[linux] Debian Install: 1 CD

Peter Abrahamsen peidran at u.washington.edu
Thu Sep 25 01:25:05 PDT 2003


Yes. Add both testing and unstable, or stable and testing, or whatever,
lines to /etc/apt/sources.list. Then, in /etc/apt/preferences, put something
like:

Package: *
pin: release a=testing
Pin-Priority: 500

Package: *
Pin: release a=unstable
Pin-Priority: 50

Changing the release names according to what you're trying to accomplish.

And with regard to the earlier part of the thread: yes, the problem is that
the release cycle is absurdly long - too long to be useful for many, many
people. If Debian could order things so that they had a, say, 3-6mo release
cycle, many more people would probably use stable. Release early, release
often. I'm convinced it's doable (and yes, I know Debian's huge). There have
been good ideas suggested in the past for how to accomplish this and keep
everyone sane, but it's never happened.

Peter

On Wed, Sep 24, 2003 at 05:43:34PM -0700, Cere M. Davis wrote:

> It's a great idea but I've never been able to get that to work. Is there

> some kind of trick to this?

>

> On Wed, 24 Sep 2003, Phillip Garland wrote:

>

> > Date: Wed, 24 Sep 2003 17:36:05 -0700 (PDT)

> > From: Phillip Garland <pgarland at u.washington.edu>

> > Reply-To: Linux/Unix Users Group at the UW <linux at u.washington.edu>

> > To: Linux/Unix Users Group at the UW <linux at u.washington.edu>

> > Subject: Re: [linux] Debian Install: 1 CD

> >

> > what about

> >

> > apt-get -t testing install $PACKAGE

> >

> > of course you will also be grabbing all of $PACKAGE's dependencies from Testing,

> > but there's no way around that.

> >

> > ~Phillip

> >

> > On Wed, 24 Sep 2003, Cere M. Davis wrote:

> >

> > >

> > >I would love to just stay on stable but my reasons for using testing are

> > >the same as most people...we want to use more current software releases

> > >since the release cycle for stable is fairly long. Currently there isn't

> > >a really straight foreward way to just grab packages that you want to use

> > >from the testing repository and still use a basic baseline repository.

> > >This is the reason, I think, that most people end up just moving whole hog

> > >over to testing.

> > >

> > >These issues seem to be a old as the hills. It's true what Tony said I

> > >think.

> > >

> > >

> > >> As a Debian developer, my answer would be: don't use testing. The point

> > >> of testing is to have a collection of fairly recent software that stays

> > >> internally consistent, and relatively free of serious bugs. It is a

> > >> development tool intended to make the transition from unstable to stable

> > >> easier for us (Debian, that is), not to be a sort of meta-stable

> > >> distribution.

> > >>

> > >> -brad

> > >>

> > >>

> > >> >

> > >> > >

> > >> > > There are other problems with testing (no Gnome 2.2 for example). However, I

> > do

> > >> > > like Debian, I just wish these problems could be fixed.

> > >> > >

> > >> > > David Fetrow <fetrow at apl.washington.edu> wrote:

> > >> > > >

> > >> > > > > As for other OSs, I haven't had the chance to try Debian. I don't like

> > >> > > > > burning 6 CDs just to install an OS (even 3 CDs kinda sucks in my

> > >> > > > > opinion). However, maybe I should try it.... :)

> > >> > > > > --Doug

> > >> > > >

> > >> > > > One CD. You can (and probably should to get latest versions) pull the

> > >> > > > rest from the net automagically via dselect/apt if you have a fast

> > >> > > > connection. One CD installs enough to get the rest.

> > >> > > >

> > >> > > >

> > >> > > > Or Knoppix, 1 heavily compressed CD Debian live version with rather

> > >> > > > more than the basics....or ClusterKnoppix (same deal

> > >> > > > with OpenMosix clustering kernel)...or even Quantian

> > >> > > > which is simlar but has various stat and math software

> > >> > > > built in.

> > >> > > >

> > >> > > > There are three versions of standard Debian around at most times:

> > >> > > >

> > >> > > > stable boring, dated and reliable (currently 3.0 aka woody)

> > >> > > > ..only security updates.

> > >> > > >

> > >> > > > testing Not cutting edge but generally useable and stable

> > >> > > > especially towards the end of the cycle

> > >> > > >

> > >> > > > unstable Adventure with every update!

> > >> > > >

> > >> > > > When "testing" moves to "stable", "unstable" moves to "testing" and

> > >> > > > a new "unstable" makes its appearance.

> > >> > > >

> > >> > > > --------------------------------------------------------------------------

> > >> > >

> > >> > >

> > >> >

> > >> > - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

> > >> > Cere Davis

> > >> > Unix Systems Administrator - CSDE

> > >> > cere at u.washington.edu ph: 206.685.5346

> > >> > https://staff.washington.edu/cere

> > >> >

> > >> > GnuPG Key http://staff.washington.edu/cere/gpgkey.txt

> > >> > Key fingerprint = B63C 2361 3B9B 8599 ECC9 D061 3E48 A832 F455 9E7FA

> > >> >

> > >> >

> > >> >

> > >> >

> > >>

> > >

> > >- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

> > > Cere Davis

> > > Unix Systems Administrator - CSDE

> > > cere at u.washington.edu ph: 206.685.5346

> > > https://staff.washington.edu/cere

> > >

> > >GnuPG Key http://staff.washington.edu/cere/gpgkey.txt

> > >Key fingerprint = B63C 2361 3B9B 8599 ECC9 D061 3E48 A832 F455 9E7FA

> > >

> > >

> > >

> > >

> >

> >

>

> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

> Cere Davis

> Unix Systems Administrator - CSDE

> cere at u.washington.edu ph: 206.685.5346

> https://staff.washington.edu/cere

>

> GnuPG Key http://staff.washington.edu/cere/gpgkey.txt

> Key fingerprint = B63C 2361 3B9B 8599 ECC9 D061 3E48 A832 F455 9E7FA

>

>

>


--
Peter Abrahamsen Seattle, Washington, USA
CEP, Phil, iSchool http://www.abrahamsen.com/peter/
University of Washington 206.525.2702 (home)
Consensus reality - did you vote for it?


More information about the Linux mailing list