imacarthur wrote:
> On 14 Mar 2009, at 20:14, Greg Ercolano wrote:
>> If #warning is a gnu only thing, I vote (a).
>
> I've been away and missed a lot of this, so I'm not sure what the
> background is (I suspect it's all my fault, however...) but for the
> record some non-gnu compilers do honour #warning messages (e.g. diab
> does, but how useful that is to most folk I can't say.)
>
It's all good -- we've #if'ed warnings on only for gnu,
to avoid getting too complicated identifying specific compilers.
In the end, #warning's are just for the developers, and since
fltk wouldn't get released without tests on unix/gnu, the errors
won't go unnoticed by us.
What's important is that #warnings don't break compiles
on machines/compilers that don't support it. (SGI's native
compiler being the one that caught me)
Albrecht suggested making an #ifdef FL_WARNING macro (controlled
by configure or some such) so that in the code, we'd just see:
#if FL_WARNING
#warning Some important message
#endif
Too bad we can't make a macro called WARNING that replaces
itself with #warning on machines that supported, and /* */'s
for machines that don't. But I'm not sure cpp lets you make
macros that resolve to other # commands.
_______________________________________________
fltk-dev mailing list
[email protected]
http://lists.easysw.com/mailman/listinfo/fltk-dev
