[Asterisk-Dev] thread-safe and reentrant functions

Jon Pounder JonP at inline.net
Wed Aug 27 11:31:31 MST 2003


At 11:59 AM 8/27/2003 -0600, you wrote:
>More comments below...

Those among us that are truly gifted can write really bad code in any language.

Personally I prefer an object oriented approach to design, but without a 
language that strictly enforces it and without the overhead it creates.


>On Wed, 2003-08-27 at 11:00, Gene Kochanowsky wrote:
> > I would recommend it because it will force you to rigorously architect
> > the application and would make it easier for others to understand
> > Asterisks architecture by just examining the code. Also I think the
> > functionality of Asterisk is a natural for multiple-inheritance. I
> > also think that it would allow you to create classes for threading and
> > thread safe code that was easily maintainable across the entire
> > application from a few classes. It would make porting to other
> > platforms a great deal easier since you could also layer Asterisk. And
> > lastly, I think it would allow you to architect Asterisk to make it a
> > great deal more extensible than it is now. Design by functional
> > decomposition is okay, but doesn't compare to designing with objects.
> >
> > Gene
> >
>
>None of these arguments seem to really point to C++ per se, at least as
>as far as I can see.  Most of your items seem to deal with the
>"architecture" of Asterisk.  Can you point to specific examples where
>the current architecture is insufficient or inadequate?  I guess I don't
>see how C++ "forces" you to rigorously architect anything...  I don't
>mean for this to be a personal attack, but is it possible that you see
>C++ as the "one true way" of programming and can't/don't want to see
>that you can "rigorously architect" a program in any language?
>
>Jared
>
>
>
>_______________________________________________
>Asterisk-Dev mailing list
>Asterisk-Dev at lists.digium.com
>http://lists.digium.com/mailman/listinfo/asterisk-dev





More information about the asterisk-dev mailing list