| [00:24:35] |
* |
columbo has joined #trac |
| [01:12:51] |
<mitsuhiko> |
matt_good: only xml people know why xmlrpc is bad |
| [01:13:02] |
<mitsuhiko> |
i agree. it has a huge overhead |
| [01:13:05] |
<mitsuhiko> |
but so has soap |
| [01:13:59] |
<coderanger> |
what overhead? bandwidth and CPU time are cheap, and if it matches the model of what you are trying to do its not worth the hassle of using a "thinner" protocol |
| [01:14:10] |
<coderanger> |
mental overhead matters way more |
| [01:15:45] |
<mitsuhiko> |
all hail jsonrpc |
| [01:16:25] |
<coderanger> |
For webapps, yeah |
| [01:16:38] |
<coderanger> |
but its no less overhead if you are using !JS |
| [01:17:03] |
<coderanger> |
just a different package for the same idea and system |
| [01:21:15] |
<mitsuhiko> |
coderanger: hahahaha |
| [01:21:23] |
<mitsuhiko> |
the overhead is much less |
| [01:21:37] |
<coderanger> |
how so? |
| [01:21:44] |
<coderanger> |
you still need to do text parsing |
| [01:21:54] |
<coderanger> |
and like I said, CPU time is cheap |
| [01:22:18] |
<mitsuhiko> |
coderanger: http://paste.pocoo.org/show/410/ |
| [01:22:25] |
<mitsuhiko> |
that's what xmlrpc looks like |
| [01:22:46] |
<mitsuhiko> |
json looks more like [1, 2, 3, 4, 5, "six", "seven", "eight"] |
| [01:23:01] |
<coderanger> |
yes |
| [01:23:06] |
<coderanger> |
thats still text paring |
| [01:23:23] |
<coderanger> |
be it XML or JSON, it rally doesnt matter |
| [01:23:25] |
<coderanger> |
if it takes 1 ms or 2, who cares |
| [01:23:43] |
<mitsuhiko> |
coderanger: if i transfer a list of 2000 entries over xmlrpc |
| [01:23:59] |
<mitsuhiko> |
it takes half a minute to transfer and parse that |
| [01:25:03] |
<coderanger> |
okay, why are you sending 2000 entries? |
| [01:25:21] |
<coderanger> |
There is no way someone would every need to look at 2000 pieces of data |
| [01:25:48] |
<coderanger> |
people generally move on after a few hundred |
| [01:27:06] |
<coderanger> |
so extract what you need to on the server, and only send the relevant data |
| [01:34:29] |
<mitsuhiko> |
coderanger: that was just an example |
| [01:36:39] |
<coderanger> |
Yes, I'm just saying that these huge dataset/extreme repetition examples don't match the way XML-RPC is actually (or at least should be) used |
| [01:36:50] |
<coderanger> |
its not a bulk data transfer protocol |
| [01:38:01] |
<mitsuhiko> |
coderanger: that doesn't matter |
| [01:38:29] |
<mitsuhiko> |
if someone would have thought more about it when he wrote the specification the amount of transported data would be much smaller |
| [01:39:17] |
<coderanger> |
bandwidth is cheap |
| [01:39:37] |
<coderanger> |
if it is easy to use XML-RPC, and in many places it is, use it |
| [01:39:53] |
<coderanger> |
your time as a developer matter more than how many packets the router sees |
| [01:40:30] |
<mitsuhiko> |
don't get me wrong |
| [01:40:36] |
<mitsuhiko> |
i like xmlrpc and use it very often |
| [01:40:43] |
<mitsuhiko> |
nevertheless it could be better |
| [01:41:28] |
<coderanger> |
it could be, but it passes the JFW test |
| [01:44:19] |
* |
tessier_ has left #trac |
| [01:46:00] |
* |
mOolight has joined #trac |
| [01:47:48] |
* |
m_g has joined #trac |
| [01:47:48] |
* |
tuxipuxi has joined #trac |
| [01:49:02] |
<mOolight> |
i miss a theme which renders really fast for ie, but i can't find it now ... sadly |
| [01:49:23] |
<coderanger> |
there aren't many themes yet |
| [01:49:34] |
<coderanger> |
considering all the ones that exist are made by me |
| [01:49:54] |
<mOolight> |
there's some not listed in trac-hacks |
| [01:50:05] |
<coderanger> |
hmm? |
| [01:50:07] |
<coderanger> |
where |
| [01:50:53] |
<mOolight> |
i'll be glad if someone can remind me |
| [01:51:22] |
<coderanger> |
I certainly don't know of any public ones |
| [01:51:42] |
<mOolight> |
thx |
| [01:55:37] |
<coderanger> |
If you find any, ask them to post them on t-h |
| [01:57:41] |
<mOolight> |
sure |
| [02:01:01] |
* |
pombreda has joined #trac |
| [02:02:10] |
* |
converter has quit IRC |
| [02:19:38] |
* |
Crhysalis has joined #trac |
| [02:19:52] |
<Crhysalis> |
Hi everyone, where has the workflow branch in the sandbox gone? is it now merged into the trunk? |
| [02:21:46] |
<coderanger> |
Nope, being restarted |
| [02:25:01] |
<Crhysalis> |
oh why is that? |
| [02:25:05] |
<Crhysalis> |
any idea when it will be ready? |
| [02:26:28] |
<coderanger> |
because it will likely be less work to start over than adapt the 0.10-based version |
| [02:26:31] |
<coderanger> |
nope |
| [02:27:25] |
<coderanger> |
what specific features were you looking for? |
| [02:29:18] |
* |
kim0 has joined #trac |
| [02:31:16] |
<kim0> |
Hi, can trac put quota on a project (or does regular Linux quota work?) |
| [02:32:24] |
<coderanger> |
quota? |
| [02:32:32] |
<coderanger> |
you mean how much disk space it can use |
| [02:32:43] |
<kim0> |
yes |
| [02:32:56] |
<coderanger> |
you could put a quota on the user trac is running as |
| [02:33:04] |
<coderanger> |
not sure I've ever heard of anyone doing that though |
| [02:33:13] |
<kim0> |
that wouldnt be per project |
| [02:33:34] |
<kim0> |
are all files in the revision control owned by trac? |
| [02:34:18] |
<coderanger> |
No |
| [02:34:32] |
<coderanger> |
You would have to run each project as a different user |
| [02:34:50] |
<kim0> |
is this easy/standard |
| [02:34:51] |
<coderanger> |
Trac only needs read access to the subversion repo |
| [02:34:56] |
<coderanger> |
No |
| [02:35:00] |
<kim0> |
ok ... |
| [02:35:14] |
<kim0> |
one more thing, can trac handle binray files in repo ? |
| [02:35:18] |
<kim0> |
I know svn can |
| [02:35:19] |
<coderanger> |
If you are concerned, disable attachments |
| [02:35:21] |
<coderanger> |
indeed |
| [02:35:31] |
<coderanger> |
well what do you mean "handle"? |
| [02:35:39] |
<kim0> |
& stores bin-diffs & all |
| [02:35:48] |
<coderanger> |
Trac is just a viewer, itl spit back whatever data it sees |
| [02:35:50] |
<kim0> |
binary difference I guess |
| [02:35:54] |
<coderanger> |
Trac doesn't do any of that |
| [02:35:57] |
<coderanger> |
subversion does |
| [02:36:07] |
<kim0> |
ok ... cool |
| [02:36:12] |
<coderanger> |
Trac is not a replacement for subversion, it is a viewer |
| [02:36:29] |
<kim0> |
I understand, and since this is binary data, trac wont do anything to it |
| [02:36:57] |
<coderanger> |
itl let you download it |
| [02:37:17] |
<coderanger> |
and if it is a known image type it will display it (if the filesize is under a given limit) |
| [02:37:40] |
<coderanger> |
Thats about it, viewing bindiffs in a web UI would be somewhat difficult |
| [02:37:52] |
<kim0> |
ah ... and can users commit to svn directly (not thru trac) ? that wouldnt cause issues ? |
| [02:38:12] |
<coderanger> |
You can't "commit through Trac" |
| [02:38:17] |
<coderanger> |
it is just a viewer |
| [02:38:33] |
<coderanger> |
like ViewVC or CVCWeb |
| [02:38:38] |
<coderanger> |
er, CVSWeb |
| [02:38:41] |
<kim0> |
so if I have a svn repo for two years, I can just put trac on it |
| [02:38:48] |
<coderanger> |
Yes |
| [02:38:53] |
<kim0> |
cool stuff |
| [02:39:02] |
* |
Synapse_ is now known as Synapse |
| [03:17:49] |
* |
_shawn has quit IRC |
| [03:51:16] |
* |
dmax has joined #trac |
| [03:52:51] |
* |
s0undt3ch has joined #trac |
| [04:16:26] |
* |
converter has joined #trac |
| [04:26:51] |
* |
converter has quit IRC |
| [04:39:41] |
* |
bryanl has joined #trac |
| [04:48:51] |
* |
kop_ has joined #trac |
| [05:09:36] |
* |
_shawn has joined #trac |
| [06:08:59] |
* |
idnar_ is now known as idnar |
| [06:12:41] |
* |
m_g has quit IRC |
| [06:25:35] |
<blinx> |
coderanger: ping |
| [06:38:13] |
* |
kim0 has left #trac |
| [06:42:12] |
* |
mOolight has quit IRC |
| [06:46:23] |
<blinx> |
anyone knows where the channellogs are? |
| [06:47:42] |
<blinx> |
ok |
| [07:06:57] |
<coderanger> |
blinx: Check the topic |
| [07:07:50] |
<blinx> |
coderanger: done ;-) |
| [07:07:57] |
<blinx> |
coderanger: I played a bit with the css |
| [07:08:03] |
<blinx> |
but now it works |
| [07:08:39] |
<blinx> |
on your testsite it doesn't work -> see the sidebar the browse source.. |
| [07:09:38] |
<coderanger> |
very possible, its not done yet |
| [07:14:52] |
<coderanger> |
Also I see no difference on my test site in the browser ... |
| [07:18:03] |
<blinx> |
/usr/lib/python2.4/site-packages/trac/db/util.py:51: Warning: Truncated incorrect DOUBLE value: ' *' |
| [07:18:06] |
<blinx> |
return self.cursor.execute(sql) |
| [07:18:09] |
<blinx> |
hjm |
| [07:18:11] |
<blinx> |
what is that for an error? |
| [07:18:33] |
<coderanger> |
I would need way more information than that |
| [07:18:55] |
<blinx> |
hm |
| [07:19:20] |
<blinx> |
# print repr(args) |
| [07:19:20] |
<blinx> |
return self.cursor.execute(sql_escape_percent(sql), args) |
| [07:19:21] |
<blinx> |
return self.cursor.execute(sql) |
| [07:19:30] |
<blinx> |
that are the lines of util.py |
| [07:19:39] |
<blinx> |
the last is line 51 |
| [07:22:06] |
<coderanger> |
more information means a traceback/what you were doing/what DB you are using |
| [07:28:49] |
* |
m_g has joined #trac |
| [07:39:42] |
<m_g> |
coderanger, hey there.. are you sure that your fix for HierWiki fixed my problem? it's still the same here apparently so i wonder if i did something wrong |
| [07:39:55] |
<m_g> |
did svn up, bdist_egg, moved the egg and restarted webserver |
| [07:40:03] |
<coderanger> |
It works on a test site here |
| [07:40:27] |
<coderanger> |
I can go look again |
| [07:40:38] |
<m_g> |
great :) |
| [07:40:53] |
<m_g> |
the heading which causes it to break is == "Just a Test" == |
| [07:41:13] |
<m_g> |
= "Just a Test" = i mean |
| [07:43:57] |
<coderanger> |
yep, the HTML looks fine |
| [07:44:10] |
<coderanger> |
the value in the title attr is indeed encoded |
| [07:44:53] |
<coderanger> |
http://coderanger.net/tracs/tracdev/wiki/WikiStart/Top |
| [07:46:52] |
<coderanger> |
Double check you don't still have the old-style macro too |
| [07:46:59] |
<m_g> |
coderanger, in my case the macro is used like [[SubWiki(Test/,1)]] .. could that make a difference? |
| [07:47:17] |
<m_g> |
i have moved SubWiki.py to SubWiki2.p |
| [07:47:48] |
<m_g> |
arg |
| [07:47:51] |
<m_g> |
wait :) |
| [07:48:00] |
<coderanger> |
http://coderanger.net/tracs/tracdev/wiki/TestPage |
| [07:49:54] |
<m_g> |
coderanger, ok i had the old .egg file lying around somewhere, my fault.. but.. your fix has changed the behavior of the plugin |
| [07:50:08] |
<m_g> |
before, it displayed the hierachy (Test/<....>) and the page name |
| [07:50:19] |
<m_g> |
now only the heading as it seems |
| [07:50:37] |
<coderanger> |
Hrmm, thought I put that back |
| [07:50:42] |
* |
coderanger checks |
| [07:53:19] |
<coderanger> |
Ahh, I see |
| [07:55:54] |
<coderanger> |
http://coderanger.net/tracs/tracdev/wiki/TestPage |
| [07:55:56] |
<coderanger> |
correct now? |
| [07:57:59] |
<m_g> |
coderanger, hmm not sure.. actually it should just display the last part of the hierachy, i.e. Top/<page names> |
| [07:59:11] |
<coderanger> |
Not from the old code |
| [07:59:46] |
<m_g> |
or does the ,1 control this probably? |
| [08:00:29] |
<coderanger> |
Nope, the name displayed is the actual name of the page |
| [08:00:41] |
<m_g> |
hmm ok.. could you commit the new version? |
| [08:00:46] |
<coderanger> |
done |
| [08:01:01] |
<m_g> |
thanks, i will try it in a few minutes |
| [08:03:47] |
* |
dj-atx has joined #trac |
| [08:04:08] |
* |
dj-atx has left #trac |
| [08:11:27] |
<m_g> |
coderanger, perfect! thanks a bunch :) |
| [08:11:32] |
<m_g> |
a ton |
| [08:38:47] |
* |
ZyX has joined #trac |
| [08:38:58] |
<ZyX> |
how do i change the title of a wiki page? |
| [08:39:18] |
<coderanger> |
title or name? |
| [08:39:23] |
<ZyX> |
name |
| [08:39:29] |
<coderanger> |
WikiRename plugin |
| [08:39:56] |
<ZyX> |
why isnt that supported directly? |
| [08:40:13] |
<coderanger> |
Because its somewhat hackish |
| [08:40:28] |
<coderanger> |
There is not currently a good way to figure who links to where |
| [08:40:36] |
<coderanger> |
so rewritting links is tricky |
| [08:40:46] |
<ZyX> |
that page isnt linked |
| [08:40:56] |
<coderanger> |
the plugin only handles explicit ([wiki:Foo]) links from other wiki pages |
| [08:43:05] |
<ZyX> |
if i dont care for linking pages, is there a simple way to rename a page? sqlite? |
| [08:43:10] |
<coderanger> |
Just use the plugin |
| [08:43:16] |
<coderanger> |
Its easier |
| [08:44:14] |
<coderanger> |
easy_install TracWikiRename to install it |
| [08:45:15] |
<ZyX> |
that command gives me http://pastebin.ca/282479 |
| [08:45:43] |
<coderanger> |
easy_install -N TracWikiRename |
| [08:45:51] |
<coderanger> |
thatl ignore dependencies |
| [08:46:02] |
<ZyX> |
hm, is that a good idea? |
| [08:46:11] |
<coderanger> |
If you already have WebAdmin installed, yes |
| [08:46:21] |
<coderanger> |
The problem is that it isnt listed on pypi |
| [08:46:34] |
<coderanger> |
which I should mention to matt at some point |
| [08:46:39] |
<ZyX> |
ok |
| [08:46:50] |
<coderanger> |
matt_good: Mind if I list webadmin and spamfilter on pypi? |
| [09:02:32] |
* |
bryanl has quit IRC |
| [09:08:46] |
* |
ZyX has quit IRC |
| [09:18:43] |
* |
bryanl has joined #trac |
| [09:47:51] |
* |
Crhysalis has quit IRC |
| [10:04:04] |
* |
pombreda has quit IRC |
| [10:04:54] |
* |
independence has joined #trac |
| [10:05:16] |
<independence> |
Can I make a robots.txt to make search robots not index my trac projects? |
| [10:12:49] |
<malikeye> |
think my python is busted |
| [10:13:01] |
<malikeye> |
can't use easy_install |
| [10:23:33] |
* |
Naked has joined #trac |
| [10:23:48] |
* |
Naked is now known as Hadaka |
| [10:34:17] |
* |
independence has left #trac |
| [10:34:40] |
<malikeye> |
nevermind, I'm a dumbass |
| [11:34:05] |
* |
dgynn has joined #trac |
| [12:51:38] |
* |
PlasmaHH has joined #trac |
| [12:52:15] |
<PlasmaHH> |
hi, im trying to setup some new trac environment with postgres as db. unfortunately my password has some @ in it. anyone knows how I can still feed it so trac uses it properly? |
| [12:59:52] |
* |
bofh has joined #trac |
| [13:01:56] |
* |
bofh is now known as l0ki |
| [13:02:01] |
* |
l0ki is now known as lowkey |
| [13:02:25] |
* |
lowkey is now known as l0wkey |
| [13:04:17] |
* |
dort has quit IRC |
| [13:09:46] |
* |
omry has quit IRC |
| [13:33:08] |
<PlasmaHH> |
is it possible/recommended to run two trac environments on the same webserver, but under two different vhosts? I seem to be getting errors in one when I used the other for a while and vv |
| [13:36:42] |
<PlasmaHH> |
this is the traceback I now always get: http://rafb.net/paste/results/3KwWyP23.html |
| [13:40:43] |
* |
l0wkey has quit IRC |
| [13:48:25] |
* |
m_g has quit IRC |
| [13:53:13] |
* |
matt_good has quit IRC |
| [14:01:35] |
* |
stepz has joined #trac |
| [14:18:16] |
* |
PlasmaHH_ has joined #trac |
| [14:20:21] |
* |
PlasmaHH has quit IRC |
| [14:20:22] |
* |
PlasmaHH_ is now known as PlasmaHH |
| [14:27:57] |
* |
julesa_ has joined #trac |
| [14:30:30] |
<julesa_> |
Can Trac receive inbound requests via email (like RT)? I looked in the faq and it wasn't obvious if this exists. |
| [14:30:37] |
* |
julesa_ is now known as julesa |
| [14:33:15] |
<julesa> |
Or is this as good as it gets? http://weblog.greenpeace.org/it/2006/07/trac_via_email_the_hard_way.html |
| [14:34:44] |
<neuralis> |
not as part of core trac, though i think i saw a plugin for it on trac-hacks. |
| [14:49:23] |
* |
matt_good has joined #trac |
| [14:50:47] |
* |
stepz has quit IRC |
| [14:59:06] |
* |
julesa has quit IRC |
| [15:01:31] |
<matt_good> |
PlasmaHH: it's a problem with the SVN bindings when they're used in multiple interpreters from the same process |
| [15:02:07] |
<matt_good> |
PlasmaHH: in each Trac section with "SetHandler mod_python" add "PythonInterpreter main_interpreter" |
| [15:02:10] |
<malikeye> |
doesn't seem to be a way to give someone the permission setting abilities from accountmgr |
| [15:02:43] |
<matt_good> |
malikeye: AccountManager doesn't do permission since that's already handled by WebAdmin |
| [15:02:55] |
<malikeye> |
ahh... then I mean webadmin :) |
| [15:03:02] |
* |
PlasmaHH has quit IRC |
| [15:03:29] |
<malikeye> |
there is of course a security concern with being able to do that, cause you can give yourself TRAC_ADMIN |
| [15:04:07] |
<matt_good> |
you need to have the TRAC_ADMIN permission to manage permissions |
| [15:04:12] |
<malikeye> |
yup |
| [15:04:19] |
<malikeye> |
maybe PERMISSION ADMIN ? |
| [15:04:26] |
<malikeye> |
CUSTOM_ADMIN ? |
| [15:04:46] |
<matt_good> |
oh, you want a new permission for managing permissions |
| [15:05:17] |
<matt_good> |
but as you said, once you give that to someone they could elevate their own permissions, so why not just give them TRAC_ADMIN to begin with? |
| [15:05:26] |
<malikeye> |
I want to give ticket_admin and most of the permission abilities |
| [15:05:38] |
<malikeye> |
exactly |
| [15:06:14] |
<matt_good> |
so what's the problem? |
| [15:06:43] |
<malikeye> |
the only way to give permission abilities is with trac_admin |
| [15:07:02] |
<malikeye> |
I don't wanna give trac_admin |
| [15:07:10] |
<malikeye> |
it will be a shared environment |
| [15:07:22] |
* |
sid506 has joined #trac |
| [15:07:47] |
<malikeye> |
but the ability to have a sub-admin would be cool |
| [15:08:08] |
<malikeye> |
unless I'm missing something |
| [15:09:06] |
<matt_good> |
I don't see the point in making a permission that would allow users to elevate their own permissions |
| [15:09:30] |
<malikeye> |
I want to have the ability for someone other than "root" to add users to a project |
| [15:09:54] |
<malikeye> |
is that possible ? |
| [15:09:58] |
<matt_good> |
ah |
| [15:09:58] |
* |
kop_ has quit IRC |
| [15:10:24] |
<matt_good> |
well, AccountManager could have a separate permission for adding or removing users |
| [15:10:26] |
<malikeye> |
sorta like a "team lead" type role |
| [15:10:50] |
<matt_good> |
ACCOUNT_CREATE, ACCOUNT_DELETE, etc. |
| [15:11:01] |
<malikeye> |
no |
| [15:11:13] |
<malikeye> |
the whole thing is open for registerations |
| [15:11:53] |
<malikeye> |
but to add like TICKET_ADMIN to the new hire so they can work on tickets |
| [15:12:02] |
<matt_good> |
ok |
| [15:12:34] |
<malikeye> |
right now it's all er nothin |
| [15:12:47] |
<matt_good> |
hrm, I suppose there could be a PERMISSION_ADMIN permission, and users would only be able to grant permissions that they have themselves |
| [15:13:11] |
<malikeye> |
that's a good idea |
| [15:13:20] |
<malikeye> |
only what they currently have |
| [15:14:01] |
<matt_good> |
right, since you also wouldn't want them creating a new account with elevated permissions |
| [15:14:10] |
<malikeye> |
exactly |
| [15:14:37] |
<malikeye> |
I wouldn't usurp the account creation either though |
| [15:14:56] |
<malikeye> |
only be able to manipulate already created accounts |
| [15:15:46] |
<malikeye> |
but that could probably be functioned out too |
| [15:16:37] |
<sid506> |
heya - does anyone know what function to use to convert a trac link to the path. For example, convert #1 to /ticket/1 |
| [15:16:54] |
<sid506> |
and WikiStart to /wiki/WikiStart ? |
| [15:19:38] |
<matt_good> |
sid506: trac.wiki.formatter.wiki_to_link |
| [15:19:57] |
<sid506> |
matt_good: thanks for that.. but it does a little too much |
| [15:20:20] |
<malikeye> |
is there a way to use a permission template ? |
| [15:20:34] |
<matt_good> |
malikeye: groups? |
| [15:20:56] |
<malikeye> |
k |
| [15:21:16] |
<sid506> |
it adds the <a href=""> around the text |
| [15:21:31] |
<sid506> |
(looks like cboos moved it too.. it is now part of the Context code) |
| [15:22:07] |
<malikeye> |
hmm, not really groups I guess.... when I do an initenv, I am assuming it sets up a default set of rules for permissions |
| [15:23:04] |
<malikeye> |
basically it sets open to anon |
| [15:24:28] |
<matt_good> |
oh |
| [15:24:36] |
<malikeye> |
I guess this is all stuff I can script out |
| [15:24:58] |
<malikeye> |
a post_script |
| [15:25:03] |
<malikeye> |
or something |
| [15:25:29] |
<matt_good> |
yeah, you can put a bunch of trac-admin commands into a file and run it like "trac-admin /your/project < commands.txt" |
| [15:25:29] |
* |
omry has joined #trac |
| [15:46:08] |
* |
stevegt has quit IRC |
| [15:46:17] |
<s0undt3ch> |
is there any python module that does word wrap? |
| [15:50:44] |
<prologic> |
prolly |
| [15:50:48] |
<prologic> |
I think urwid does |
| [15:50:52] |
<prologic> |
but thats a curses interface |
| [15:51:17] |
<matt_good> |
s0undt3ch: yeah, "textwrap" |
| [15:51:24] |
<matt_good> |
Trac uses it for the email formatting |
| [15:51:38] |
<prologic> |
there ya go :) |
| [15:51:38] |
<s0undt3ch> |
matt_good: no probs with unicode are there? |
| [15:52:02] |
<matt_good> |
it works right if you use unicode objects |
| [15:52:24] |
<matt_good> |
passing it utf-8 encoded strs will cause problems |
| [15:52:42] |
<s0undt3ch> |
oh |
| [15:52:50] |
<s0undt3ch> |
hmm, that's not good :\ |
| [15:53:14] |
<matt_good> |
well, just decode the str objects to an actual unicode object first |
| [15:53:15] |
<s0undt3ch> |
although text passed to with might all be ascii |
| [15:53:34] |
<s0undt3ch> |
s/to with/to it/ |
| [15:53:45] |
<s0undt3ch> |
k, thanks |
| [15:54:22] |
<matt_good> |
textwrap doesn't know what encoding a str is using, so multi-byte characters screw up the character count |
| [15:59:55] |
* |
sproust has joined #trac |
| [16:00:43] |
<sproust> |
morning ladies |
| [16:00:50] |
<prologic> |
speak for yourself :) |
| [16:00:57] |
* |
prologic checks |
| [16:00:59] |
<prologic> |
yup all man :) |
| [16:01:23] |
<sproust> |
blitz intro: me trac user, dying for a subversion wiki backend. |
| [16:01:40] |
<sproust> |
another co-worker also has the same problem. |
| [16:01:41] |
<prologic> |
heh |
| [16:01:51] |
<prologic> |
why a subversion wiki backend ? |
| [16:01:56] |
<s0undt3ch> |
yeah, why? |
| [16:02:06] |
<prologic> |
you just need the current backend to support merging |
| [16:02:17] |
<sproust> |
looked at the TighterSubversionIntegration page and the old patch and seriously thinking about reviving it. |
| [16:02:23] |
<prologic> |
which wouldn't be hard with the right plugin given that the right hook exists now |
| [16:02:25] |
<prologic> |
(if it does) |
| [16:02:33] |
<sproust> |
in fact, i'm in the office today sunday in Beverly Hills to look at this. |
| [16:02:45] |
<sproust> |
I should be at the beach, really. |
| [16:02:57] |
<prologic> |
then go |
| [16:02:58] |
<prologic> |
:) |
| [16:02:59] |
<sproust> |
soundtech: why? I hate editing stuff in a TEXTAREA. I want to use emacs |
| [16:03:08] |
<sproust> |
i want to grep, rename, add attachments, etc. |
| [16:03:13] |
<s0undt3ch> |
ah |
| [16:03:19] |
<prologic> |
oh ic |
| [16:03:19] |
<s0undt3ch> |
good reason :) |
| [16:03:21] |
<sproust> |
seriously. |
| [16:03:22] |
<prologic> |
yeah I hate that too |
| [16:03:24] |
<prologic> |
I'd rather use vim |
| [16:03:31] |
<prologic> |
how would you accomplish this ? |
| [16:03:37] |
<sproust> |
should have been like that from the start IMO, but I guess the original forces were different. |
| [16:04:40] |
<matt_good> |
malikeye: I've updated the Trac trunk with permissions like we talked about, though you'll probably want them in the WebAdmin plugin |
| [16:04:43] |
<sproust> |
so I had a quick look around, and I would do this: change trac/wiki/model.py to support an abstraction for the wiki backend. |
| [16:04:56] |
<sproust> |
e.g. probably create another kind of WIkiPage object. |
| [16:05:06] |
<sproust> |
and dispatch to a specific wiki backend. |
| [16:05:17] |
<malikeye> |
matt_good: uh... ok |
| [16:05:21] |
<sproust> |
I would also have to modify some of the other code that makes use of the wiki table, e.g. macros.py |
| [16:05:46] |
<sproust> |
prologic: can you expand on your plugin and "right hook" idea? |
| [16:05:54] |
<sproust> |
prologic: me not that familiar with trac source code. |
| [16:05:58] |
<sproust> |
(very pretty code BTW) |
| [16:06:21] |
<sproust> |
I'd like do code this in a branch, can I get perms to the repo? |
| [16:06:41] |
<sproust> |
(here's me: http://furius.ca/) |
| [16:07:12] |
<sproust> |
BTW I created Nabu for similar reasons, I wanted to create blog pages without editing TEXTAREAs. |
| [16:07:26] |
<sproust> |
(maybe I should just embed an Emacs in Firefox instead :-)) |
| [16:08:29] |
<matt_good> |
sproust: we don't usually give repo permissions without someone submitting some patches first |
| [16:08:41] |
<sproust> |
I see. I'll work on a patch then. |
| [16:08:47] |
<matt_good> |
also, take a look at http://trac.edgewall.org/ticket/590 |
| [16:09:12] |
<prologic> |
well coderanger told me ages ago that some hook was needd |
| [16:09:14] |
<prologic> |
in order to build a plugin that added merging into wiki-page edits |
| [16:09:21] |
<prologic> |
ie: two or more people editing the same page but different parts and mergning them |
| [16:09:26] |
<matt_good> |
there's a protocol for launching an external editor and then posting the content back |
| [16:09:38] |
<sproust> |
about 590: I also looked at trac-wiki.el for Emacs, that's probably easier, and partial solution that doesn't help others... |
| [16:09:39] |
<prologic> |
this was back in 0.9.6 |
| [16:09:57] |
<sproust> |
prologic: about merging, I assume you mean keeping a repo and the current DB storage in sync automatically? |
| [16:09:59] |
<prologic> |
I think a hook was needed just before the page was saved |
| [16:10:33] |
<matt_good> |
sproust: I think there's also a Firefox plugin that supports opening any text area in an external editor |
| [16:10:57] |
<sproust> |
matt: saw it, not very reliable, and have no time to hack FF, it's a bit of madness |
| [16:11:24] |
<prologic> |
sproust, no merging of changes in wiki-page edits |
| [16:11:35] |
<prologic> |
currently two or more users cannot edit the same page at the same time in Trac |
| [16:11:37] |
<sproust> |
prologic: i see. |
| [16:11:42] |
<sproust> |
right |
| [16:11:43] |
<prologic> |
the page becomes locked instead |
| [16:12:09] |
<sproust> |
that behaviour would be conservative for the subversion wiki backend. |
| [16:12:20] |
<prologic> |
perhaps |
| [16:12:26] |
<sproust> |
all: would modifying trac/wiki/* files be sufficient to implement this? Is this welcome? |
| [16:12:31] |
<prologic> |
but it's not such a bad idea building it into the current wiki engine |
| [16:12:49] |
<sproust> |
I saw the discussion on doing this or not, I'm heavily on the "doing this only makes sense" side. |
| [16:13:37] |
<sproust> |
the subversion format would be somethign like this: |
| [16:13:42] |
<sproust> |
MyFile.txt : wiki page |
| [16:13:51] |
<sproust> |
MyFile.attach : dir that contains attachments. |
| [16:13:55] |
<sproust> |
voila. |
| [16:14:07] |
<sproust> |
In a separate repo, for simplicity. |
| [16:14:43] |
<sproust> |
also, having a s |
