[OGo-GNUstep-Port] exception when entering a project

Sebastian Reitenbach gnustep-port@opengroupware.org
Tue, 27 Nov 2007 18:18:59 +0100


gnustep-port@opengroupware.org wrote: 
> 
> On 27.11.2007, at 17:28, Helge Hess wrote:
> 
> > On 27.11.2007, at 15:53, Sebastian Reitenbach wrote:
> >> (gdb) frame 1
> >> #1  0x0ba90e28 in -[NSObject(KeyValueCoding)  
> >> setValue:forUndefinedKey:]
> >> (self=0x86859688, _cmd=0x2bad3510,
> >>    anObject=0x86859768, aKey=0x7f6f27e8) at NSKeyValueCoding.m:415
> >> 415           [self handleTakeValue: anObject forUnboundKey: aKey];
> >> (gdb) print *self
> >> $1 = {isa = 0x230d38a0}
> >> (gdb) po self
> >> <EOGenericRecord: description Project attributes={dbStatus =  
> >> inserted;
> >> endDate = "2028-12-30 23:00:00 +0000"; isFake = 0; kind = "<NSNull:
> >
> >
> > I think we broke this with the last patch. EOGenericRecord does  
> > define -takeValue:forKey: which we now do not call anymore (but - 
> > setValue:forKey: instead).
> > This seems to break all 'old style KVC' code. Notably this does NOT  
> > break on MacOS (which supports both styles [-setValue:forKey: first  
> > checks for -takeValue:forKey:]]). So I still consider that a basic  
> > GS issue.
> 
> 
> This used to be a gnustep-base problem which I fixed quite a couple of  
> months ago. Sebastian, which version of gnustep-base do you use? If  
> it's an old version, please try a recent one from svn.
It is -base version 1.15.1, the unstable that got announced some weeks ago.

Sebastian

> 
> Cheers,
> 
>    Marcus
> 
> -- 
> Marcus Mueller  .  .  .  crack-admin/coder ;-)
> Mulle kybernetiK  .  http://www.mulle-kybernetik.com
> Current projects: http://www.mulle-kybernetik.com/znek/
> 
> 
>