Date: Mon, 25 Oct 2004 10:34:56 -0700 From: Gary Kline <kline@tao.thought.org> To: Roman Kennke <roman@ontographics.com> Cc: freebsd-ports@freebsd.org Subject: Re: RELEASE_X_Y_Z branches/tags maintained?? Message-ID: <20041025173456.GA95850@thought.org> In-Reply-To: <1098697521.666.30.camel@moonlight> References: <1098641975.705.10.camel@moonlight> <1098692436.666.17.camel@moonlight> <20041025083705.GA16273@anembo.nu.org> <1098697521.666.30.camel@moonlight>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Oct 25, 2004 at 11:45:21AM +0200, Roman Kennke wrote:
> Am Mo, den 25.10.2004 schrieb Christopher Vance um 10:37:
> > >> > I have a question regarding the branches/tags of the ports tree for
> > >> > stable releases. Are they in any way maintained. For instance I would
> > >> > like to see security fixes and corrections like changed download URLs be
> > >> > committed there.
> >
> > You have a choice between
> >
> > (1) a system with fewer packages/ports, but each one related to
> > several supported OS versions,
> >
> > or
> >
> > (2) a system with more packages/ports, but they're not tied to any OS version.
> >
> > If you want something like (1) on FreeBSD, you can always capture the
> > ports tree as it was when your OS version was released (it's even
> > tagged for you) and update only those parts you care about. You get
> > to follow any advisories yourself (try portaudit). But if it breaks,
> > you get to fix all the pieces yourself.
>
> Maybe, if there is _enough_ interest, somebody (starting with me??)
> could start a separate (from FreeBSD) project, that aims to maintain a
> stable FreeBSD ports tree. It could start out with a subset of ports,
> architectures and OS versions for the beginning, and scale when
> resources are available. It could occasionally grab a tagged ports tree
> and develop a stable version out of it.
>
> What do you think?
>
<CAUTIONARY_NOTE>
I've got tarballs of portupgrades for 2 ports,
one now broken. So this may get me flamed.
... .
</CAUTIONARY_NOTE>
I think your idea has lots of merit, Roman--to stick my
two cents' worth in. Porting isn't that hard once you've
found and fixed <<whatever>> problems. Most bugs aren't that
hard to fix; some are bloody murder. After that, to create
a port for FBSD is a lot of grungy detail work. For
example, creating the patch files, then the new
distfiles and the ancillary files that make certain
that everything Just-Works{tm}.
Once you've done a few ports--either your own hacking
or someone else's--creating a port gets pretty easy.
Takes a few hours/port/architecture. Before aiming for
a separate project, it may pay to work within the ports
group for awhile.
gary
PS: I'm personally grateful for the best OS--anywhere.
But that gratitude and $6 may buy a double latté.
--
Gary Kline kline@thought.org www.thought.org Public service Unix
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20041025173456.GA95850>