Around 20 o'clock on Jun 1, Bart Massey wrote:
> Is this a bug?
>
> > int f(&int[*] a) { return a[0]; }
> > f(&((int[1]){0}))
> -> f (&(int[1]) { 0 });
> <stdin>:13: Invalid lvalue
>
> I hate it when I have to ask that...
We still don't have anonymous lvalues. If you like you can use:
f (reference ((int[1]) { 0 })))
-keith