Re: Plans for 2.8 - GNOME Managed Language Services?

[Rob Adams <readams readams net>]

Certainly I can do a brain-dead C implementation of any algorithm that
will be slower than it's well-designed C# counterpart.  This proves
nothing.

I'm not trying to say that java/c#/etc are inferior; in fact I think
that the small performance gain by C is rarely worth it except perhaps
in scientific computing applications.

On Sat, 2004-03-27 at 10:19 -0800, Bob Smith wrote:

> Actually, if you want to get technical, there are some cases where an
> algorithm using GC will be faster then one managing its memory itself.
> The GC can move the effort of freeing memory to a part of the program
> that is not time critical which can sometimes be difficult doing it by
> yourself.
> 
> On Fri, 2004-03-26 at 15:22, Rob Adams wrote:
> > On Fri, 2004-03-26 at 18:01 -0500, Manuel Amador (Rudd-O) wrote:
> > 
> > > AFAIK, code written in C# and compiled to the Mono runtime should run
> > > equally fast as the C equivalent.  Note I said run, meaning that I don't
> > > count:
> > 
> > That's not quite true.  There is additional overhead in running things
> > like gc and runtime checks that C doesn't do.  If you want pure
> > performance, you're going to do better with C code.  The reasons for
> > using java or C# or whatever have nothing to do with performance, but
> > can rather be justified by saying that the performance is not hurt very
> > much, or that the faster development times outweigh the runtime
> > performance problems.
> > 
> > In the real world, the performance impact is quite small.  But it is
> > incorrect to say that they will run equally fast.
> > 
> > -Rob
> > 
> > _______________________________________________
> > desktop-devel-list mailing list
> > desktop-devel-list gnome org
> > http://mail.gnome.org/mailman/listinfo/desktop-devel-list
> -- 
> Bob Smith <bob thestuff net>