bug-gnustep
[Top][All Lists]
Advanced

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

Re: [PATCH] gnustep-gui cleanups


From: Nicola Pero
Subject: Re: [PATCH] gnustep-gui cleanups
Date: Thu, 30 Jan 2003 15:46:01 +0000 (GMT)

> NP> Yes - we shouldn't have all those warnings around - thanks for
> NP> fixing them.  I applied your patches :-)
> 
> I forgot to mention: RTFConsumer.m left unchanged
> 

Thanks - sorry, I didn't see the warning as that's not generated by older
compilers, and didn't want to change 'self' to a hardcoded class name as
that potentially limits subclassing (not really relevant here, maybe, but
still psychologically disturbing).

The code is somewhat right and the compiler could be smarter and figure
out what's happening, and not generating the warning :-)

if it doesn't find the method in interfaces or protocols, it looks for it
in the current implementation context, but does not look for it in
implementation contexts of superclass(es) which might have already
compiled, so it doesn't realize the method has been implemented in the
same compilation unit, and ends up generating this sort of spurious
warning.

I have clean testcases, but don't have the time to fix/improve the
compiler now -- maybe in the future.

Still, the code was right but clunky.  It had a few `handy' methods, which
were not public though, and never used in the code itself :-)  I think it
was a historical relict of changes we did in the past.  I just killed
those (currently pointless) methods and cleaned the code.

Thanks for pointing out that this issue still needed resolving!





reply via email to

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