DjangoBot: #django-dev

change the telenieko@ for whatever ;)

jacobkm: i actually dont know. i remember coming across it and it turned out to be more difficult for than just re-doing it. i remember some folks in #git having some idea though

Meh, in that case I don't care.

<g>

telenieko: That's what I have, minus the prefix stuff

:)

Alex_Gaynor: I also did it whithout the prefix, and I'm now redoing it with prefix 'cause git branch -r is a full mess!

git svn init is slowwww

jacobkm: rebuilding the repo from the filesystem (access to the direct svn repo) would be quite fast ;)

brosner: again, meh

i just haven't figued out how to translate that to work remotely. so not really worth it now that i think about it.

mm.. jacob could you do a clean repo with a prefix and put it tgz'd somewhere? that be much faster :)

telenieko: maybe at some point; I'm done git-wrangling for today

ok, I'll tell my home pc to do the clone. I can't unplug the laptop now! :)

time to go

c u later

Does anyone see anything in this diff: http://dpaste.com/57590/ that would cause this error: http://dpaste.com/57589/ , I'm working on #4117

hi all

Hey

how many new git repos are now? :)

Counting the one on my hard drive no one but me gets to touch?

It's just a local repo, if any of you guys have access to it, I have a problem :D

ah,.. git remote add alex hack+local:///home/alex/dev/django

npe, no luck. Did I guess the folder atleast? :)

Close, /home/alex/django_src/

argh! hack+local must be php based, it doesn't work!

;o)

Yeah, PHP isn't strongly typed so hack is assumed to be a string, and local is a file obj, somehow they get combined into an array :P

iep, but works on IIS :! (if that's a feature)

Haha

we have a ticket about runnning on IIS, we'll be at the same level as PHP then!! hehe. That's post-1.0 I hope

hehe

For the record, #2039

jesus the patch on there is unpython, self != This :D

def __init__(This) < That rocks

uhu

specially the cap on T

even in j?v? it's called this and not This ;)

That code scares me

it's almost c#

as far as i've seen, they've got everything is cappitalized

I wonder where's Env defined

Plus the * import

Didn't see the * :)

or where: Write(chunk) # comes from... probably also the *

going for dinner see you in a while!

have fun telenieko :)

btw the PyISAPIe code that server handler is using is also ridden with references to This, * imports and upper-case-fever

This.Back()

visual basic in my python?

Should we add a line in contributing.txt: Please try to make your diffs as if you'd read PEP8

xD

Does anyone have any opinions on #4117

a) attrs=({}, {}) b) attrs={} (applies to all)

or: attrs={}, attrs_widget1={}, attrs_widget2={} ...

You asked, I just made as if I tought about it :P

iterable attrs involves backwards-incompatibility

and special-casing

The main case I'm thinking about is I'd like to have the id attr set on the ul that contains the checkboxes or radio widgets

Alex_Gaynor: nested dict, maybe?

attrs = {'common-attr-1': value, 'widget1': {'attr-only-for-w1': foo}, 'widget2': {'attr-only-for-w2': bar}}

seems a bit magical

hylje: what would you suggest?

Hrm, I think that might be too much stuff, at that point you're better off redoing the renderer, I don't have a problem with the ability to specify attrs for the ul tag as opposed to the <input>s though

separate argument dict for the subwidget-specific stuff in the way you propose

i'm not overly familiar with how widgets work; i though suggest the widgets are specified by element type (e.g. ul, select, input)

which should be clear enough and have enough granularity for the purpose

I think what I like best is to keep the one attrs dict, any regular attrs are placed on both, anything in root is placed on the ul and anything in sub is placed only on the inputs(those names could be changed), thoughs?

er, what?

Let me make an example

is there any particular reason there's a RadioFieldRenderer but not a CheckboxFieldRenderer?

http://dpaste.com/57612/ so in this example both the inputs and ul would have class, and the UL would get the 'root' style, and the inputs would get the 'sub' style(they overide the general one)

insin: #4592

ta

Another option would be to give the ul tag a seperate arg with it's attrs(but still add the id I think), this is more backwards compatible

but fairly specific

considering other things which could use the same functionality

anyone have any long hanging fruit for 1.0 that i could start working on?

or help out with :)

The other alternative would be to just have this patch add id, and then leave everything else for the ticket about making CSS work better with newforms

forsaken: think of a number between 1 and 7500 and you have high chances of finding a good ticket.

1/7 about

Any opinions on my last suggestion?

telenieko, heh, okay :) that's what i figured

is it best for C.objects.filter(b=B()) to throw a DoesNotExist error when B is a subclass of A (meaning multi-table inheritance)? i am curious if that behavior should be improved so it evaluates to an empty list

perhaps this has been reported before?

forsaken: you can go to the Roadmap and take a ticket marked for 1.0 if you prefer. But any ticket you work on will be helpfull

forsaken: http://tinyurl.com/4xb4k5 -- a good start

forsaken: ^^ tickets flagged for 1.0 without patches, grouped by component.

jacobkm, yea, thanks

i'll poke around

jacobkm: hey Jacob, fellow lawrencian here. I just rebased the view permission patch to nfa HEAD. How many ad astras will it take for it to be reopened for consideration? ;)