| [00:06:03] |
* |
converter has quit IRC |
| [00:15:29] |
* |
cmlenz has joined #trac |
| [00:16:07] |
* |
prologic_ has joined #trac |
| [00:18:47] |
* |
prologic_ has quit IRC |
| [00:19:50] |
* |
prologic_ has joined #trac |
| [00:26:06] |
<prologic_> |
hey guys, any suggestions on a very tightly secured trac instnace ? |
| [00:29:24] |
* |
jrydberg has joined #trac |
| [00:31:41] |
* |
prologic has quit IRC |
| [00:32:52] |
* |
prologic_ is now known as prologic |
| [01:23:22] |
* |
asmodai pokes cmlenz |
| [01:23:35] |
<asmodai> |
cmlenz: .10-dev sets new tickets on prio highest and severity blocker by default? |
| [01:23:50] |
<cmlenz> |
that'd be a bug |
| [01:24:24] |
<cmlenz> |
.10-dev doesn't have severity by default |
| [01:24:38] |
* |
jrydberg has quit IRC |
| [01:24:42] |
<cmlenz> |
not seeing that here |
| [01:24:53] |
* |
asmodai blinks |
| [01:25:22] |
<asmodai> |
I nuked site-packages/trac, share/trac before installing, anything I've missed? (did upgrade and wiki upgrade) |
| [01:26:17] |
<otaku42> |
prologic: what kind of security are you thinking of? |
| [01:28:49] |
* |
Sonderblade has joined #trac |
| [01:29:09] |
* |
tuxipuxi has joined #trac |
| [01:35:35] |
* |
wbb4_ has quit IRC |
| [01:36:03] |
* |
wbb4 has joined #trac |
| [01:37:22] |
<vorsprung> |
prologic, security on systems is all to do with the three As |
| [01:37:36] |
<vorsprung> |
prologic, access, authentication and audit |
| [01:38:49] |
<vorsprung> |
prologic, so lock down the access to the server to restrict it. Perhaps it could be on an intranet that is secure. Or a vpn. Or if it's on the public internet, maybe the firewall blocks all IPs except a few |
| [01:39:39] |
<vorsprung> |
prologic, authentication- look at using apache SSL certificates. If you already have a PKI then look at using that with apache or trac |
| [01:40:38] |
<vorsprung> |
prologic, audit- any ticket tracking system, including trac is good at audit trials. Think about additional audit controls like apache logging of incoming requests |
| [01:48:07] |
<asmodai> |
cmlenz: see http://trac.tendra.org/newticket |
| [01:52:15] |
<cmlenz> |
any defaults in trac.ini? |
| [01:52:24] |
<cmlenz> |
default_priority/default_severity? |
| [01:52:41] |
<cmlenz> |
also, if this *is* a default install, why do you have a severity dropdown at all? |
| [01:53:06] |
<cmlenz> |
ah no, the environment is older |
| [01:53:57] |
<asmodai> |
It's an environment from 0.8 IIRC |
| [01:54:02] |
<asmodai> |
I kept following trunk :) |
| [01:54:26] |
<cmlenz> |
well, check the default values in trac.ini... apparently they don't match any item in the two dropdowns |
| [01:54:42] |
* |
maxb_ has joined #trac |
| [01:54:44] |
<asmodai> |
ok, priority I see: |
| [01:54:45] |
<asmodai> |
trac.ini:default_priority = major |
| [01:55:09] |
<asmodai> |
Nothing for severity though. |
| [01:55:24] |
* |
maxb_ is now known as maxb |
| [01:56:49] |
* |
asmodai adds a default for severity, just funny it defaults to either the highest in the list or the first alphabetically |
| [02:05:34] |
<Getty> |
i also so that the sorting is not always used |
| [02:05:46] |
<Getty> |
i sorted my project parts but they are still sorted alphabetical |
| [02:15:00] |
* |
wbb4_ has joined #trac |
| [02:15:00] |
* |
wbb4 has quit IRC |
| [02:19:03] |
<cmlenz> |
asmodai: the browser does that, it's simply how HTML <select>s work |
| [02:19:09] |
<cmlenz> |
trac doesn't preselect any option |
| [02:19:17] |
<cmlenz> |
except if there's a match default_ option in trac.ini |
| [02:47:56] |
<asmodai> |
cmlenz: gotcha |
| [02:48:35] |
* |
PingYeh has joined #trac |
| [03:06:54] |
* |
_idostyle has joined #trac |
| [03:24:06] |
* |
idostyle has quit IRC |
| [03:24:08] |
* |
_idostyle is now known as idostyle |
| [03:25:18] |
* |
diacono has joined #trac |
| [03:26:52] |
<diacono> |
well I find myself scratching my head trying to install trac |
| [03:27:06] |
<diacono> |
any gentoo users here? |
| [03:27:47] |
* |
jerome_ has quit IRC |
| [03:28:17] |
<diacono> |
is there a prefered way of getting trac up and running? |
| [03:29:13] |
<cmlenz> |
http://trac.edgewall.org/wiki/TracOnGentoo |
| [03:29:18] |
<cmlenz> |
not sure how up to date that is |
| [03:29:36] |
<diacono> |
I've seen that |
| [03:29:48] |
<vorsprung> |
diacono, if I can make it work on MS Windows 2003 you can make it work on a nice linux |
| [03:30:33] |
<diacono> |
but that article is a manual build... |
| [03:31:50] |
<diacono> |
let me digg out how far I got |
| [03:35:53] |
<diacono> |
I've followed http://trac.edgewall.org/wiki/TracModPython |
| [03:36:10] |
<diacono> |
but instead of getting trac, I get apacher directory browsing |
| [03:36:30] |
<diacono> |
I'm running this as a vhost |
| [03:36:55] |
<otaku42> |
diacono: just an idea: did you restart apache after you changed configuration? |
| [03:37:05] |
<diacono> |
yes |
| [03:37:39] |
<otaku42> |
diacono: is mod_python enabled/loaded? is it configured as handler for the corresponding vhost? |
| [03:38:00] |
<otaku42> |
s/it/mod_python+trac/ |
| [03:38:07] |
<diacono> |
it loaded from httpd.conf |
| [03:38:16] |
<diacono> |
LoadModule python_module modules/mod_python.so |
| [03:39:42] |
* |
davix has joined #trac |
| [03:42:17] |
* |
diacono has quit IRC |
| [03:42:20] |
* |
diacono_ has joined #trac |
| [03:44:35] |
<diacono_> |
the handler is there |
| [03:47:06] |
* |
jerome has joined #trac |
| [03:48:08] |
<diacono_> |
here is the vhosts: http://www.copypot.com/103 |
| [03:50:54] |
<diacono_> |
actually the handler was set to / instead of /project1 |
| [03:51:25] |
<diacono_> |
however once I change it to /project1 apache does not respond |
| [03:52:33] |
<diacono_> |
I can see mod_python in the error log as I have set it to debug: |
| [03:52:34] |
<diacono_> |
[notice] mod_python: Creating 32 session mutexes based on 150 max processes and 0 max threads |
| [03:53:40] |
<otaku42> |
diacono_: no idea, the configuration i use for my trac installation is the same (apart from the path to the trac environment) |
| [03:55:16] |
<diacono_> |
in fact, apache wont even respond on the root directory either |
| [04:08:39] |
* |
davix has left #trac |
| [04:52:15] |
* |
whitelynx has quit IRC |
| [05:08:50] |
* |
whitelynx has joined #trac |
| [05:09:33] |
* |
chjunior has joined #trac |
| [05:31:51] |
* |
mjg2 has quit IRC |
| [05:35:12] |
* |
bijan has joined #trac |
| [05:45:12] |
<prologic> |
vorsprung, thanks for your suggestions |
| [05:45:15] |
<prologic> |
sorry was out for a while |
| [05:45:43] |
<prologic> |
I'm looking to a. secure it down b. allow (in futore) access to certain people perhaps via vpn and auth |
| [05:45:48] |
<prologic> |
and of course it will be audited :) |
| [05:50:46] |
* |
wbb4_ is now known as wbb4 |
| [05:51:46] |
* |
milk-it has joined #trac |
| [05:54:21] |
* |
Champion_du_Mond has joined #trac |
| [06:03:57] |
* |
mjg2 has joined #trac |
| [06:17:59] |
* |
coderanger_ has joined #trac |
| [06:19:47] |
* |
whitelynx|firest has joined #trac |
| [07:02:13] |
* |
converter has joined #trac |
| [07:07:08] |
* |
_idostyle has joined #trac |
| [07:22:43] |
* |
idostyle has quit IRC |
| [07:22:54] |
* |
_idostyle is now known as idostyle |
| [07:23:40] |
* |
danbeck has joined #trac |
| [07:28:49] |
* |
twiggy has joined #trac |
| [07:30:05] |
<twiggy> |
Hey I'm at a loss here, I've followed the wiki page for plugins, moved the eggs into the plugin dir for the project, ran easy install on the eggs, checked permissions, enabled them in the ini file but I still can't seem to get any plugins working |
| [07:30:23] |
<misc> |
evil_twin: @debug |
| [07:30:23] |
<evil_twin> |
misc: Error: "@debug" is not a valid command. |
| [07:30:26] |
<misc> |
evil_twin: debug |
| [07:30:26] |
<evil_twin> |
misc: Error: "debug" is not a valid command. |
| [07:30:30] |
<misc> |
evil_twin: faq debug |
| [07:30:30] |
<evil_twin> |
misc: Error: No factoid matches that key. |
| [07:30:35] |
<misc> |
groumph |
| [07:30:38] |
<misc> |
evil_twin: faq log |
| [07:30:39] |
<evil_twin> |
misc: Error: No factoid matches that key. |
| [07:30:56] |
<misc> |
evil_twin: come on, i am sure you you know what i am talking about ! |
| [07:30:56] |
<evil_twin> |
misc: Error: "come" is not a valid command. |
| [07:31:04] |
<misc> |
twiggy: ok, nothing in log ? |
| [07:31:17] |
<twiggy> |
misc: lemme check |
| [07:31:48] |
<twiggy> |
nothing in my log dir |
| [07:31:53] |
<twiggy> |
unless that's the wrong place |
| [07:32:15] |
<misc> |
are they enabled ? |
| [07:32:22] |
* |
klasstek has joined #trac |
| [07:32:22] |
<misc> |
by default, they are not |
| [07:32:57] |
<twiggy> |
misc: aye, in my ini file I have (for example) webadmin.* = enabled under components |
| [07:33:12] |
<twiggy> |
I can disable and enable builtin plugins fine, but 3rd party won't go |
| [07:33:24] |
<misc> |
twiggy: well, the log may help |
| [07:33:56] |
<twiggy> |
logging is enabled but I have no log file :o/ |
| [07:34:24] |
<misc> |
permission problem, maybe ? |
| [07:34:32] |
<misc> |
evil_twin: logging ? |
| [07:34:32] |
<evil_twin> |
misc: Error: "logging" is not a valid command. |
| [07:34:57] |
<twiggy> |
I'll check |
| [07:37:49] |
<twiggy> |
my log file is set to log/trac.log, type is file, level is debug, correct right? |
| [07:39:29] |
<twiggy> |
hmm this is interesting, now when I click any links it sends me to a blank page, this may be a problem with my build I'll go double check that |
| [07:45:34] |
<diacono_> |
BINGO! got it up and running... |
| [07:45:46] |
<diacono_> |
no idea what fixed it |
| [07:46:19] |
<diacono_> |
but I basically downloaded buildix and stole the apache conf files |
| [07:46:35] |
* |
agile has quit IRC |
| [07:46:44] |
<diacono_> |
now, my css on trac is not working |
| [07:46:54] |
<diacono_> |
any ideas why? |
| [07:47:05] |
* |
bijan has quit IRC |
| [07:48:28] |
<diacono_> |
in fact, nor is javascript and the trac logo |
| [07:49:45] |
<twiggy> |
misc: found it, egg cache is currently in /root/.python-eggs, I installed them as root, how would I change that? |
| [07:49:47] |
* |
otaku42 is now known as otaku42_away |
| [07:50:25] |
<misc> |
twiggy: there is a ven variable , i think |
| [07:50:27] |
<misc> |
env |
| [07:50:37] |
<twiggy> |
there we go, yep, thankya |
| [07:51:51] |
<misc> |
i thin PYTHON_EGG_DIR, or something like this, but i do not know everything :) |
| [07:54:42] |
<twiggy> |
that was it :oD |
| [07:58:08] |
* |
l0wkey has joined #trac |
| [08:03:52] |
<diacono_> |
so has anyone else had the css issue? |
| [08:08:23] |
* |
chjunior has quit IRC |
| [08:12:15] |
* |
coderanger1 has joined #trac |
| [08:13:59] |
* |
converter has quit IRC |
| [08:20:29] |
* |
agile has joined #trac |
| [08:26:01] |
* |
coderanger_ has quit IRC |
| [08:28:02] |
* |
olafnils is now known as roo9 |
| [08:30:01] |
* |
chandlerc has quit IRC |
| [08:30:37] |
* |
xjjk has joined #trac |
| [08:35:04] |
* |
converter has joined #trac |
| [08:39:14] |
* |
stretch has joined #trac |
| [08:44:01] |
* |
diacono_ has quit IRC |
| [08:58:19] |
* |
coderanger1 has quit IRC |
| [09:02:50] |
* |
coderanger_ has joined #trac |
| [09:06:24] |
* |
chandlerc has joined #trac |
| [09:25:23] |
<pacopablo> |
cmlenz: around? |
| [09:25:30] |
<cmlenz> |
yup |
| [09:25:44] |
<pacopablo> |
got a couple questions about markup :) |
| [09:26:24] |
<cmlenz> |
sure |
| [09:27:56] |
<pacopablo> |
is it possible to extend the XInclude support to be able to call clearsilver templates? |
| [09:28:08] |
<pacopablo> |
similar to how tracistan can? |
| [09:28:21] |
* |
pacopablo is looking for relevant code |
| [09:28:27] |
<cmlenz> |
I guess it would but not sure I see the point ;-) |
| [09:29:21] |
<coderanger_> |
It would ease migrations :P |
| [09:29:21] |
<pacopablo> |
ideally, so that it could be extended in oter manners asa well so that one could include other types of templates |
| [09:29:25] |
<pacopablo> |
that too |
| [09:29:46] |
<pacopablo> |
and it would allow me to make a tracmarkup plugin :) |
| [09:29:59] |
<pacopablo> |
so that markup can be easily used before the migration |
| [09:31:09] |
<cmlenz> |
what I'd like is a wholesale migration including a compatibility layer for clearsilver, not trying some kind of mix-and-match template system |
| [09:31:32] |
<cmlenz> |
there's an example/trac directory in the markup repos that has the beginnings of a migration, including the compatibility layer |
| [09:32:17] |
<pacopablo> |
I looked at that, but admittedly couldn't see many of the changes |
| [09:32:37] |
<pacopablo> |
what does the compatibility layer look like? |
| [09:33:01] |
<pacopablo> |
and is template proccessing going to remain mostly the same? |
| [09:33:17] |
<pacopablo> |
as in, is the return value of process_request goign to change? |
| [09:33:37] |
<cmlenz> |
well, first it's page based, so that's the granularity for migration |
| [09:33:45] |
<cmlenz> |
it changes, yeah |
| [09:34:04] |
<cmlenz> |
includes the data dict |
| [09:34:31] |
<cmlenz> |
http://markup.edgewall.org/browser/trunk/examples/trac/trac/web/main.py#L199 |
| [09:35:01] |
<cmlenz> |
http://markup.edgewall.org/browser/trunk/examples/trac/trac/web/chrome.py#L237 |
| [09:38:20] |
* |
converter has quit IRC |
| [09:39:42] |
<pacopablo> |
so if there are two return values, then it uses clearsilver, and if there are three, it uses markup? |
| [09:40:23] |
<cmlenz> |
yup |
| [09:40:50] |
<cmlenz> |
that's all just an adhoc prototype of course |
| [09:41:28] |
* |
Sonderblade has quit IRC |
| [09:47:58] |
<skugg> |
what's the 'master' trac.ini file called on debian? I (want to) have a group of trac installs that inherit a master config and just tweak the bits they need |
| [09:48:30] |
<pacopablo> |
check /usr/lib/python2.x/site-packages/trac/__siteconfig__.py |
| [09:48:37] |
<pacopablo> |
but it's probably /usr/share/trac/conf |
| [09:48:46] |
<pacopablo> |
and trac.ini inside there |
| [09:49:27] |
<cmlenz> |
that's only with the current dev version though |
| [09:49:48] |
<cmlenz> |
or isn't it? |
| [09:49:54] |
<cmlenz> |
jeeze I've lost track :-P |
| [09:50:02] |
<cmlenz> |
it isn't |
| [09:50:04] |
<pacopablo> |
I think it's in 0.9 |
| [09:50:48] |
<cmlenz> |
yeah, I confused that with the global plugins dir |
| [09:50:53] |
<coderanger_> |
trunk added a site plugins folder |
| [09:52:02] |
* |
coderanger_ has quit IRC |
| [10:02:15] |
* |
xjjk has quit IRC |
| [10:02:16] |
<skugg> |
pacopablo: there doesn't seem to be such a file on my system. |
| [10:02:21] |
<alect> |
morning |
| [10:02:33] |
<pacopablo> |
skugg: you have to create it |
| [10:03:43] |
* |
xjjk has joined #trac |
| [10:05:15] |
<skugg> |
pacopablo: sorry, I meant the __siteconfig__.py one. it doesn't exist here. I'm using unreleased 0.10, by the way. |
| [10:05:35] |
<skugg> |
nm. i found the dir now. |
| [10:05:39] |
<pacopablo> |
ahh, k |
| [10:05:40] |
<skugg> |
(and you were right) |
| [10:06:35] |
<skugg> |
i was temporarily thrown off track by debian creating an empty /etc/trac directory, but creating a trac.ini or conf/trac.ini in there doesn't seem to do anything. |
| [10:07:10] |
<skugg> |
i don't want to put my custom stuff in /usr/share/trac though, as this will presumably be changed on upgrade. |
| [10:07:13] |
<skugg> |
:( |
| [10:07:35] |
<pacopablo> |
simply modify the location in the __siteconfig__.py script |
| [10:07:54] |
<pacopablo> |
though /usr/share/trac tends to be the default for source based installes |
| [10:09:59] |
* |
cdigs has joined #trac |
| [10:12:50] |
* |
Gruf has quit IRC |
| [10:13:30] |
* |
Gruf has joined #trac |
| [10:14:08] |
* |
matt_good has joined #trac |
| [10:15:15] |
* |
cdigs has quit IRC |
| [10:21:10] |
<pacopablo> |
any good way to include image maps into wiki pages? |
| [10:21:19] |
<pacopablo> |
{{{#!html ? |
| [10:22:21] |
<matt_good> |
pacopablo: cmlenz made one on markup.edgewall.org |
| [10:22:33] |
<matt_good> |
take a look at that |
| [10:22:43] |
<pacopablo> |
thanks |
| [10:23:05] |
* |
cmlenz is fighting with generating bytecode from template expressions in markup... what a PITA |
| [10:24:33] |
* |
pisi has quit IRC |
| [10:24:40] |
<cmlenz> |
anyone know a good guide to python bytecode? |
| [10:25:48] |
<alect> |
this for compiling markup? |
| [10:26:23] |
<cmlenz> |
for compiling individual expressions in markup |
| [10:26:31] |
<alect> |
yes |
| [10:27:22] |
* |
s0undt3c1 has joined #trac |
| [10:27:46] |
<cmlenz> |
or just how to find out how a function like this: http://markup.edgewall.org/browser/trunk/markup/eval.py#L262 compiles to bytecode |
| [10:28:03] |
<cmlenz> |
I'm not sure how to drill-down into dis results |
| [10:30:46] |
<alect> |
you could inspect the bytecode it generates and reverse engineer it |
| [10:31:05] |
<cmlenz> |
yeah, but *how* is the question :-P |
| [10:31:19] |
<alect> |
you can get the bytecode from code objects in the inspect module |
| [10:33:45] |
<matt_good> |
cmlenz: are you looking for a textual listing of the assemly or what? |
| [10:33:58] |
<cmlenz> |
yeah |
| [10:34:16] |
<matt_good> |
and dis.dis(function) doesn't give you that? |
| [10:35:32] |
<asmodai> |
Hey cmlenz, alect |
| [10:35:46] |
<cmlenz> |
that includes nested code object that I don't know how to get to |
| [10:36:58] |
<matt_good> |
cmlenz: this is what I get on Python 2.4 for the _visit_attribute code: http://rafb.net/paste/results/trwFo753.html |
| [10:37:15] |
<asmodai> |
cmlenz: http://www.python.org/pycon/2006/papers/45/ does this help? |
| [10:38:16] |
<cmlenz> |
asmodai: thanks, I'll have a look |
| [10:38:22] |
<cmlenz> |
matt_good: how did you get that exactly? |
| [10:38:56] |
<matt_good> |
well, I pasted the code for that function in the interpreter |
| [10:39:03] |
<matt_good> |
then import dis |
| [10:39:09] |
<matt_good> |
dis.dis(_visit_attribute) |
| [10:39:32] |
<cmlenz> |
thanks |
| [10:39:39] |
<matt_good> |
:) |
| [10:40:31] |
<cmlenz> |
I have the "loading a name from the context instead of from locals" working here, but still need to get interchangable getitem/getattr working |
| [10:40:32] |
<alect> |
looks like dis can disassemble code objects from inspect too |
| [10:41:13] |
<asmodai> |
cmlenz: http://holdenweb.blogspot.com/2006/02/implementation-of-python-bytecode.html |
| [10:41:37] |
<cmlenz> |
thx |
| [10:41:49] |
<asmodai> |
cmlenz: perhaps this helps too http://www.python.org/dev/peps/pep-0330/ |
| [10:42:01] |
<asmodai> |
http://docs.python.org/lib/bytecodes.html |
| [10:42:06] |
<alect> |
asmodai is our own personal web crawler |
| [10:42:10] |
<alect> |
:) |
| [10:42:16] |
<cmlenz> |
heh |
| [10:42:18] |
<matt_good> |
cmlenz: so the dotted notation isn't working w/ the AST? |
| [10:42:30] |
<asmodai> |
alect: *grin* For some reason I am quite able to navigate search engines getting the information I want/need. |
| [10:42:31] |
<cmlenz> |
it is |
| [10:42:41] |
<cmlenz> |
matt_good: but I want MORE SPEED :-P |
| [10:42:46] |
<matt_good> |
heh |
| [10:42:59] |
<matt_good> |
is it still slow on Python 2.5? |
| [10:43:04] |
<cmlenz> |
currently only kid is slower than markup |
| [10:43:10] |
<cmlenz> |
yeah, a bit |
| [10:43:15] |
<cmlenz> |
slower |
| [10:43:18] |
<cmlenz> |
not exactly slow |
| [10:43:31] |
* |
asmodai hands cmlenz parrot code ;) |
| [10:43:31] |
<cmlenz> |
even though the AST stuff is a lot faster |
| [10:45:17] |
<alect> |
perhaps it would be easier just to compose python code, and compile that? |
| [10:45:25] |
<alect> |
s/code/source |
| [10:45:26] |
<asmodai> |
cmlenz: http://www.wooji-juice.com/blog/python-globals-bytecode.html |
| [10:45:41] |
<cmlenz> |
alect: not sure what you mean |
| [10:46:18] |
<matt_good> |
I think compile("code") |
| [10:46:25] |
* |
s0undt3ch has quit IRC |
| [10:46:27] |
<alect> |
translate markup expressions into python source, compile() that |
| [10:46:33] |
* |
s0undt3c1 is now known as s0undt3ch |
| [10:47:13] |
<cmlenz> |
I think that'd be more work as I'll be doing the parsing |
| [10:47:24] |
<cmlenz> |
I'm building on compiler.pycodegen |
| [10:47:36] |
<cmlenz> |
only overriding those visitor methods where I want different behavior |
| [10:47:56] |
<asmodai> |
perhaps http://bytecodehacks.sourceforge.net/bch-docs/bch/index.html helps too |
| [10:48:20] |
* |
asmodai stops :) |
| [10:48:31] |
* |
alect hands asmodai a cookie |
| [10:49:09] |
<pacopablo> |
alect: think it will be easy to patch the current trunk with security? |
| [10:49:26] |
<cmlenz> |
asmodai: thanks a lot, nice bedtime reading material :-P |
| [10:49:44] |
<alect> |
paco: should be yeah |
| [10:50:14] |
<pacopablo> |
cmlenz: think security will get merged afte 0.10 is released? |
| [10:50:31] |
<cmlenz> |
that's the plan |
| [10:50:35] |
<pacopablo> |
good |
| [10:50:50] |
<pacopablo> |
now let's get 0.10 out so I don't have to wait any longer :) |
| [10:50:52] |
<cmlenz> |
but we've got a *lot* on the plate for "after 0.10 is released" :-P |
| [10:50:58] |
<alect> |
tracobjects (just the object bit) would make life easier re security |
| [10:51:12] |
<Getty> |
was tracobjects? |
| [10:52:04] |
<Getty> |
aehm what is :) |
| [10:52:54] |
* |
MACscr has joined #trac |
| [10:52:59] |
<matt_good> |
Getty: something cboos was working on for the class heirarchy of the Trac model classes |
| [10:53:04] |
<alect> |
check the TracObjects page on wiki |
| [10:53:26] |
<Getty> |
ah ok |
| [10:53:31] |
<MACscr> |
was "Download in Other Format" not available for trunks until v10? |
| [10:53:38] |
<Getty> |
do you btw work with an ORM inside of trac? or plain querywar? |
| [10:54:05] |
<pacopablo> |
plain query |
| [10:54:13] |
<matt_good> |
Getty: it's all straight SQL right now. I'm experimenting w/ integrating SQLAlchemy |
| [10:54:21] |
<pacopablo> |
though I think a move to sqlalchemy is slated for "after 0.10" :) |
| [10:54:35] |
<asmodai> |
cmlenz: Hope so |
| [10:55:03] |
<alect> |
sqlalchemy! |
| [10:55:06] |
<alect> |
+1 |
| [10:55:46] |
<cmlenz> |
bbl |
| [10:55:49] |
* |
cmlenz has quit IRC |
| [10:56:16] |
<Getty> |
matt_good: i dont now SQLalchemy, cause i'm coming from PHP side, but actually i started the next generation of an ORM of PHP that is very good. the python ORMs are all very different to that one. anyway, was just interested :-) |
| [10:58:33] |
<asmodai> |
alect: I was wondering about it |
| [10:58:43] |
<asmodai> |
alect: would it be a good move to use SQLAlchemy? |
| [10:58:50] |
<asmodai> |
alect: Have not messed with it yet, just readstuff |
| [10:59:15] |
<matt_good> |
asmodai: it's pretty nice |
| [10:59:31] |
<pacopablo> |
if we really want to go with an ORM, SQLAlchemy is definitely the one to go with |
| [10:59:35] |
<alect> |
matt is the expert |
| [10:59:37] |
<matt_good> |
it fits in well to replace the current DB layer |
| [10:59:44] |
<Getty> |
my i see some examples of use of sqlalchemy? |
| [10:59:49] |
<alect> |
but i think it's clear we need to ditch plain sql |
| [10:59:50] |
<asmodai> |
matt_good: So is it officially in the sights? |
| [10:59:51] |
<Getty> |
then i can compare it to the technology we got on php |
| [10:59:52] |
<pacopablo> |
sqlalchemy.org :) |
| [10:59:55] |
<Getty> |
thx |
| [11:00:09] |
<alect> |
and from my limted experience with the others, it wasn't that fun |
| [11:00:14] |
<Getty> |
some sqlalchemy guy who wanna see the other side? :-) |
| [11:00:17] |
<alect> |
and sqlalchemy was pretty cool |
| [11:00:20] |
<Getty> |
http://pear.php.net/DB_DataObject |
| [11:00:31] |
<matt_good> |
asmodai: well, it's officially in *my* sights, and barring some terrible disaster I would think it should be merged |
| [11:00:34] |
<alect> |
omg, blasphemer |
| [11:00:52] |
<asmodai> |
matt_good: :) |
| [11:01:13] |
<Getty> |
you must setup the tables in python? |
| [11:01:22] |
<Getty> |
or can sqlalchemy also autogenerate that stuff from DB? |
| [11:01:27] |
<pacopablo> |
Getty: no, sqlalchemy has inspection |
| [11:01:31] |
<Getty> |
ok good |
| [11:01:35] |
<Getty> |
was near your first negative point ;) |
| [11:01:39] |
<Getty> |
hrhr |
| [11:01:53] |
<pacopablo> |
but usually one sets them up in python, because sqlalchemy can create the database and table layout for you too |
| [11:02:19] |
<MACscr> |
can someone tell me how to enable "Download in other format" Download Zip, in trac? |
| [11:02:28] |
<matt_good> |
Getty: you can also use the "active mapper" extension which lets you just define the model classes and then it builds the tables for you |
| [11:02:40] |
<Getty> |
i see first negative point |
| [11:02:43] |
<Getty> |
it still gives back results |
| [11:02:53] |
<Getty> |
an ORM doesnt need the split of result and object for mapping |
| [11:03:05] |
<pacopablo> |
MACscr: I believe that download in zip is only in trunk atm |
| [11:03:12] |
<pacopablo> |
Getty: what do you mean? |
| [11:03:29] |
<MACscr> |
paco: thats all im looking for |
| [11:03:46] |
<pacopablo> |
Getty: you might be thinking of what active mapper does. |
| [11:03:46] |
<Getty> |
ah! another bad point.. defining of select totally complex... |
| [11:03:50] |
<Getty> |
no |
| [11:04:09] |
<Getty> |
i dont wanna pollute that channel, but if i'm allowed i give some example |
| [11:04:13] |
<Getty> |
how the flow is on DB_DataObject |
| [11:04:31] |
<Getty> |
(i put it in as less lines as possible ;) ) |
| [11:04:51] |
<alect> |
pastebin |
| [11:04:55] |
<alect> |
@pastebin |
| [11:04:55] |
<evil_twin> |
alect: "pastebin" is http://trac.pastebin.com |
| [11:05:05] |
<Getty> |
yeah ok or this.... but it fits always in one line anyway ;) |
| [11:05:29] |
<matt_good> |
Getty: you can avoid the lower-level db parts of sqlalchemy if you want, but they're provided for you if you need them |
| [11:05:29] |
* |
whitelynx has quit IRC |
| [11:06:28] |
<MACscr> |
Paco: should this Download Format option be on by default? |
| [11:06:37] |
<MACscr> |
and do you have any idea if its available in .95? |
| [11:07:21] |
<matt_good> |
MACscr: no, it will be in 0.10 |
| [11:07:24] |
<Getty> |
http://paste.debian.net/9049 |
| [11:07:35] |
<MACscr> |
ah man |
| [11:07:37] |
<MACscr> |
thanks though |
| [11:07:54] |
<Getty> |
matt_good: actually i would say its never needed if you lost the database in mind a bit, cause its just "out of view" if you think straight in your code |
| [11:08:11] |
<Getty> |
matt_good: of course you could probably take performance advantages, but if you want that, you dont use an ORM |
| [11:08:17] |
<pacopablo> |
Getty: does DB_DataObject support multi-column primary keys? |
| [11:08:26] |
<Getty> |
pacopablo: yeah |
| [11:08:27] |
<MACscr> |
Matt_good: is their a plugin I can install to get the functionaliy now? |
| [11:08:31] |
<Getty> |
why not? |
| [11:08:43] |
<Getty> |
but the concept gives no advantage for this |
| [11:08:51] |
* |
_idostyle has joined #trac |
| [11:08:51] |
<matt_good> |
MACscr: I don't think so |
| [11:08:56] |
<Getty> |
or on the other site: if you got one key or many, it doesnt care for the concept |
| [11:09:20] |
<pacopablo> |
Getty: but I need to use the concept :) |
| [11:09:31] |
<Getty> |
the question is what is the concept ;) |
| [11:09:37] |
<Getty> |
the ORM gives always a "special" view |
| [11:09:43] |
<pacopablo> |
multi-row primary keys |
| [11:09:59] |
<Getty> |
i have took a view on about 5 python ORMs and 3 PHP ORMs |
| [11:10:0 |
