[Info-vax] Dave Cutler, Prism, DEC, Microsoft, etc.

[toby]

On Dec 17, 2:42 am, Michael Kraemer <M.Krae... at gsi.de> wrote:
> toby schrieb:
>
>
>
>
>
> > On Dec 16, 9:40 pm, JF Mezei <jfmezei.spam... at vaxination.ca> wrote:
>
> >>toby wrote:
>
> >>>In a standard C compiler where mystring is an array,
> >>>mystring  ,  &mystring  ,  or  &mystring[0]  are all the same pointer.
>
> >>In ANSI C, &mystring is not accepted.
>
> > My gcc 4.1.2 with -ansi -pedantic accepts it. It always gives a
> > warning, "assignment from incompatible pointer type".
>
> > e.g.
> >     char a[], *p;
> >     p = &a;
>
> which seems correct, IMHO.
> p is of type "char *",
> &a is of type "char **".

No,  &a  for an array has the same bit pattern as  a  and  &a[0]

> So mystring and &mystring are not the same,
> mystring and &mystring[0] are.


If a had been a pointer e.g. char*, then yes, then  &a  would be the
address of pointer a and typed char** - not the value of a. But arrays
behave slightly differently.