Subject: Re: pkg/15128: pkgsrc/lang/guile is too interactive
To: Urban Boquist <email@example.com>
From: gabriel rosenkoetter <firstname.lastname@example.org>
Date: 01/03/2002 17:31:58
Content-Type: text/plain; charset=us-ascii
On Thu, Jan 03, 2002 at 11:10:08PM +0100, Urban Boquist wrote:
> The patch makes the problem go away, but if it isn't correct I guess
> the package should be marked IS_INTERACTIVE instead. But that seems a
> bit unnecesary.
Or just not test SIGWINCH if there's no TTY. I mean, we know ours
works, right? (That reeks of hubris, of course. And, anyway, pkgsrc
oughtn't necessarily be NetBSD-dependent. Cf, Zoularis.)
I'm not sure that just making the problem disappear means things
are actually working right. Feeding a program that wants a TTY
/dev/null isn't the same thing as feeding it a TTY, I don't think.
But I'm speaking without having looked at any of the code involved,
much less how we implement SIGWINCH, so hopefully someone with more
of a clue will comment.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (NetBSD)
Comment: For info see http://www.gnupg.org
-----END PGP SIGNATURE-----