svn commit: r304840 - in head/x11: nvidia-driver nvidia-driver-173 nvidia-driver-71 nvidia-driver-96

[Chris Rees]

On 25 Sep 2012 16:23, "Alexey Dokuchaev" <danfe at freebsd.org> wrote:
>
> Author: danfe
> Date: Tue Sep 25 15:23:14 2012
> New Revision: 304840
> URL: http://svn.freebsd.org/changeset/ports/304840
>
> Log:
>   Fix the problem that PKGVERSION for 173.14.xx legacy driver went
backwards
>   when it was updated to .35 on the 12th.  That moment, it kept
PORTREVISION
>   1 from the master port.  When the master port was updated and
PORTREVISION
>   removed on the 22nd, PKGVERSION of nvidia-driver-173 went backwards as
its
>   PORTREVISION dropped from 1 to 0.  While I am at it, provide a comments
in
>   all makefiles that keeping PORTREVISION setting (even when it is
seemingly
>   zero) is important.
>
>   Reported by:  erwin
>
> Modified:
>   head/x11/nvidia-driver-173/Makefile
>   head/x11/nvidia-driver-71/Makefile
>   head/x11/nvidia-driver-96/Makefile
>   head/x11/nvidia-driver/Makefile
>
> Modified: head/x11/nvidia-driver-173/Makefile
>
==============================================================================
> --- head/x11/nvidia-driver-173/Makefile Tue Sep 25 14:55:49 2012
 (r304839)
> +++ head/x11/nvidia-driver-173/Makefile Tue Sep 25 15:23:14 2012
 (r304840)
> @@ -6,6 +6,8 @@
>  #
>
>  DISTVERSION=   173.14.35
> +# Explicitly set PORTREVISION as it can be overridden by the master port
> +PORTREVISION=  1

I don't know if a special comment is needed; this is pretty standard slave
behaviour.

However, it is a common problem, perhaps we could put a note into the
Porter's Handbook in PORTREVISION or master/slave sections?

"It is usually incorrect to set PORTREVISION to 0, except in a slave port
where it should override the master port"

Hm, how best to put that into docspeak?

Chris