On Friday 30 July 2004 15:05, Michael Matz wrote:
> Hi,
>
> On Wed, 28 Jul 2004, George Staikos wrote:
> >   (really, explicitly specifying the signedness may only help half of
> > the cases I've seen.  don't go there...)
>
> Don't know what you want to say with this.  "unsigned char" and "signed
> char" are perfectly defined.  Only "char" has the problem of platform
> dependend signedness.

On Friday 30 July 2004 11:53, Richard Dale wrote:
> I defined a dcop slot like this:
>
> k_dcop:
>     char foobar(char c);
>
> And it generated this code, which doesn't compile:

But:

k_dcop:
	signed char foobar(signed char c);

Works fine..

-- Richard