[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Chicken-hackers] Re: Making chicken-install amenable to automation
From: |
Timothy Beyer |
Subject: |
Re: [Chicken-hackers] Re: Making chicken-install amenable to automation |
Date: |
Wed, 10 Jun 2009 01:36:36 -0700 |
User-agent: |
Wanderlust/2.14.0 (Africa) SEMI/1.14.6 (Maruoka) FLIM/1.14.7 (Sanjō) APEL/10.6 MULE XEmacs/21.4 (patch 21) (Educational Television) (i386--freebsd) |
At Tue, 09 Jun 2009 13:38:51 +0400,
Aleksej Saushev wrote:
>
> Timothy Beyer <address@hidden> writes:
>
> > The non-versioned egg files makes adding chicken packages to FreeBSD ports
> > a challenge,
> > similar to the situation you describe on NetBSD.
>
> This isn't the problem actually, the problem is maintainance of such a
> package.
>
> In reality this causes pain even when not using package management system.
>
What I meant was, adding/contributing a package usually implies that the
contributor will be the maintainer as well, at least when the package has
relatively few users.
> > The only real alternative that I can think of (which would be
> > "versioned") would be subversion checkouts during the fetch phase,
> > but I seriously doubt that any of the BSDs or any of the source-based
> > Linux distributions can do such a task in an elegant or standardized
> > way.
>
> We do that in pkgsrc-wip, and it works nice on all systems I have access
> to (NetBSD, FreeBSD, Dragonfly). But this won't enter main distribution
> due to the policy.
>
Well, as it turns out, I found a port (although it currently uses a locally
hosted tarball snapshot) that in 2005 performed a cvs checkout instead of the
fetch target. However, the method is not documented in the "Porter's
Handbook", is obscure, and is probably shunned by the community. I don't know
if the situation differs for packages in pkgsrc-wip (maybe they have the code
prewritten to do version control checkouts) going into the main pkgsrc.
Tim