2008-7-19

00:02 elliott
wow
00:02 elliott
that WAS close
00:05 Gulopine crosses his fingers that 7967 didn't break any file stuff
00:07 Gulopine
175 files changed in one form or another
00:07 Gulopine
2 conflicts
00:07 Gulopine
not bad :)
00:07 Gulopine
hah, and they're exactly where i expected
00:08 cramm
:)
00:39 Newmaniese
Sweet! NFA hit trunk!!!
00:46 Gulopine sighs
00:46 Gulopine
i hate doctests :(
00:46 Gulopine
or, more appropriately, i hate the model_forms doctests
00:47 Newmaniese
Gulopine: are these test issues with your file storage patch?
07:13 zerok
if anyone can find some time for it: http://code.djangoproject.com/ticket/6427 is beyond "Accepted" IMO ;-)
08:19 tail
freakboy3742: do you have a minute to see if the patch I added to #7611 is the right solution?
08:19 DjangoBot
08:19 freakboy3742
Looking now
08:21 freakboy3742
The reverse approach is a novel idea
08:21 freakboy3742
however, I'd rather see something systemic, rather than a once off
08:21 freakboy3742
The broader issue is something that will potentially affect other applications
08:21 tail
by that, do you mean something like #4788?
08:21 DjangoBot
08:22 freakboy3742
Possibly
08:22 freakboy3742
although there are other options
08:22 freakboy3742
e.g., a decorator for 'only test if url is deployed'
08:23 freakboy3742
Sorry - I have to run - my son needs some attention
08:23 tail
np, thanks
08:23 freakboy3742
I will be back later this evening if you want to dig into this further
14:06 jbronn
the svn:ignore *.pyc property should be set in django/core/files
14:09 freakboy3742
jbronn: I'm on it. If you find any others let me know.
17:52 ChrisHasenpflug
slow day today :-) everyone must be wiped from the big merge
20:15 Alex_Gaynor
brosner: #7827
20:15 DjangoBot
20:15 brosner
hmm, i think this is a duplicate
20:16 brosner
but is correct. it needs tests
20:16 Alex_Gaynor
tests it is then!
20:17 eddyMul
I got r7985, and django/conf/project_template/urls.py is missing the 'r' for admin
20:17 DjangoBot
20:17 eddyMul
should I file a bug?
20:18 brosner
not really needed
20:18 Alex_Gaynor
brosner: Where should the tests go?
20:19 brosner
check regressiontests/modeladmin
20:19 Alex_Gaynor
ok
20:19 brosner
but that might not be the best place. hmm.
20:19 brosner
actually admin_views may be better
20:20 Alex_Gaynor
Ok
20:36 Alex_Gaynor
brosner: Here's a first pass at the tests: http://dpaste.com/66136/ do these look right(admin_views tests aren't running for me right now, trying to sraiten that out)
20:38 brosner
your tests dont test the failure case
20:38 brosner
and theres no reason for a custom manager
20:38 brosner
just an overridden queryset method in the ModelAdmin
20:39 Alex_Gaynor
Ah, ok, I'll fix that up
20:41 brosner
and tons of typos, but i am sure running it finds those ;)
20:41 Alex_Gaynor
:D
20:53 Alex_Gaynor
brosner: Ok this should be much better(and actually work): http://dpaste.com/66145/
20:54 brosner
yay for runtime tests
20:54 Alex_Gaynor
These are my first non doc-tests and I gotta say, I like it
20:54 Alex_Gaynor
(mostly because my screen isn't one giant green string)
20:55 brosner
hehe
20:55 brosner
yeah no kidding
20:55 brosner
ok prefect
20:55 brosner
now
20:55 brosner
go find the duplicate. i *think* there is one, not sure
20:55 brosner
well, you are the duplicate. go find the original ticket.
20:55 Alex_Gaynor
Haha, ok
20:56 ChrisHasenpflug
brosner: do you have any idea what's up with the test that's failing since the merge?
20:57 ChrisHasenpflug
i think it was failing pre-merge?
20:57 brosner
they were failing prior
20:57 brosner
i just ignored them at the time (my bad) :P
20:57 ChrisHasenpflug
haha
20:57 ChrisHasenpflug
any idea what the root cause may be?
20:57 brosner
they were completely unrelated to what i did. which is odd. one sec let me verify something.
20:57 Alex_Gaynor
brosner: Ok, it's #7510 want me to close mine in favor of it, or the reverse
20:57 DjangoBot
20:59 brosner
close yours in favor of theirs. and i am not sure what they mean by this not being dry. are the has_*_permission method not honoring queryset?
21:01 Alex_Gaynor
I don't get it either :/
21:01 brosner
if the object is not in the queryset for the ModelAdmin then likely should raise an object not foud
21:01 brosner
customizing the queryset really does give a new meaning to ModelAdmin
21:02 Alex_Gaynor
Yeah, it raises a 404 if the user would have permision and the object doesn't exist
21:09 Alex_Gaynor
brosner: Full tests pass with this for me, FYI
21:10 brosner
and they fail with your change?
21:10 brosner
(which would be a good thing) ;)
21:10 brosner
*without
21:11 Alex_Gaynor
Yeah, without the changes the change view ones fail, the changelist ones pass
21:12 brosner
good
21:20 carljm
anyone around who wants to verify that I'm not off my rocker, and ticket #7843 is as bad as I think it is?
21:20 DjangoBot
21:20 carljm
I think r7986 totally broke all FileField and ImageFields uploaded prior to today
21:20 DjangoBot
21:21 carljm
if i'm right, seems like something that ought to get reverted post-haste
21:25 dwsBeta2
i'm seeing a doctest failure in modeltests.delete.models, and tracked the introduction back to r7967. :(
21:25 DjangoBot
21:26 ChrisHasenpflug
dwsBeta2: that test was failing in the nfa branch before the merge
21:26 dwsBeta2
is someone working on it?
21:27 brosner
i believe it began on the branch when i did the ModelAdmin validation, not 100% sure though (for those looking to track it down)
21:27 brosner
which seems odd since they are completely unrelated as far i can tell
21:27 dwsBeta2
i'll help, but the nfa code is all new to me.
21:27 ChrisHasenpflug
unfortunately buildbot doesnt go back far enough to see where it was passing
21:44 uzi
dwsBeta2: you're not the only one... #7825
21:44 DjangoBot