On Fri, Oct 01, 1999 at 09:17:07AM -0700, longship@keyway.net wrote:
> Wouldn't it be nice is this information was collated at one location so
> that people could build on what's already been done and not have to try
> everything new every time. That's what open source is about--sharing.
> What I am proposing is to help collate the information, do the research
> so that people coming after me don't have to "keep up with the various
> lists". This is a no-brainer. It needs to be done and it needs to be
> done now.
Great. Go for it. You might like to get in touch with Joey Hess (who
does Debian Weekly News), since he's also collating much of the same
information for inclusion in DWN. A look through past issues would
probably turn up most of the important existing issues.
It's not that we're out to make life hard for you - if it's not being
done it's probably more to do with nobody having the right combination
of time, enthusiasm and skill to do it than anything else. In a
voluenteer orgainzation there are very few people with who the buck
*definately* stops, and most of them seem to be very busy already.
> > actually imagine this being a more serious problem these days with
> > things like GNOME having so many librarie to get right.
> Those are a nightmare. That's why it's important for somebody who's already
> got things working with Slink to document it and share with others so that
> everybody doesn't have to reinvent the wheel.
There is a Slink version of the last GNOME release - have a look on the
GNOME web site - and AFAIK the plan is to spin a similar set of packages
for the upcoming GNOME release. I think there's a link from the Slink
section of the web site.
[Perhaps stable is OK?]
> What's extreme is that hundreds of Debian users have stable upgraded systems
> and nobody has bothered to document it and post a How-To.
There is documentation for those upgrades I'm using on the web site
(actually, I'm not sure about the GNOME stuff but it's not precisely
hidden). There's also the APT registry someone already pointed out to
you - it really depends upon the enthusiasm and interest of the
developers.
> And yes, I could go without GIMP and Enlightenment and even X Windows, too.
IME GIMP from Slink works fine. Don't know about E - I'm more a wm2
(the window manager with no features but which looks very pretty) person.
> What your telling me is that when using Debian, I have to get used to using
> out of date stuff. I won't buy that when the solution is so easy--document
> what you've done and put it up on a damn Web site so others can benefit
> from your experience.
Documenting everything properly and consistently is non-trivial - there
are an awful lot of packages in Debian, with a huge number of
interdependancies. Currently, the choice is between stable (rock solid
but a bit old) and unstable (which may break but almost always works).
I imagine that part of the reason nobody's gone to the trouble of fixing
the problem yet is that for the sort of person who likes to run up to
date things, unstable is stable enough and nobody's thought it
worthwhile to make something in between yet. The name is something of
an overstatement.
There's nothing intrinsicly wrong with running old things, so long as they
aren't broken - some of the machines I help administer run SunOS quite
happily. We could upgrade them, but for the most part there's no particular
reason to do so.
> before updating my S.u.S.E. machines to Debian. My server will remain
> S.u.S.E. until potato is Stable--hopefully before the end of the year.
I admire your faith :-) . I'd say more like early next year (although
it should freeze in November, which may be good enough for you).
> > Otherwise, giving it a day or two before installing new packages and
> > paying attention to bug reports and the lists should help you steer
> > clear of anything really serious. Using apt, you can track just the
> > list of packages you need rather than the entire distribution.
[semi-stable]
> I will do that. I will also see what kind of interest there is on this.
Quite a bit of interest I think, although I wouldn't anticipate anything
happening until after Potato is out the door - the intention is to
freeze next month, and the best time to make changes would seem to be at
the start of the release cycle.
> I, too, was thinking that this would result in what may be called a
> semi-stable distribution. Basically it means having another chain of
> programs with updated contributions which include instructions on integration
> with Slink. Dependencies would have to be worked out to build .deb packages,
> but that will work itself out in testing, which can be a simple go/no-go.
I think the current thinking is more "unstable with a more blunt edge"
than "stable with a few fancy bits" IYSWIM. The intention is not to be
any more compatible with slink than potato is, but rather to have some
kind of testing requirement (eg, a time period without any bug reports)
after which packages could be moved into semi-stable once all their
dependancies were satisfied there. Trying to make everything completely
compatible with stable would be a lot of effort for questionable gain
(where do you draw the line about what's important enough to upgrade?),
and would eliminate one of the potential benefits - the possiblity of
freezing from semi-stable to give greatly reduced freeze times.
--
Mark Brown mailto:broonie@tardis.ed.ac.uk (Trying to avoid grumpiness)
http://www.tardis.ed.ac.uk/~broonie/
EUFS http://www.eusa.ed.ac.uk/societies/filmsoc/
Attachment:
pgpKDelKcDZue.pgp
Description: PGP signature