Re: Defect in XSH utimes

["H. Peter Anvin" <yyy@xxxxxxxxx>]

Ulrich Drepper wrote:
> Paul Eggert wrote:
>
> > If the Solaris futimesat() interface is carved in stone, then we could
> > call the variant with struct timespec 'utimesat', as H. Peter Anvin
> > suggested (as the leading 'f' is somewhat weird anyway).
> That's just because Peter doesn't understand the interface completely.
> Without a name (i.e., pass NULL) futimesat corresponds to what is
> futimes on many platforms.  Therefore the 'f' is completely fine.
Saying "it not inconsistent if you use it in this edge case" doesn't 
mean it's not inconsistent.

	-hpa