Re: How to keep debian current??
Hi, Celejar:
On Sunday 23 May 2010 21:42:44 Celejar wrote:
> On Sat, 22 May 2010 03:03:42 +0200
> "Jesús M. Navarro" <jesus.navarro@undominio.net> wrote:
>
> ...
>
> > My simple rule about Debian has always been:
> > * Stable, if you just want to use Debian.
> > * Testing, if you want a peek over what Debian will be on next release
> > and want to help to hunt down the non-obvious bugs (probably because you
> > depend on the quality of Debian Stable and that's what you can do to help
> > going for it).
> > * Sid, if you look for fun and have at least a mild desire to become a
> > day a DD. If you don't want to open and follow a lot of bugs, provide
> > patches from time to time and follow the devel lists, you'd probably be
> > better out of the loop and stay on Stable or Testing.
>
> You omit two very good reasons (although they certainly aren't
> dispositive, and will not be relevant for many) to use Sid: support for
> newer hardware, and inclusion of newer software.
>
> I know that this has all been rehashed a million times already, but I
> just wanted to clarify your summary.
Well, I did omit those two reasons, but I didn't forget about them: it is that
I don't consider them reasons to run Sid, much less to counsel anyone else to
do it. Neither new hardware nor new software are reasons (on my own regard,
of course) to run a whole in-development OS (which is what Sid is).
You stay with known-to-work hardware and software on production environments
to start with (non-production environments can, of course, qualify on both
the "peeking into the future" and "for fun" clauses). If you can't stay on
the known path, then you backport what you need; this way at least you'll
have a clear notice about where your (future) problems are coming from.
And then, if once everything considered it really becomes attractive to go for
a future, in-development version (basically, too much of a burden on the
backporting effort), you still have a case for Testing, not Sid (I for one
I'm open to consider Testing as an alternative on my workplace... for new
projects that can provide strong reasons to do it and once Testing
is "frozen" so probably it will already be Stable by the time the project
goes "live").
Cheers.
Reply to: