> > btw, i just gave abi a compile last night i was a bit miffed to see the
> > spew of errors from mswordview (a bit shamed really). And i wondered about
> > the errors from constructs like
> >
> > unsigned char test:3;
> >
> > I think that this results only in a warning with gcc's -traditional option, but
> > id like to be reassured that bitfields are an actual part of ansi c, and that
> > there no hidden compiler dependancy in using them ?, anyone know ?
>
> To me it looked like it was dissatisfied with the "char" part of it, not
> the ":3" part of it. "unsigned int test:3;" ought to work beautifully?
Yeah, bitfields are in the spec, but they are always prefixed by either
"signed" or "unsigned" and that's it.
The above should be:
unsigned test:3;
The Word8 spec, however, make extensive use of things like
un/signed <type> foo:4;
which is just another example of Microsoft's non-standard extensions
(unnamed unions are another example). You probably wrote the above line
after reading too many MS headers.
It should work fine the correct way.
Justin
