[cairo] Re: Re: Re: Re: cairo c++ bindings

[Mike Hearn]

On Fri, 25 Nov 2005 15:15:07 -0500, MenTaLguY wrote:
> It's more an issue of the standard ABI for the particular OS/processor
> combination.  I don't remember offhand which it has been an issue for in
> the past (the last time the issue came up for me was a long time ago),
> although I want to say Alpha was one of them.

Yes, I can believe that.
 
> There's also the issue of -- if you have cairo callbacks that can throw
> C++ exceptions -- you will need to be _very_ careful in structuring
> cairo itself so that throwing an exception will not cause a resource
> leak, because any cleanup bits in the C call stack will get ignored as
> the stack is unwound.  

Yes, I touched on exception safety in my original mail. You're probably
right, it may not be worth it. A JOGL style approach might be better (with
swappable bindings, one checked, one not).

thanks -mike