gnustep-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: RFC: [GDL2/base] unableToSetNullForKey: / unableToSetNilForKey:


From: David Ayers
Subject: Re: RFC: [GDL2/base] unableToSetNullForKey: / unableToSetNilForKey:
Date: Wed, 26 Mar 2003 14:33:32 +0100
User-agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.3b) Gecko/20030210

Hi Manuel,

Manuel Guesdon wrote:

On Tue, 25 Mar 2003 00:07:24 +0100 David Ayers <address@hidden> wrote:

>| David Ayers wrote:
>| >| > Should/could we implement something like GSObjCReplaceMethod(Class >| > class, struct objc_method *old, struct objc_method *new) or should >| > this be handled differently?
>| >
>| Hi Richard,
>| >| I know this is a rather "brutal" approach (but then again so was the >| behavior suggestion), yet I believe it should work for both runtimes and >| be useful in similar situations. Would you consider the attached patch >| as acceptable? If so I would commit.

This kind of method would also solve other problems, for example my 'combined 
key' problem :-)

Well actually the problem can be solved the way it is now by leaving the implementation of this special case handling for valueForKeyPath:/takeValueForKeyPath: in NSDictionary as base's and (I believe) Foundation don't override this in NSDictionary. Or do you want to extend this mechanism to all (i.e. custom) objects? I can't really see the usefullness of a key like "'some.key.with.periods'" for any type of object except NSDictionary... But if you like, I can see if I can replace NSObject's implementation for GDL2.

Cheers,
Dave







reply via email to

[Prev in Thread] Current Thread [Next in Thread]