Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 14 Jul 2009 16:00:15 +0800
From:      Li-Wen Hsu <lwhsu@FreeBSD.org>
To:        Erwin Lansing <erwin@FreeBSD.org>
Cc:        python@freebsd.org
Subject:   Re: Ports with duplicate LATEST_LINKS
Message-ID:  <20090714080015.GA60177@FreeBSD.cs.nctu.edu.tw>
In-Reply-To: <20090714075502.GO83265@droso.net>
References:  <200907091332.n69DWgXF055913@pointyhat.freebsd.org> <1e39c0a90907091948i5b11a4fdrb0d75cd08f245eac@mail.gmail.com> <20090710145547.GE86673@droso.net> <1e39c0a90907120222m4d0d7736ga6a7221e514b836b@mail.gmail.com> <20090713173739.GJ83265@droso.net> <20090714005451.GA90576@FreeBSD.cs.nctu.edu.tw> <20090714075502.GO83265@droso.net>

next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Jul 14, 2009 at 09:55:02 +0200, Erwin Lansing wrote:
> On Tue, Jul 14, 2009 at 08:54:51AM +0800, Li-Wen Hsu wrote:
> > On Mon, Jul 13, 2009 at 19:37:40 +0200, Erwin Lansing wrote:
> > > On Sun, Jul 12, 2009 at 05:22:14PM +0800, Li-Wen Hsu wrote:
> > > > >>
> > > > >> I think these two are not an issue. ?Default Python version is switched to 2.6.
> > > > >>
> > > > >> But how this report generated? ?Switching happened before
> > > > >> databases/py25-bsddb repocpoied.
> > > > >>
> > > > > The script is in Tools/scripts/check-latest-link. ?Could this be caused
> > > > > by the installed python version on the system it runs on?
> > > > 
> > > > Sounds possible, and that's what bsd.python.mk does.
> > > > databases/py25-bsddb is a slave port of databases/py-bsddb,
> > > > which generates python 2.5 package for some ports depend on
> > > > specified python version.  When a system with python 2.5
> > > > as the default setting, databases/py-bsddb and databases/py25-bsddb
> > > > should generate same package.  So now the problem is, is it OK for them
> > > > have same LATEST_LINK?  Or we can just ignore this problem, since this
> > > > should not effect official package build, and the latest links on the ftp.
> > > > 
> > > As you probably saw on the ports list, this also broke INDEX (not
> > > noticed before because the INDEX script has wedged it zfs mount).  It
> > > looks like overriding LOCALBASE to /nonexistent does fix both issue as
> > > the script will no longer see the locally installed python version.
> > > This solves this issue.
> > 
> > Sorry that I am a bit confusted about what you mean.  Are you suggesting
> > me to override these ports' LOCALBASE to /nonexistent ?  Or this means
> > that everything works fine now?
> > 
> The latter.  I was just explaining what I changed in the script to make
> it less confused.  No need for you to change anything.

That's great.  Thanks for the explanation.

-- 
Li-Wen Hsu <lwhsu AT FreeBSD.org>
http://lwhsu.org



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20090714080015.GA60177>