[Gnash-dev] Re: shared pointers

[Mateusz Loskot]

Rob Savoye wrote:
> Mateusz Loskot wrote:
>> Rob Savoye wrote:
>>> Markus Gothe wrote:
>>>> GCC is a gnuism with it's own non-iso extensions, not a real compiler
>>>> for except x86-based systems imo. ;)
> 
>> Honestly, I was hoping Rob will tell us the truth about it :-)
> 
>   Actually, since I avoid using the extensions,

Yes.

> what's left is a nicely ANSI conforming C++ compiler,
> especially if you use the newer 4.x version.

Right.
The only think that's missing for me is to get rid of problem described
here (casting pointer-to-function to pointer-to-object, and vice versa):

http://www.trilithium.com/johan/2004/12/problem-with-dlsym/

but this problem applies to C++ Standard than to particular compiler.
But GCC could be nice and fix it even then :-)

> I find this statement amusing, as for many years most people
> thought it worked better on all the non x86 architectures. :-) The x86
> really only started to catch up with GCC 3.x.
> 
>   Personally, it's VC++ that I find obnoxious, as it doesn't even really
> support the ANSI C++ spec.

I found Visual C++ 2005 as well supporting ANSI C++.
However, it's a crap because does not support C99.

Here is my (mloskot) story and funny replies from MS :-)

http://forums.microsoft.com/MSDN/ShowPost.aspx?PostID=259043&SiteID=1

Cheers
-- 
Mateusz Loskot
http://mateusz.loskot.net