure, I don't think so |
| [06:44:00] |
<boorad> |
hmm, not working anymore - still does when I set them in process_request() |
| [06:44:16] |
<cmlenz> |
ah wait, cboos did something in that area a couple weeks ago |
| [06:44:33] |
<cboos> |
what did I do? |
| [06:44:49] |
<boorad> |
the ones where I set them in match_request() use the parent class process_request(), so match_request() is the only place I can do it |
| [06:45:02] |
<cmlenz> |
cboos: creating the hdf lazily, no? |
| [06:45:12] |
<cmlenz> |
i.e. after match_request() |
| [06:45:12] |
<cboos> |
uh? |
| [06:45:16] |
<cboos> |
ah! |
| [06:45:30] |
<cmlenz> |
see, I remember your stuff better than you do ;-) |
| [06:45:37] |
<boorad> |
lol |
| [06:45:52] |
<boorad> |
where, in web.main? |
| [06:46:06] |
<cboos> |
yeah, actually the change I did was preventing building the hdf stuff for the Chrome requests... |
| [06:46:18] |
<cboos> |
not sure it's related to that issue |
| [06:46:38] |
<cmlenz> |
pretty sure it is... you need to match first to determine whether the handler needs the HDF etc |
| [06:47:00] |
<cmlenz> |
but of course populating req.hdf from match_request() should be considered a hack I think we shouldn't support ;-) |
| [06:47:06] |
<boorad> |
it's nice to do it in my plugin's match_request(), ya know? |
| [06:47:24] |
<boorad> |
how else would I do it? |
| [06:47:37] |
<cmlenz> |
I don't know your use case |
| [06:47:46] |
<boorad> |
it felt weird, but no other place (in my plugin) seemed feasible |
| [06:48:08] |
<cboos> |
ok, well, now I understand what boorad tried to do... no, it won't work anymore due to my changes, sorry ;) |
| [06:48:24] |
<boorad> |
I'm trying to set a variable as to which 'site-wide' tab is selected - see the top right at http://www.dsource.org |
| [06:48:37] |
<boorad> |
it works on that version in production, but not on my local |
| [06:49:33] |
<boorad> |
I set req.hdf["dsource.site"] = 1 and then the cs template makes that the selected tab |
| [06:49:42] |
<cmlenz> |
hmm, if clearsilver wasn't so limited you could do that in the site_header.cs, I guess |
| [06:49:46] |
<boorad> |
thoughts on how to get around this, do it the right way? |
| [06:50:19] |
<boorad> |
I agree - in fact, due to the limitations, I am hacking header.cs, instead of keeping my stuff out of Trac proper :( |
| [06:50:22] |
<cmlenz> |
basically you're abusing match_request to implement a request filter/middleware |
| [06:50:38] |
* |
boorad raises hand guiltily |
| [06:50:59] |
<cmlenz> |
we have a ticket/proposal for request filters somewhere |
| [06:51:07] |
<cmlenz> |
by alect IIRC |
| [06:51:14] |
* |
sam`_ is now known as sam` |
| [06:51:31] |
<BaconTastesGood> |
how do I disable anonymous browsing of the top level project list? |
| [06:51:39] |
<boorad> |
just trying to skate by until some of these changes are implemented |
| [06:52:03] |
<boorad> |
cmlenz, cboos: any thoughts on how to do this in the short-term? |
| [06:52:36] |
* |
cboos have a look at the code, waiting for the solution to jump at his face... |
| [06:52:53] |
<boorad> |
the places where I'm (ab)using the match_process() is in a plugin that inherits from WikiModule |
| [06:53:19] |
<boorad> |
so I could maybe have my plugin override process_request() and then hand off to a newly created WikiModule object ? |
| [06:54:01] |
<boorad> |
I tried that (as far as I knew how), but it didn't work |
| [06:54:17] |
<cmlenz> |
you'd have to disable the real WikiModule |
| [06:54:48] |
* |
danbeck has joined #trac |
| [06:54:51] |
<boorad> |
can't... Do you see that Home and Site are actually wiki pages on dsource.org? Those two are the plugins that we're talking about |
| [06:55:37] |
<cmlenz> |
well, but you'd have your inherited WikiModule to replace it... anyway, not going to work because the tabs would only be right on actual wiki pages |
| [06:56:25] |
<alect> |
we discussed the request filters in relation to the kidalike i think? |
| [06:56:43] |
<boorad> |
progress on the new templating engine? |
| [06:56:52] |
<alect> |
or no, that was the actual tree filtering |
| [06:57:15] |
<cmlenz> |
yeah, request filtering is lower-level... but still higher than WSGI-level filters |
| [06:57:23] |
<cboos> |
boorad: you could hack Chrome.populate_hdf, and call 'handler.populate_hdf' if 'populate_hdf' is a defined method on your handler... |
| [06:57:32] |
<alect> |
monkey patch, mmm |
| [06:57:37] |
<alect> |
makes you feel warm and fuzzy |
| [06:57:48] |
<cmlenz> |
but for boorad's use case, a django-ish "template context processor" would be the best option |
| [06:58:05] |
<boorad> |
sounds painful ;) |
| [06:58:35] |
<alect> |
cboos, i tried the >> patch |
| [06:58:54] |
<cboos> |
yes, and what do you think about it? |
| [06:59:15] |
<boorad> |
cboos: I'll give it a try - what does this do for handlers that don't have that method? just trap it in Chrome? |
| [06:59:33] |
<alect> |
http://trac.pastebin.com/707399 |
| [06:59:39] |
<alect> |
^^ that did not render as i would expect |
| [06:59:48] |
<cboos> |
boorad: call hasattr(handler, 'populate_hdf') before... |
| [06:59:55] |
<boorad> |
cboos: thx |
| [07:00:00] |
<alect> |
is pec toast? |
| [07:00:39] |
<cboos> |
alect: I'll try... can you paste me also what is the output you'd expect? |
| [07:01:14] |
<alect> |
you'll see when you try it. the tables are at the same level |
| [07:02:18] |
<alect> |
i like the >> stuff, although perhaps newlines should be converted to line breaks? |
| [07:02:52] |
<cmlenz> |
what >> stuff? |
| [07:03:01] |
<alect> |
citations |
| [07:03:02] |
<cboos> |
wiki citation syntax |
| [07:03:07] |
<cboos> |
in #124 |
| [07:03:15] |
<cmlenz> |
blockquote? |
| [07:03:25] |
<alect> |
yeah |
| [07:03:32] |
<cboos> |
yes, along with list and regular indentation improvements |
| [07:03:44] |
* |
perchr has quit IRC |
| [07:04:08] |
<cboos> |
alect: looks good to me, what patch did you try? |
| [07:04:11] |
<cmlenz> |
hmm, single > would make it compatible with email quoting |
| [07:04:17] |
* |
idostyle has joined #trac |
| [07:04:28] |
<cboos> |
cmlenz: it's compatible with email quoting, actually |
| [07:04:32] |
<alect> |
yeah that's the goal, i just put too many in :) |
| [07:04:35] |
* |
cmlenz is a markdown fan |
| [07:04:40] |
<boorad> |
cboos: beautiful, and it's far less abusive / hackish. thanks! |
| [07:05:08] |
<cboos> |
your welcome ;) |
| [07:05:08] |
<alect> |
i used whatever patch was on there about two hours ago |
| [07:05:19] |
<alect> |
i can't get to pec atm so i can't check |
| [07:05:34] |
<alect> |
md5 is 793c2c5fdac28cb13a95fbd7675b490f |
| [07:05:34] |
<cboos> |
strange ... let me check if I have some local modifications... |
| [07:05:34] |
<boorad> |
cmlenz, cboos: is it worth it to add that to the IRequestHandler? or is my use case rare? or will request filters do the trick later? |
| [07:06:14] |
<cmlenz> |
I'd rather implement it properly for the templating change in 0.11 |
| [07:06:25] |
<boorad> |
k, np - thx for time everyone |
| [07:06:25] |
<cboos> |
boorad: if added, certainly another name than populate_hdf should be chosen ... it's way too general |
| [07:06:27] |
<cmlenz> |
you'd be able to just do that in your template, actually |
| [07:06:50] |
<boorad> |
what has been chosen for templating? |
| [07:07:00] |
<alect> |
cboos: this is what mine renders like: http://trac.pastebin.com/707408 |
| [07:07:02] |
<boorad> |
kidalike, I saw? |
| [07:07:30] |
<alect> |
it's top secret hush hush |
| [07:07:37] |
<alect> |
wink wink say no more say no more :) |
| [07:09:00] |
<cboos> |
alect: mine like http://trac.pastebin.com/707412 |
| [07:09:18] |
<cmlenz> |
kidalike, yeah |
| [07:09:23] |
<alect> |
oh well :) |
| [07:09:29] |
<alect> |
i'll try it later again |
| [07:09:31] |
* |
cmlenz restarted p.e.c |
| [07:09:46] |
<cmlenz> |
(well, apache on p.e.c, anyway) |
| [07:09:57] |
<cboos> |
what was it on p.e.c, that time? |
| [07:10:26] |
<cmlenz> |
dunno, load was very low, memory usage okay... just hanging |
| [07:10:34] |
<cmlenz> |
I have no idea, tbh |
| [07:10:59] |
<alect> |
chris, what do you think about IMIMETypeConverter? |
| [07:11:16] |
<cboos> |
christopher I assume :) |
| [07:11:20] |
* |
sam` has quit IRC |
| [07:11:33] |
<alect> |
hehe yeah :) |
| [07:12:35] |
<cboos> |
alect: btw, as you tried #124, could you also try #2703? |
| [07:12:52] |
<alect> |
ooh |
| [07:12:55] |
<alect> |
i like that idea |
| [07:13:00] |
<cboos> |
(ticket comments permalinks + reply to) |
| [07:14:28] |
<alect> |
how did you get around the comment re-numbering? |
| [07:14:51] |
<cmlenz> |
alect: I like the idea, but the API and naming of things not so much ... need to take a closer look at what it does before further commenting though :-P |
| [07:15:11] |
<cboos> |
I distinguish between "permanent" changes and changes that might go away (currently only the attachment changes) |
| [07:16:02] |
<cboos> |
If we implement comment/change deletion, a good compromise would be anyway to leave a placeholder saying that a comment was deleted. |
| [07:16:04] |
<alect> |
cmlenz: mkay |
| [07:16:32] |
<cboos> |
An alternative would be to use the 'old' field for the comment to leave the comment number there. |
| [07:17:03] |
<cboos> |
That's what I would like to use anyway to "remember" the replied-to comment... |
| [07:17:22] |
<cmlenz> |
cboos: actually, the same is true for deleting ticket attachments... leaving them there but with a deleted/obsoleted flag or sth |
| [07:17:39] |
<cmlenz> |
in my book, *any* change to a ticket should be permanent |
| [07:17:55] |
<alect> |
seems to work well cboos |
| [07:18:12] |
<alect> |
except the [Reply] is ugly :P |
| [07:18:35] |
<cboos> |
hm, well, that's the best I can do :( sorry ;) |
| [07:18:48] |
<cboos> |
what would you suggest? |
| [07:19:08] |
<cmlenz> |
a button? |
| [07:19:23] |
<alect> |
Ditch the [] i think |
| [07:19:35] |
<alect> |
i suspect a button would be a bit too ... intrusive |
| [07:19:46] |
<cmlenz> |
the convention in trac is to use buttons for initiating edits and stuff |
| [07:19:52] |
<alect> |
just plain "reply" looks pretty good |
| [07:20:32] |
<cboos> |
yeah, that's what I thought too. If it would be a button, it should be *above* the line, but I couldn't figure how to do that. |
| [07:20:41] |
<alect> |
looks pretty nasty with a button |
| [07:20:55] |
<alect> |
has there been any talk about styling form controls? |
| [07:21:11] |
<cmlenz> |
yeah, that we do too much of that :-) |
| [07:21:24] |
<cmlenz> |
styling form controls sucks big time |
| [07:21:25] |
<cboos> |
alect: about #T124, I figured out the patch is not complete ... will fix that |
| [07:21:32] |
<cmlenz> |
coz you can't null out the styles |
| [07:21:51] |
<alect> |
what do you mean null out? |
| [07:22:11] |
<cmlenz> |
you can't have a site deployment go back to the default (native) look of form controls |
| [07:22:21] |
<cmlenz> |
if the trac css does too much control styling |
| [07:22:28] |
<alect> |
aah |
| [07:24:09] |
<alect> |
perhaps the "reply" anchor could be styled to look more like a button...? |
| [07:24:16] |
* |
FabulousGeek has joined #trac |
| [07:24:22] |
* |
alect shrugs |
| [07:25:01] |
* |
FabulousGeek has left #trac |
| [07:26:51] |
<cmlenz> |
cboos: what's the a/ b/ stuff in those patches? |
| [07:26:52] |
* |
tolsen has quit IRC |
| [07:27:19] |
<cboos> |
patches produces by mercurial :) just use `patch -p1` |
| [07:27:31] |
<cboos> |
oops, now I wiki format my IRC messages ... |
| [07:27:35] |
<alect> |
hahaha |
| [07:27:39] |
<alect> |
heh |
| [07:28:40] |
<alect> |
i like that new timeline full wiki formatting |
| [07:28:55] |
<alect> |
trac-hacks is running completely unpatched now |
| [07:31:21] |
* |
MasterC has joined #trac |
| [07:35:45] |
<alect> |
i'm renaming some of the methods in that patch cmlenz |
| [07:36:15] |
<cmlenz> |
a better idea for the interface name? |
| [07:36:35] |
<alect> |
IContentConverter? |
| [07:36:51] |
<cmlenz> |
transformer or translator maybe? |
| [07:37:08] |
<cmlenz> |
converter isn't bad though |
| [07:37:53] |
<cmlenz> |
IContentConverter is better, but a tad too generic... I think I haven't quite grasped the scope of this critter yet |
| [07:38:09] |
<alect> |
:) |
| [07:38:12] |
* |
sam` has joined #trac |
| [07:38:14] |
<alect> |
the method names were a bit unclear i think |
| [07:38:31] |
<alect> |
eg. conversions() -> get_content_conversions() |
| [07:38:38] |
<alect> |
and convert() -> convert_content() |
| [07:39:01] |
<cmlenz> |
get_supported_conversions maybe? |
| [07:39:48] |
<alect> |
righto |
| [07:40:09] |
<cboos> |
... to add a bit to the conversation, I plan to add a similar interface, but which will dedicated to the diffs |
| [07:40:20] |
<cboos> |
e.g. IContentChangeConverter, |
| [07:41:19] |
<alect> |
what will it do? |
| [07:41:32] |
<alect> |
different diff outputs? |
| [07:41:33] |
<cboos> |
with the same (input_mime_type, output_mimetype) selection mechanism, and convert will take old, new (and eventually base) content |
| [07:41:44] |
<cboos> |
yes |
| [07:41:47] |
<cboos> |
#53 |
| [07:41:56] |
<alect> |
aha |
| [07:42:49] |
<alect> |
sneaky |
| [07:43:15] |
<cboos> |
also, it would be extra nice to have 3-way diffs for the SCM systems that supports that ;) |
| [07:43:39] |
<alect> |
like mercurial? ;) |
| [07:43:50] |
<cboos> |
... like mercurial of course ;) |
| [07:45:31] |
<alect> |
what about adding CSV ticket export to trac core? |
| [07:45:41] |
<alect> |
or XML |
| [07:45:48] |
<bwmcadams> |
that's already available |
| [07:45:50] |
<bwmcadams> |
on the reports system |
| [07:45:57] |
<alect> |
people seem to request that facility quite frequently |
| [07:46:12] |
<alect> |
yeah, but i'm talking per-ticket |
| [07:46:14] |
<bwmcadams> |
oh... |
| [07:46:15] |
<bwmcadams> |
hmm |
| [07:46:18] |
<bwmcadams> |
i'm actually working on a patch |
| [07:46:33] |
<bwmcadams> |
that changes the 'alternate report format' code in query and reports to be an Extension interface |
| [07:46:41] |
<alect> |
ah |
| [07:46:50] |
<bwmcadams> |
i could probably add something into it for per ticket too, without much difficulty |
| [07:47:05] |
<bwmcadams> |
i can see people wanting pdf and such as well |
| [07:47:06] |
<alect> |
hmm, perhaps that should use IContentConverter as well... |
| [07:47:25] |
<alect> |
have a look at #2296 |
| [07:48:00] |
<bwmcadams> |
my business users are demanding Excel export of reports... blech |
| [07:48:12] |
<bwmcadams> |
so i figured while i was doing it i may as well contribute something useful to trac by making it easy to add alternate outputs |
| [07:48:17] |
<alect> |
that's just plain wrong |
| [07:48:18] |
<alect> |
:) |
| [07:48:21] |
<bwmcadams> |
yes well |
| [07:48:28] |
<bwmcadams> |
they love trac, |
| [07:48:37] |
<bwmcadams> |
they didn't bother using bug tracking when we used bugzilla |
| [07:48:49] |
<bwmcadams> |
so it's a step in the right direction..but they like pretty excel reports for when they have their meetings |
| [07:49:11] |
<alect> |
bugzilla is unreal. drives me nuts using it |
| [07:49:58] |
<bwmcadams> |
yup |
| [07:50:00] |
<alect> |
the patch on #2296 adds the ability to add arbitrary mime conversions, and adds the facility to the ticket page |
| [07:50:10] |
<bwmcadams> |
but we DO have features we need from bugzilla |
| [07:50:12] |
<bwmcadams> |
that aren't in there |
| [07:50:15] |
<alect> |
but it doesn't include a CSV exporter by default, which i think it probably should |
| [07:50:15] |
<bwmcadams> |
that i'll probably end up coding |
| [07:50:20] |
<alect> |
such as? |
| [07:50:30] |
<bwmcadams> |
saved queries |
| [07:50:34] |
<bwmcadams> |
(not reports, the query interface) |
| [07:50:47] |
<alect> |
ah, yes that would be very useful |
| [07:51:01] |
<bwmcadams> |
multi-select boxes for picking users to assign and cc, driven from the auth table |
| [07:51:06] |
<bwmcadams> |
since we require login to access our instances |
| [07:51:23] |
<bwmcadams> |
oh, i did write a nice plugin that adds quips to trac |
| [07:51:36] |
<bwmcadams> |
I keep meaning to put it on trac-hacks...it adds a random quip to the top of the screen :) |
| [07:51:40] |
<cboos> |
bwmcadams: about saved queries, there's a ticket for that. |
| [07:51:44] |
<bwmcadams> |
[yes, my users wanted quips back] |
| [07:52:08] |
<alect> |
heh |
| [07:53:47] |
* |
klasstek has joined #trac |
| [07:54:04] |
<bwmcadams> |
i mean of all the important things i could have written, they HAD To have quips... |
| [07:54:06] |
<BaconTastesGood> |
is there a way to disallow the top level project browser? i.e. so that anonymous users can't see what all the different projects are? |
| [07:54:18] |
<bwmcadams> |
what top level project browser? |
| [07:54:21] |
<bwmcadams> |
the svn browser? |
| [07:54:33] |
<BaconTastesGood> |
no, like if you go to http://mydomain.com/trac and it lists the projects |
| [07:54:35] |
<alect> |
the project index |
| [07:54:37] |
<BaconTastesGood> |
right |
| [07:54:39] |
<BaconTastesGood> |
that thing =) |
| [07:54:44] |
<bwmcadams> |
oh.that's just an apache thing innit? |
| [07:54:46] |
<alect> |
no i don't think so |
| [07:54:50] |
<BaconTastesGood> |
I'm using tracd |
| [07:54:57] |
<bwmcadams> |
oh, i'm doing it al lwith mod_python |
| [07:55:25] |
<bwmcadams> |
bbiab - meetings... |
| [07:55:27] |
<BaconTastesGood> |
are permissions handled by choice of Web server? |
| [07:55:28] |
<alect> |
here's the new content converter api: http://trac.pastebin.com/707490 |
| [07:55:48] |
<bwmcadams> |
alect: I'll have to look about integrating it into what i'm doing then.. |
| [07:55:53] |
<cboos> |
bwmcadams: Custom Query "saved as" Report -> #1735 |
| [07:56:11] |
<alect> |
i figured application/x-trac-ticket for tickets? |
| [07:58:32] |
<cboos> |
alect: the new content converter API is ok, except in the example, we should not promote using 9 for the quality ratio (better use 8 instead) |
| [07:59:15] |
<alect> |
ok |
| [07:59:40] |
<cboos> |
speaking about api changes, what about adding the formatter as an argument to the render_macro method? |
| [07:59:58] |
<BaconTastesGood> |
any tips on permissions/user management using trac-admin? Specifically, is there a shortcut to say "Add all permissions" or "Remove all permissions" without listing everything? |
| [08:01:55] |
<alect> |
cboos: feels a bit hackish checking the arity, but other than that it sounds like a good idea, and i can't think of a better way to do it other than adding a new method (eg. render_macro_formatter) |
| [08:02:27] |
<cboos> |
svn style: render_macro2 ;) |
| [08:02:32] |
<alect> |
heh |
| [08:02:48] |
<alect> |
it's a tricky thing, api compatibility |
| [08:03:49] |
<cboos> |
yes, checking the arity is a bit hackish. We should certainly add simple api versioning to the Interface/Component stuff |
| [08:04:00] |
* |
cmlenz has quit IRC |
| [08:04:44] |
<alect> |
or use setuptools dependencies in trac itself |
| [08:05:00] |
* |
perchr has joined #trac |
| [08:05:01] |
<alect> |
so plugins say "trac==0.10dev" |
| [08:05:17] |
<alect> |
i'll remove IHTMLPreviewRenderer |
| [08:05:52] |
<BaconTastesGood> |
how do I disable svn in the conf file again? |
| [08:06:19] |
<cboos> |
alect: I would prefer a finer grained versioning scheme... that way we could handle evolution of each interface independantly |
| [08:06:47] |
<alect> |
true, so plugins depending on an interface that hasn't changed will still work |
| [08:06:56] |
<cboos> |
yes, take the example of the macro interface we are just discussing. |
| [08:08:22] |
<alect> |
agreed |
| [08:08:29] |
<cboos> |
the 0.10 WikiSystem will now how to handle IWikiMacroProvider version 1 (render_macro) and version 2 (render_macro_formatter), |
| [08:09:06] |
<alect> |
or, version 2 could be just a different method signature |
| [08:09:19] |
<alect> |
if the api versioning handles that |
| [08:10:18] |
<cboos> |
sure, that would the way to avoid annoying renames ;) |
| [08:10:27] |
<alect> |
yeah :) |
| [08:10:32] |
<BaconTastesGood> |
where is a list of components? |
| [08:10:43] |
<cboos> |
About/plugins |
| [08:10:49] |
<BaconTastesGood> |
TracIni Wiki entry says "See the Plugins page on About Trac" |
| [08:10:58] |
<BaconTastesGood> |
but for some reason the Wiki doesn't link to it |
| [08:10:59] |
<cboos> |
(be logged as admin, though) |
| [08:11:08] |
<cboos> |
(see above reason) |
| [08:11:29] |
<BaconTastesGood> |
ah, I need to be *_ADMIN? |
| [08:11:45] |
<alect> |
bacon does taste good, it's true |
| [08:11:56] |
<cboos> |
don't remember which one, I think it's the mighty TRAC_ADMIN itself ;) |
| [08:12:12] |
<BaconTastesGood> |
is there an easy way to add all permissions to a user using trac_admin? |
| [08:12:14] |
<alect> |
so Interfaces would need versions, and probably Components |
| [08:12:18] |
<BaconTastesGood> |
instead of typing them all out? |
| [08:12:19] |
<alect> |
TRAC_ADMIN is all permissions |
| [08:12:36] |
<alect> |
or you can assign them to a group, then assign the users to that group |
| [08:12:36] |
<BaconTastesGood> |
so simply doing trac_admin permissions add admin TRAC_ADMIN gives a superset? |
| [08:12:40] |
<BaconTastesGood> |
gotcha |
| [08:13:09] |
<alect> |
probably modules themselves |
| [08:13:17] |
<alect> |
api versioning smells like effort and pain |
| [08:15:20] |
<alect> |
this IContentConverter could be used for the query alternate formats |
| [08:15:23] |
* |
alect does that |
| [08:15:40] |
* |
ged_ has quit IRC |
| [08:15:54] |
<cboos> |
alect: interface versioning: don't use it until you need it... i.e. if we introduce an API change, then we simply add a static member version = 2; all components are then assumed to have "version == 1" |
| [08:16:18] |
<alect> |
yes good idea |
| [08:16:48] |
<alect> |
you'd probably need to change the implement() method |
| [08:17:18] |
<alect> |
eg. implements(IContentConverter=1, IRequestHandler) |
