| [00:15:00] |
* |
onny has joined #trac |
| [00:50:50] |
<onny> |
sorry for this offence question, but is there a php alternative for trac ? the software is great, but python unfortionatly disqualifies on my system :( |
| [00:51:50] |
<onny> |
something with a similar feature set, most importantly milestone management and bug ticket system, the other features are optional, its just for 3 ppl to manage a small software project, plan featuers and track bugs |
| [00:53:30] |
<coderanger> |
Not that I know of |
| [00:53:44] |
<coderanger> |
there is something called primo planner, that I think is in PHP |
| [00:53:54] |
<coderanger> |
but its not a terribly good replacement |
| [00:54:00] |
<coderanger> |
why can't you use python though? |
| [00:54:11] |
<coderanger> |
Its installed on basically every flavor of unix by default |
| [00:55:32] |
<onny> |
in apache i will need mod_python ? |
| [00:55:48] |
<onny> |
would the install of trac be easy and not time consuming if my server would have this python plugin ? |
| [00:55:50] |
<coderanger> |
No |
| [00:56:06] |
<coderanger> |
Trac runs on a large number of frontends |
| [00:56:15] |
<coderanger> |
it can run as a simple CGI script (though thats slow) |
| [00:56:35] |
<coderanger> |
you can also use FastCGI or tracd (a webserver that comes with Trac) |
| [00:56:37] |
<onny> |
hm ok. its a standard debian server system and if it lacks python i can let it install for free by the maintainer, but a php version would be so much less time consuming |
| [00:56:52] |
<onny> |
yea that all sounds to complicated |
| [00:57:01] |
<coderanger> |
You sure debian doesn't include python by default? |
| [00:57:10] |
<onny> |
i want to unpack an archieve, setup a database, create a admin account and start |
| [00:57:15] |
<onny> |
for more hacking i dont have time |
| [00:57:36] |
<onny> |
im still at work after 15 hours now, and i want to go home soon ;) |
| [00:57:43] |
<coderanger> |
Okay, so use CGI |
| [00:57:45] |
<coderanger> |
or tracd |
| [00:57:50] |
<onny> |
im not familiar with debian, that is 100% done by an employee |
| [00:58:02] |
<coderanger> |
they require no more config than PHP |
| [00:58:04] |
<onny> |
but im happy if he does more important stuff too, time is money, my money |
| [00:58:34] |
<onny> |
ok i will try that, once i have finished some other tasks, if it doesnt run atfer 10 minutes i search for another tool |
| [00:58:42] |
<coderanger> |
umm, okay |
| [00:58:55] |
<coderanger> |
you will probably then spend more time search than it would take to finish the setup :P |
| [00:59:14] |
<coderanger> |
other things you can look at are jtrac (in Java) and rtrac |
| [00:59:16] |
<onny> |
i know i can read that on the trac website too, but its just a yes no question, so excuse my lazyness: can i use trac commercially ? |
| [00:59:22] |
<coderanger> |
Yes |
| [00:59:26] |
<onny> |
thx alot m8 =) |
| [00:59:32] |
<onny> |
u safed me some time |
| [00:59:42] |
<coderanger> |
You can do anything, including redistirbute it, except claim you wrote it |
| [00:59:50] |
<onny> |
i continue with my work then... maybe i check back here when i run into strange install problems |
| [00:59:59] |
<coderanger> |
(thats the short version anyway) |
| [01:00:18] |
<onny> |
unfortionatly my server maintainer is at home and sleepin, so i cant ask him for assistance on this, he surly knows python in sleep |
| [01:00:23] |
<onny> |
i know python just by name |
| [01:00:49] |
<onny> |
i know i should sleep too, but no serious coder has more than a 5 day week :P |
| [01:01:23] |
<onny> |
escpially if he wants to pay his next month apartment and office fees .. :( |
| [01:04:23] |
* |
Sweetshark has joined #trac |
| [01:05:24] |
<bitsweat> |
if your time is super important, use a hosted trac |
| [01:05:54] |
<bitsweat> |
http://trac.edgewall.org/wiki/CommercialServices |
| [01:15:01] |
<onny> |
thats a ncie idea but unfortionatly a nodo thing for us |
| [01:19:52] |
* |
Sweetsha1k has quit IRC |
| [02:05:30] |
* |
LionsMane has quit IRC |
| [02:34:48] |
* |
m_g has joined #trac |
| [03:06:13] |
* |
bsdguru_away has quit IRC |
| [03:28:56] |
* |
pygi has quit IRC |
| [03:41:11] |
* |
thm has joined #trac |
| [04:11:48] |
* |
oversize has joined #trac |
| [04:27:36] |
<alect> |
evening |
| [04:38:03] |
* |
pygi has joined #trac |
| [04:44:53] |
* |
bryanl has joined #trac |
| [05:22:34] |
<onny> |
coderanger i contacted the server maintainer and he agreed to take over the whole project management software setup, he decided to go with mantis+subversion, so i had not to waste time on this topic for a single minute ;) |
| [05:23:26] |
<onny> |
i btw recommened him trac, but he said it has an experimental mysql support or something, and we use mysql in our company |
| [05:23:41] |
<onny> |
so he wants to take no risk, no matter how nice trac is... to bad :/ |
| [05:23:59] |
<pygi> |
I used mysql before, and worked like a charm |
| [05:25:37] |
<onny> |
hm, i think he has not yet begun installing the other 2 software packages, i coudl still recommend trac |
| [05:25:47] |
<onny> |
persoanlly i liked the easy of use of trac |
| [05:26:33] |
<onny> |
but one question (which ofc i could also find in the docs, but it would take 1000x longer) does trac looks on source codes on a server with versioning? like cvs or so ? |
| [05:27:14] |
<onny> |
i know im very unexperienced with cvs and version management, because i never in my live used any, so excuse this stupid question ;) |
| [05:27:34] |
<pygi> |
trac can't use cvs :) |
| [05:27:53] |
<pygi> |
it can use svn, git, bzr, monotone, and such tho |
| [05:28:16] |
<onny> |
hm whatever this means, but i assume its the directory structure we save a projects source code on the server with version management ? |
| [05:28:30] |
<onny> |
i did just programm 1-man projects the last 10 years ;) |
| [05:29:12] |
<onny> |
so all i know about cvs is how to update a source code on my computer from other ppl projects i want the latest daily code from |
| [05:29:40] |
<pygi> |
svn commit ; svn update ; svn copy ; svn delete ... |
| [05:29:46] |
<pygi> |
hm, have I missed anything important ... |
| [05:29:50] |
<pygi> |
svn add |
| [05:29:56] |
<pygi> |
svn import |
| [05:36:59] |
<onny> |
i think he installs tracs now too and then we try both systems simultanous for some days and make a final decision of which one we will use ;) |
| [05:37:49] |
* |
_dmax has joined #trac |
| [05:38:15] |
<pygi> |
^_^ |
| [05:42:34] |
* |
s0undt3c1_ has joined #trac |
| [05:44:52] |
* |
pygi has quit IRC |
| [05:46:46] |
* |
pygi has joined #trac |
| [05:50:11] |
<oversize> |
hello, is it possible to have the browser show different trees for different users? for ex: the repos has /repos/public/ /repos/private/ folders and only authenticated users can see /repos/private ? |
| [05:52:24] |
* |
lnayrb has joined #trac |
| [05:54:06] |
* |
dmax has quit IRC |
| [05:54:08] |
* |
_dmax is now known as dmax |
| [05:55:46] |
* |
s0undt3ch has quit IRC |
| [05:55:46] |
* |
s0undt3c1_ is now known as s0undt3ch |
| [05:55:46] |
* |
s0undt3ch_ is now known as s0undt3ch |
| [05:55:49] |
* |
s0undt3ch_ has joined #trac |
| [06:04:30] |
* |
jborg has joined #trac |
| [06:06:34] |
* |
bryanl has quit IRC |
| [06:10:45] |
* |
stevegt has quit IRC |
| [06:13:46] |
* |
cookin_ has joined #trac |
| [06:15:49] |
<bitsweat> |
the trac browser supports svn authz |
| [06:19:30] |
<bitsweat> |
oversize: http://trac.edgewall.org/wiki/FineGrainedPermissions |
| [07:02:06] |
* |
Moi has joined #trac |
| [07:02:32] |
<Moi> |
hi |
| [07:02:42] |
<Moi> |
am I in the 64 studio room? |
| [07:03:56] |
* |
Moi has quit IRC |
| [07:04:15] |
* |
Moi has joined #trac |
| [07:06:06] |
<Moi> |
I'm having problems getting into 64 studio, i installed picked only low res for xserver, but my monitor just keeps switching on and off when xserver starts, like the res used is too high... |
| [07:06:49] |
<Moi> |
I can switch to term, but dont know the commands for setting down the screen res, anyone help? |
| [07:07:49] |
<pygi> |
what are you talking about? :) |
| [07:09:20] |
<Moi> |
Erm, I'm in the right place for help with 64 studio yeah? |
| [07:10:36] |
<pygi> |
no :P |
| [07:11:48] |
<Moi> |
well, regardless, can anyone tell me the terminal command to switch the screen resolution down to say 800x600? |
| [07:12:19] |
<pygi> |
just change xorg.conf |
| [07:13:24] |
<Moi> |
ah, will do, ta. am very much a linux noob still |
| [07:20:04] |
* |
s0undt3ch has quit IRC |
| [07:20:04] |
* |
s0undt3ch_ is now known as s0undt3ch |
| [07:20:08] |
* |
s0undt3ch_ has joined #trac |
| [07:43:59] |
* |
PlasmaHH has joined #trac |
| [07:44:40] |
<PlasmaHH> |
hi, how can I put images into my wiki pages? |
| [07:49:51] |
* |
Moi_ has joined #trac |
| [07:50:40] |
* |
Moi has quit IRC |
| [07:57:16] |
* |
oversize has quit IRC |
| [07:57:19] |
* |
gak has joined #trac |
| [07:57:31] |
* |
oversize has joined #trac |
| [08:09:16] |
* |
dgynn has joined #trac |
| [08:16:33] |
* |
gakman has quit IRC |
| [08:16:41] |
* |
Moi_ has quit IRC |
| [08:23:26] |
* |
idnar has quit IRC |
| [08:25:19] |
* |
idnar has joined #trac |
| [08:51:52] |
<coderanger> |
PlasmaHH: The Image macro |
| [09:08:36] |
* |
lnayrb is now known as bryanl |
| [09:23:05] |
* |
cookin__ has joined #trac |
| [09:35:03] |
* |
LionsMane has joined #trac |
| [09:36:19] |
* |
malikeye has quit IRC |
| [09:40:28] |
* |
cookin_ has quit IRC |
| [09:41:00] |
* |
cookin_ has joined #trac |
| [09:41:38] |
* |
cookin__ has quit IRC |
| [09:48:53] |
* |
pygi has quit IRC |
| [09:49:53] |
* |
pygi has joined #trac |
| [09:52:56] |
* |
cookin__ has joined #trac |
| [09:54:15] |
* |
_shawn has quit IRC |
| [09:55:01] |
* |
_shawn has joined #trac |
| [10:11:31] |
* |
cookin_ has quit IRC |
| [10:55:38] |
* |
joost_ has joined #trac |
| [11:03:43] |
<joost_> |
Hi, need some help with trac+mod_python (httpd segfault) |
| [11:04:45] |
<joost_> |
tracd works fine, only mod_python gives me problems, similar to ticket #4456 and #3658 |
| [11:04:53] |
<joost_> |
trac-0.10.3, httpd-2.2.3, mod_python-3.2.10, clearsilver-0.10.3, sqlite-3.3.6, python-sqlite2-2.3.2 |
| [11:15:25] |
* |
hotte has joined #trac |
| [11:34:32] |
* |
lnayrb has joined #trac |
| [11:35:20] |
* |
bryanl has quit IRC |
| [11:35:31] |
* |
Tetraboy has joined #trac |
| [12:01:58] |
* |
LionsMane has quit IRC |
| [12:02:37] |
* |
LionsMane has joined #trac |
| [12:31:30] |
* |
openjack has joined #trac |
| [12:33:28] |
<openjack> |
I'm trying to install Trac on Centos, I'm also using Ravencore |
| [12:33:28] |
* |
jborg has quit IRC |
| [12:34:26] |
<openjack> |
I grabbed the clearsilver and python-clearsilver rpms from rpmfind since they weren't found in the Dag repository |
| [12:34:50] |
<openjack> |
now when I hit cgi-bin/trac.cgi I get an error saying clearsilver not installed |
| [12:34:55] |
<openjack> |
but it clearly is installed |
| [12:35:11] |
<openjack> |
any ideas? |
| [12:47:23] |
* |
malikeye has joined #trac |
| [12:54:34] |
* |
hotte has quit IRC |
| [12:59:55] |
* |
joost_ has quit IRC |
| [13:06:50] |
<openjack> |
TracError: ClearSilver not installed (/usr/lib/python2.3/site-packages/neo_cgi.so: cannot open shared object file: No such file or directory) |
| [13:06:55] |
<openjack> |
but that file does exist |
| [13:07:11] |
<openjack> |
what else should I do? |
| [13:25:28] |
<coderanger> |
check selinux |
| [13:29:16] |
* |
francois-simond has quit IRC |
| [13:42:59] |
<tag-> |
I love how the fact that the javascript console in mozilla will say '[Uncaught Exception "Error: ...whatever happened"]' |
| [13:43:18] |
<tag-> |
and more than that, I love how this has created a new term in my QA department, "Exception error" |
| [13:43:36] |
<tag-> |
"We keep getting exception errors when we do X" |
| [13:45:26] |
* |
Morth has joined #trac |
| [13:45:33] |
<Morth> |
hi guys |
| [13:45:38] |
<Morth> |
I have a problem |
| [13:46:06] |
<Morth> |
Im running a L2j server and im almost done setting it up but im stumped on this next part |
| [13:46:49] |
<Morth> |
anyone here? |
| [13:47:08] |
* |
Morth has left #trac |
| [13:47:44] |
<wardi> |
it's ok, we can come out now. Morth is gone |
| [13:52:04] |
* |
stevegt has joined #trac |
| [14:00:51] |
* |
m_g has quit IRC |
| [14:13:27] |
* |
whitelynx has joined #trac |
| [14:39:54] |
<openjack> |
I'm still having problems with ClearSilver, I checked that selinux is turned off, but I'm still getting the same error |
| [14:40:02] |
* |
stevegt has quit IRC |
| [14:40:04] |
<openjack> |
I've reinstalled the rpms |
| [14:42:44] |
* |
LionsMane has quit IRC |
| [14:43:40] |
<PlasmaHH> |
coderanger: unfortunately the needed syntax is not described on http://trac.edgewall.org/wiki/WikiFormatting though it promises to do so. |
| [14:44:17] |
<coderanger> |
PlasmaHH: Its on WikiMacros |
| [14:44:38] |
<coderanger> |
openjack: Chroot? |
| [14:44:59] |
<PlasmaHH> |
then the "see below" that promises to explain the image macro is a bit misleading... |
| [14:46:57] |
<coderanger> |
fix it ;-) |
| [14:47:21] |
<openjack> |
not chrooted |
| [14:47:47] |
<PlasmaHH> |
:P I assume there is no way to add captions below an image...? |
| [14:49:19] |
<coderanger> |
PlasmaHH: Just put text below it? |
| [14:49:44] |
<coderanger> |
openjack: Did you compile anything from source, or use any RPMs that were compiled for your system? |
| [14:49:52] |
<coderanger> |
er, werent |
| [14:50:03] |
<PlasmaHH> |
coderanger: I would like to not have the description in the code flow. more like (sorry to mention it) mediawiki does it, with smaller text that clearly belongs to the image. |
| [14:50:19] |
<openjack> |
I didn't compile anything from source, any additions were done using yum |
| [14:50:25] |
<openjack> |
thats why I'm confused |
| [14:50:35] |
<coderanger> |
PlasmaHH: Either use HTML then, or make your own macro |
| [14:51:18] |
<PlasmaHH> |
.o0O( wasnt the wiki there so one does not need html? ) |
| [14:55:22] |
<coderanger> |
openjack: You said earlier you got some RPMs from elsewhere though |
| [14:55:32] |
<coderanger> |
PlasmaHH: Doesn't mediawiki use a macro though? |
| [14:56:59] |
<PlasmaHH> |
coderanger: well, you can call it macro too, dunno how they call it but it has a similar syntax. no () but | for parameters and caption etc. |
| [14:57:26] |
<coderanger> |
PlasmaHH: So write a macro to do that. Someone always has to do it first ;-) |
| [14:58:03] |
<PlasmaHH> |
coderanger: usually the people searching for such functionality dont have the time, since they search for functionality because they have to get something done, like documentation with lots of images ;) |
| [14:58:36] |
<coderanger> |
PlasmaHH: Such is the danger of using software as young as Trac (compared to MediaWiki). |
| [15:00:13] |
<openjack> |
coderanger: ya, at first I downloaded the rpms, then I used yum |
| [15:00:38] |
<PlasmaHH> |
well, I see the wiki in trac as an additional feature, while for mediawiki its the core functionality. and as such for what trac combines (which is its core functionality) there are lots of things making the individual parts better, but none that seem to combine it the way it is. (although I would like to have a way to replace tracs wiki by mediawiki, since using tracs wiki only for one project) |
| [15:00:43] |
<coderanger> |
openjack: Okay, and were those RPMs for your actual distro or for something else (including another verison) |
| [15:00:49] |
<openjack> |
dag |
| [15:01:15] |
<openjack> |
coderanger: trac, clearsilver, and python-clearsilver all came from the dag wieers repo |
| [15:01:28] |
<openjack> |
I have several other packages from that repo as well |
| [15:02:20] |
<openjack> |
I'm using CentOS, the dag repo is the only additional repo that is setup |
| [15:03:23] |
<coderanger> |
PlasmaHH: The way something like mediawiki gets so much functionality like that is generally just being around and people writing all the little bits they need. Trac has a good start (trac-hacks), but it can't compare in terms of man-hours spent on things like that. |
| [15:03:40] |
<thm> |
openjack: does 'ldd /usr/lib/python2.3/site-packages/neo_cgi.so' work? |
| [15:04:03] |
<openjack> |
thm: linux-gate.so.1 => (0xffffe000) |
| [15:04:03] |
<openjack> |
libz.so.1 => /usr/lib/libz.so.1 (0xf7fb7000) |
| [15:04:03] |
<openjack> |
libpthread.so.0 => /lib/tls/libpthread.so.0 (0xf7fa5000) |
| [15:04:03] |
<openjack> |
libc.so.6 => /lib/tls/libc.so.6 (0xf7e7a000) |
| [15:04:03] |
<openjack> |
/lib/ld-linux.so.2 (0x56555000) |
| [15:04:14] |
<openjack> |
thm: does that mean it worked? |
| [15:04:19] |
<thm> |
yep |
| [15:04:23] |
<thm> |
just an idea |
| [15:05:31] |
<PlasmaHH> |
coderanger: maybe its just me thinking like that, but I would have thought that whoever wrote the image macro would have thought of having captions. I mean there are things like size, html attributes, different locations etc. maybe its only me from a scientific background being used to have no image without dedicated captions |
| [15:07:50] |
<coderanger> |
PlasmaHH: Like I said, someone has to be the first to want something like that for everything, congrats :) |
| [15:08:27] |
<coderanger> |
The Image macro is pretty simple, and correlates directly to an <img> tag |
| [15:08:35] |
<PlasmaHH> |
it just that I would have thought that whoever uses the image macro first would want it ;) ( ok, I could be the first to use the image macro, but I somehow doubt it) |
| [15:08:51] |
* |
thm has quit IRC |
| [15:09:00] |
<coderanger> |
PlasmaHH: Heh, I've never needed captions, but then I use very few images |
| [15:09:01] |
<PlasmaHH> |
the direct correlation might be the problem then, since a caption would need a little bit of div magic around it |
| [15:09:19] |
<PlasmaHH> |
images are great to illustrate things ;) |
| [15:09:28] |
<coderanger> |
PlasmaHH: Well you could generalize it somewhat |
| [15:09:48] |
<coderanger> |
PlasmaHH: make a processor that adds a caption to whatever wiki text it encloses |
| [15:10:24] |
<PlasmaHH> |
I really might do this one day, but not now as I have too much work to do, and since I prefer mediawiki ;) |
| [15:10:36] |
* |
pisi has joined #trac |
| [15:11:02] |
<pisi> |
Hi. with genshi templates, what is the equivalent of site_footer.cs ? |
| [15:11:55] |
<PlasmaHH> |
but I have to go to bed now, need to be at work in 8 hours |
| [15:13:20] |
<coderanger> |
pisi: site.html |
| [15:15:02] |
<pisi> |
coderanger: in <project>/templates ? |
| [15:15:13] |
<coderanger> |
pisi: You need to position your elements with match blocks though |
| [15:15:15] |
<pisi> |
coderanger: I was thinking about editing layout.html directly |
| [15:15:40] |
<pisi> |
coderanger: I have google analytics code to paste somewhere... and looking for that somewhere now.. |
| [15:16:09] |
<coderanger> |
use site.html and a py:match block |
| [15:17:22] |
<pisi> |
coderanger: could you elaborate ? |
| [15:17:38] |
<coderanger> |
pisi: Same idea as the way layout.html works |
| [15:17:57] |
<coderanger> |
You may have to read up on XPath |
| [15:18:56] |
<pisi> |
py:match html/body ? |
| [15:19:21] |
<coderanger> |
if you just want to shove it at the end |
| [15:21:25] |
<pisi> |
I think i have to read up on genshi instead... |
| [15:22:04] |
<coderanger> |
that too |
| [15:24:20] |
<pisi> |
ok, got it, thanks. ${select('*|text()')} was the magic |
| [15:27:46] |
* |
pisi has quit IRC |
| [15:33:41] |
<openjack> |
I found some hope on the clearsilver issue, my python is 64bit and my clearsilver (neo_cgi) is 32 bit |
| [15:34:55] |
* |
cookin__ has quit IRC |
| [15:40:41] |
* |
_shawn has quit IRC |
| [15:50:51] |
* |
lnayrb has quit IRC |
| [15:52:49] |
* |
cookin_ has joined #trac |
| [16:10:57] |
* |
_shawn has joined #trac |
| [16:22:12] |
* |
boorad has quit IRC |
| [16:23:09] |
* |
boorad has joined #trac |
| [16:23:59] |
* |
oversize has quit IRC |
| [16:24:51] |
* |
cookin_ has quit IRC |
| [16:30:02] |
* |
bnovc has quit IRC |
| [16:32:29] |
* |
openjack has left #trac |
| [16:36:08] |
* |
boorad_ has joined #trac |
| [16:38:11] |
* |
onny2 has joined #trac |
| [16:42:50] |
* |
boorad__ has joined #trac |
| [16:46:11] |
* |
onny has quit IRC |
| [16:46:55] |
* |
onny2 has quit IRC |
| [16:47:07] |
* |
onny has joined #trac |
| [16:51:26] |
* |
boorad has quit IRC |
| [16:53:02] |
* |
The_Tick has joined #trac |
| [16:53:19] |
<The_Tick> |
so I'm trying to find documentation on moving from sqlite to postgresql |
| [16:53:22] |
<The_Tick> |
any suggestions? |
| [16:56:39] |
<pygi> |
there is a script to do so in : |
| [16:56:41] |
<pygi> |
@hacks |
| [16:56:41] |
<evil_twin> |
pygi: "hacks" is http://trac-hacks.org <-- Plugins, Macros, etc. |
| [16:57:52] |
<The_Tick> |
ya, but I can't find any docs on the recommended version of postgresql |
| [16:57:54] |
<The_Tick> |
things like that |
| [16:58:09] |
<pygi> |
I guess newest :P |
| [16:58:16] |
<The_Tick> |
heh |
| [16:58:22] |
<The_Tick> |
I suppose that's what I'll use |
| [16:59:04] |
* |
boorad_ has quit IRC |
| [17:05:58] |
<Eridius> |
The_Tick: there's a plugin that lets you move data through Trac |
| [17:06:09] |
<Eridius> |
I was hacking on it to add the capability to move more pieces of data |
| [17:06:14] |
* |
Eridius still needs to hack in Attachment support |
| [17:06:52] |
<Eridius> |
http://trac-hacks.org/wiki/DatamoverPlugin |
| [17:07:11] |
<Eridius> |
granted, current version doesn't have my hacks yet |
| [17:07:22] |
* |
Eridius needs to do attachments, then submit a diff |
| [17:08:33] |
<The_Tick> |
Eridius: we're going to do an in place change of databases |
| [17:08:45] |
<The_Tick> |
not move between 2 instances |
| [17:08:52] |
<The_Tick> |
if I'm reading this right the latter is what this does |
| [17:12:28] |
<Eridius> |
true, but you can create a dummy second environment with the new database |
| [17:12:31] |
<Eridius> |
and then migrate data to it |
| [17:13:07] |
<Eridius> |
I'm migrating typosphere to dreamhost, and in my tests I have problems with sqlite database locking. so I'm working on this in order to migrate to MySQL (DH doesn't provide Postgres :/ ), so I'm doing the dummy environment approach |
| [17:13:10] |
<The_Tick> |
or just do a hotbackup |
| [17:13:29] |
<Eridius> |
a dummy env with no associated repo is pretty durn easy to do |
| [17:13:35] |
<The_Tick> |
you can't install postgres? |
| [17:13:48] |
<Eridius> |
somehow I doubt dreamhost would like me running a database server |
| [17:13:55] |
<pygi> |
:-P |
| [17:14:00] |
<The_Tick> |
could contact them first |
| [17:14:39] |
<Eridius> |
I'm pretty sure they'd say no. they don't like long-running processes |
| [17:15:04] |
<pygi> |
you can't run servers on shared hostings |
| [17:38:21] |
* |
dgynn has quit IRC |
| [17:39:01] |
<Eridius> |
hrm... why does Attachment.select() require parent_type and parent_id params? |
| [17:39:14] |
<Eridius> |
it produces a generator... but it's only ever going to give a single object |
| [17:39:38] |
<coderanger> |
You can have more than one attachment per wiki page/ticket |
| [17:40:05] |
<Eridius> |
ooh, that's the id of the ticket? |
| [17:40:13] |
<Eridius> |
I thought that was a attachment id |
| [17:40:14] |
<coderanger> |
or wiki page |
| [17:40:23] |
<coderanger> |
and the type is 'wiki' or 'ticket' |
| [17:40:31] |
<Eridius> |
ok, so attachments don't have their own id |
| [17:40:38] |
<Eridius> |
I figured the 'id' field meant, uh, attachment id ;) |
| [17:40:44] |
<Eridius> |
but I see it's uniqued on the type/id/filename combo |
| [17:40:44] |
<coderanger> |
they do, thats what the select returns |
| [17:40:51] |
<Eridius> |
no, that's a ticket/wiki id |
| [17:41:09] |
* |
Eridius is trying to figure out the proper way to get a list of all attachments, since Attachment.select() won't do it for me |
| [17:41:22] |
<Eridius> |
I mean, I can get all rows, but I want to get Attachment objects |
| [17:41:30] |
<Eridius> |
I suppose I can select filename as well and just use the constructor |
| [17:41:46] |
<coderanger> |
Attachment.select returns all attachments on the given parent object |
| [17:41:55] |
<Eridius> |
right, but I want all Attachments. period |
| [17:42:17] |
<coderanger> |
then just iterate it |
| [17:42:30] |
<coderanger> |
use WikiSystem.get_pages to get all wiki page names |
| [17:42:33] |
<Eridius> |
the point is, I have to do my own query, and build Attachment object |
| [17:42:39] |
<coderanger> |
No, you don't |
| [17:42:53] |
<coderanger> |
Use attachment.select on each wiki page and ticket |
| [17:42:53] |
<Eridius> |
ok, doing a separate query for every page/ticket in the database would be pretty inefficient |
| [17:43:00] |
<Eridius> |
I don't really want to do 1000 queries just to get a list of attachments |
| [17:43:10] |
<Eridius> |
and THEN do queries for each attachment |
| [17:43:16] |
<Eridius> |
(to fetch the attachment data) |
| [17:43:35] |
<coderanger> |
Its an admin task |
| [17:43:39] |
<coderanger> |
doesn't need to be fast |
| [17:43:52] |
<Eridius> |
I'd prefer my browser not to time out, thank you |
| [17:44:04] |
<coderanger> |
how many attachments do you have? |
| [17:44:22] |
<Eridius> |
693 |
| [17:44:30] |
<Eridius> |
1082 tickets |
| [17:44:35] |
<coderanger> |
the major time issue is likely to be copying the files anyway |
| [17:44:56] |
<Eridius> |
true... but I'd rather not be deliberately inefficient |
| [17:45:05] |
* |
Eridius ponders hardlinking files again |
| [17:45:12] |
<Eridius> |
that would be pretty fast, and attachments really *shouldn't* be modified on-disk |
| [17:45:27] |
<coderanger> |
this has to run on windows |
| [17:45:29] |
<coderanger> |
so no |
| [17:45:35] |
<coderanger> |
:P |
| [17:45:36] |
<Eridius> |
BAH! |
| [17:46:00] |
<Eridius> |
so what does os.link do on windows? |
| [17:46:15] |
<coderanger> |
writting Trac code is fun, unix/windows, 2.3/2.4/2.5, sqlite/postgres/mysql, cgi/fcgi/mod_python/tracd |
| [17:46:24] |
<Eridius> |
mmm... |
| [17:46:36] |
<coderanger> |
:) |
| [17:54:39] |
* |
mcella has joined #trac |
| [18:09:55] |
* |
mcella has quit IRC |
| [18:25:32] |
* |
malikeye has quit IRC |
| [18:26:55] |
<Eridius> |
what's the fastest way of turning a generator into an array? |
| [18:28:56] |
* |
malikeye has joined #trac |
| [18:35:28] |
<coderanger> |
list() |
| [18:35:44] |
<Eridius> |
thanks |
| [18:35:53] |
<coderanger> |
unless you are doing this from a cursor |
| [18:35:54] |
* |
Eridius had been using [c for c in theGenerator] since I didn't know |
| [18:36:05] |
<Eridius> |
how do you do it from a cursor? |
| [18:36:07] |
<coderanger> |
in which case the answer is .fetchall() |
| [18:36:12] |
<Eridius> |
(in this case I was using WikiSystem(env).get_pages() |
| [18:36:12] |
<Eridius> |
) |
| [18:36:28] |
<coderanger> |
ahh |
| [18:36:34] |
<coderanger> |
try to avoid doing it though |
| [18:36:43] |
<coderanger> |
as the memory load can add up quickly |
| [18:37:14] |
<Eridius> |
oh? |
| [18:37:26] |
<Eridius> |
I wanted to provide a list of wiki pages for you to migrate attachments from |
| [18:38:09] |
<coderanger> |
yeah, in CS the only option is to listify it |
| [18:38:19] |
<coderanger> |
in Genshi I think you can just pass the generator right through |
| [18:58:40] |
<Eridius> |
coderanger: any recommendations for best way to copy files? |
| [18:58:46] |
<Eridius> |
or any python libraries that provide this functionality |
| [18:58:54] |
<alect> |
shutils |
| [18:59:03] |
<alect> |
-s |
| [18:59:20] |
<Eridius> |
shutil.copyfile()? |
| [19:01:45] |
<Eridius> |
does that create intermediate directories? |
| [19:03:14] |
<Eridius> |
I guess not |
| [19:09:59] |
<coderanger> |
os.makdirs will I think |
| [19:10:16] |
<Eridius> |
yeah, I found it, thanks |
| [19:20:56] |
* |
stevegt has joined #trac |
| [19:21:13] |
<Eridius> |
coderanger: my attachments code appears to be working |
| [19:21:17] |
* |
Eridius is now testing copying *all* attachments |
| [19:21:26] |
<coderanger> |
coolness :) |
| [19:21:46] |
<Eridius> |
hehe, taking a while |
| [19:21:57] |
<Eridius> |
ah, done |
| [19:22:16] |
<Eridius> |
well, it appears to have worked |
| [19:22:53] |
<Eridius> |
woot |
| [19:23:22] |
<Eridius> |
hrm... ideally I'd only show wiki pages that have attachments |
| [19:30:02] |
<Eridius> |
what's the best way to remove duplicates from a list? |
| [19:30:16] |
<Eridius> |
list(set(thelist)) |
| [19:30:16] |
<Eridius> |
? |
| [19:30:21] |
<alect> |
yep |
| [19:31:22] |
<coderanger> |
remember to do the 2.3 armoring for sets |
| [19:31:25] |
<Eridius> |
? |
| [19:31:41] |
<Eridius> |
coderanger: I haven't touched python in years, aside from this hacking here. I don't know what you're talking about |
| [19:31:52] |
<coderanger> |
try: \n set = set \n except NameError: \n from sets import Set as set |
| [19:32:00] |
<Eridius> |
I see |
| [ |
