On Tue, Nov 25, 2008 at 06:49:02PM +0000, Darren J Moffat wrote:

> Bart Smaalders wrote:
> >John Levon wrote:
> >>On Sun, Nov 23, 2008 at 07:35:28PM -0800, Bart Smaalders wrote:
> >>
> >>>variant.arch             sparc, x86, zos
> >>Shouldn't these match up with what uname -p returns?
> >>
> >>regards
> >>john
> >
> >Probably so... anyone tried zos :-)?
> >
> >Seriously, though, I'm not sure that i386 captures this variant
> >fully.
> 
> Maybe it should be `uname -m` rather than `uname -p` but then that gives 
> sun4v rather than space.  But maybe that is what we want.

To quote Bart:

"In IPS, options that may be selected or not selected, such as various
locales, documentation, etc, are referred to as facets.  Options which
are mutually exclusive are called variants."

We could invent a new string 'x86' and make things even more confused
than they are already: I doubt *anyone* can clearly elucidate when you
should use uname -p, uname -m, and uname -i. 'i386' isn't an ideal
string, but at least it's familiar to uname users.

> I think that `uname -p` is far too limiting for variant.arch on x86 but 

For what purpose? Are you imagining a system where /usr/bin/ls can be
either 32 or 64 bit?

regards
john
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

Reply via email to