[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: KVC
From: |
Helge Hess |
Subject: |
Re: KVC |
Date: |
Tue, 29 May 2007 23:46:41 +0200 |
Hi Chris,
On May 29, 2007, at 19:15, Chris Farber wrote:
Is this what you are referring to?
don't know :-) I only know that OGo works 98.5% fine on Cocoa but
throws plenty of KVC exceptions on gnustep-base :-/
In OGo we rather often override those two with custom implementations
(sometimes for no good reason ;-):
-valueForKey:
-takeValue:forKey:
but usually none of the others. In fact the 'nil' thing should never
be called because we do implement -takeValue:forKey: and never fail
due to 'nil' :-) Maybe the GS code checks the method signature and
resorts to the nil thing for base types before even attempting to
call -takeValue:forKey:? No idea.
BTW: IMHO you should not assume that the documentation represents the
actual Cocoa behaviour :-) Especially in compat mode ...
Thanks,
Helge
PS: I have _absolutely_ no idea why Apple changed the method name. It
can't be just to please the Java setXYZ guys (in this case they would
have also reverse the ordering of the arguments to setKey:toValue: or
something) ...
--
Helge Hess
http://www.helgehess.eu/
- KVC, Helge Hess, 2007/05/28
- Re: KVC, Nicola Pero, 2007/05/28
- Re: KVC, Helge Hess, 2007/05/29
- Re: KVC, Richard Frith-Macdonald, 2007/05/29
- Re: KVC, Helge Hess, 2007/05/29
- Re: KVC, Chris Farber, 2007/05/29
- Re: KVC,
Helge Hess <=
- Re: KVC, Helge Hess, 2007/05/31