Re: ObjC additions (Was: Re: PROPOSAL: Objective-C++)

[Philippe C.D. Robert]

Hi,

just to mention it, I support the idea of having real ObjC exceptions, of 
course!

On Tue, 20 Nov 2001 20:20:53 +0100
"Philippe C.D. Robert" <phr@3dkit.org> wrote:
> > > > 1. Exception handling.
> > Seconded. I'd like to see this as well. Personally, though, I'd prefer 
> > something along the lines of Python, which has try...except...finally but 
> > also try...except...else (the else gets executed in case no exception has 
> > occurred).
> > > > 2. Class variables. A "must add" in my opinion, as the current way of
> > > > defining them using static variables is far from being acceptable. 
> > Yep.
> > > > 3. Blocks, as described by Brad Cox.
> > Yep.
> > > > 4. Syntactic sugars for constant property objects.
> > Yep.
> 
> Uhm ... just one question: why? There are rate situations where such features 
> really are needed. So while I agree that there are (some) situations where 
> such (theoretical/academical) add ons might be 'nice', I much more fear 
> languages which are too complicated, bloated, have too many features etc. ... 
> and I know what I am talking about, I earn my living writing C++ code...;-)
> 
> Smart (the car company) once had a brilliant PR strategy saying: Reduce to 
> the max! This is exactly what we need and what makes ObjC such a beautiful 
> language! 
> 
> > > 5. how about operators? 
> > Please, no. IMHO, you don't gain much with this, but it makes debugging a 
> > nightmare. I'm in the process of debugging a program which makes use of 
> > operator overloading in C++. I don't want operators. No. :-/
> 
> Well.... I should actually agree, esp. after my statement above, *but* did 
> you ever do maths in ObjC ... *Very* ugly and *very* annoying (not to speak 
> of string handling in OpenStep/ObjC...).
>  
> Anyway, let me throw in another issue. Coming from a parallel world (SGI 
> machines with only 1 CPU are bastards...grin) I'd vote for another feature:
> 
> - synchronisable blocks and methods
> 
> But then, maybe this would do harm to the simplicity of the language as 
> well...
> 
> -Phil
> -- 
> Philippe C.D. Robert
> Software Engineer
> Silicon Graphics, Inc.
> 
> _______________________________________________
> Discuss-gnustep mailing list
> Discuss-gnustep@gnu.org
> http://mail.gnu.org/mailman/listinfo/discuss-gnustep
> 


-- 
Philippe C.D. Robert
Software Engineer
Silicon Graphics, Inc.