Date: Wed, 25 Jun 2003 16:53:23 -0500 (CDT) From: Mark Linimon <linimon@lonesome.com> To: Christian Weisgerber <naddy@mips.inka.de> Cc: freebsd-ports@freebsd.org Subject: Re: Ports that don't run on !i386 Message-ID: <Pine.LNX.4.44.0306251647580.31099-100000@pancho> In-Reply-To: <bdd11o$6mc$1@kemoauc.mips.inka.de>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 25 Jun 2003, Christian Weisgerber wrote:
> I can slap a NOT_FOR_ARCHS=alpha on it, but that sounds too final.
> Nobody it going to try to fix a port once it is declared not to
> run. Besides, it's likely that there are more architectures affected.
AFAICT the plan ought to be as follows: if the port can never run
on alpha (i.e. it's i386-architecture-specific by design, not by
accident), then that's the time to use NOT_FOR_ARCHS or ONLY_FOR_ARCHS.
Otherwise the Right Thing IMHO is to do something like
.if ${MACHINE_ARCH} = alpha
BROKEN = "will not currently compile on alpha"
.endif
This way the bento logs will still be produced, but users curious
to find out about the state of a particular port on their architecture
won't have to wait through an entire compile cycle only to get
frustrated. The bento logs can then still serve as gentle reminders
to ... someone ... that these things don't work.
And I'm not just saying "someone ought to do something", I just
turned in a bunch of PRs to try to make some things that currently
say BROKEN into NOT_FOR_ARCHS and vice versa. More are needed.
As for the problem of getting people people to do testing and
patching to fill in the gaps, I unfortunately don't have any
brilliant insight, but would welcome some.
Mark Linimon
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.LNX.4.44.0306251647580.31099-100000>