Date: Sat, 10 Jun 2006 23:35:54 -0700 From: Doug Barton <dougb@FreeBSD.org> To: Sam Lawrance <boris@brooknet.com.au> Cc: FreeBSD Ports <ports@FreeBSD.org>, Boris Kovalenko <boris@tagnet.ru>, Peter Losher <Peter_Losher@isc.org> Subject: Re: FreeBSD Port: net/quagga Message-ID: <448BB9CA.8050901@FreeBSD.org> In-Reply-To: <64E1E60A-9AFF-4C93-A8D1-D6CC8B1BCBDC@brooknet.com.au> References: <448B0127.1040107@isc.org> <448B655E.70003@FreeBSD.org> <448B6EDF.4040401@isc.org> <448B74BF.20203@FreeBSD.org> <64E1E60A-9AFF-4C93-A8D1-D6CC8B1BCBDC@brooknet.com.au>
next in thread | previous in thread | raw e-mail | index | archive | help
Sam Lawrance wrote:
> In that case we should remove the check from portlint, until
> USE_RC_ORDER is no longer required by supported systems. Otherwise it's
> just another confusing portlint warning.
I disagree. I would rather have to change a small number of ports back that
really do need it, than to continue having ports that don't need it
needlessly spam /etc/rc.d. I maintain a lot of ports that have weirdness in
them that portlint doesn't like. It's a tool, not a report card. :)
> The port author removed it because portlint said it was deprecated.
Even if that is true, it sparked a lively discussion, so it's not all bad.
> Can USE_RC_ORDER still be used on all supported FreeBSD versions without
> negative effect?
As far as I know, yes, but that's not really the point. For those systems
that are past the local_startup merge, it's better to have them install
their ports startup script in $PREFIX, since that is the direction that all
of this is ultimately moving towards. I have no objection to preserving
legacy behavior for those ports (and systems) that really do need it, but
those that don't shouldn't be hamstrung by it.
> If there is a need to conditionalize behaviour on
> OSVERSION, it could be handled in bsd.port.mk (depending on how many
> ports need it, really).
I agree with both parts of your statement here. My intent was to provide a
solution that would work immediately to make life a little easier on the
port author. If this really does turn out to be a larger problem, then a
larger solution is a SMOP.
Doug
--
This .signature sanitized for your protection
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?448BB9CA.8050901>