Re: amd64 into mainstream
On Monday 18 April 2005 6:31pm, Christopher Browne wrote:
> On 4/18/05, Ed Cogburn <firstname.lastname@example.org> wrote:
> > > I full agree here: Ubuntu is more attractive to the average end user.
> > > But I do not understand why everybody is so upset about this. After
> > > all, there is no "one size fits all" distribution.
> > No, I'm not saying Ubuntu will kill Debian or vice-versa, or anything
> > like, but I am saying that the more energy and momentum for a desktop
> > system Ubuntu takes from Debian, the longer it will be for Debian itself
> > to get its act together on the desktop, because everyone who wants to see
> > Debian on the desktop are now saying "Why not just use Ubuntu?", and are
> > moving to it.
> It seems to me that this misses part of the real problem.
> As far as I can see, the main merit of Ubuntu isn't anything to do
> with "desktop" support, but rather to do with the fact that it is
> several years more up to date than Debian/stable.
I don't know how many people are using Ubuntu on servers, but most comments I
see are about people using Ubuntu on desktops. A /. thread I saw some
weeks ago about Kubuntu was a fairly large discussion and the topics ranged
over many things about Ubuntu/Kubuntu/Debian but there wasn't anyone there
talking about Ubuntu on a server, the topics were mainly about Ubuntu being
the desktop platform that Debian itself has never bothered to create....
along with the usual GNOME/KDE flames. :)
> I find it ludicrous that Debian/stable still has PostgreSQL 7.2.1 as
> the "latest official" version of PostgreSQL even though there have
> been a whole ream of security updates and other fairly severe bug
> fixes in the binary compatible 7.2.x series. And despite the
> PostgreSQL project getting accused of having long release cycles, that
> doesn't even touch the fact that there have been three _MAJOR_ release
> cycles (7.3.x, 7.4.x, and 8.0.x) since then.
> And I usually _am_ something of a curmudgeon on stability of releases;
> at work, we never wound up touching 7.3.x because by the time we were
> ready to consider an upgrade from 7.2, 7.4 had been out for a while.
> And we're now just starting to _think_ about 8.0 upgrades...
> I am running Debian/unstable on my desktop pretty happily; I just find
> it painful that the "stable" release is so woefully out of date.
I agree. But I really see several problems here, not just one bogeyman we
can point to. The threads in debian.devel after the infamous meeting where
"Debian" decided to "get rid of" some lesser ports pointed to many different
problems in Debian's organization, most of which require different solutions.
The quotes above are used because I don't really know the true nature of that
meeting, only that it was interpreted that way by some DDs. The fact that
there is so much suspicion involved is another indicator of serious trust
issues, mainly between some DDs and the ftp masters of debian.org. That
alone is a major problem I hope Brandon can get to the bottom of. The
Server/Desktop problem is tangential but still related to the delay problems.
The example you chose to use (not being critical here - your example is a good
one - but also indicative of the Desktop/Server problem as well), a SQL
database server kind of illustrates the dilemma that a Debian-for-everyone
now has. The Server people want up-to-date SQL, PHP, and Apache, while the
Desktop people are clamoring for KDE, GNOME, nvidia drivers, OpenOffice,
games, etc, etc. Yet Debian's infrastructure is having a hard time keeping
up with everything at once, especially the package auto-building, and
*especially* on those older ports with (relatively) less powerful machines to
act as auto-builders.
Right now the current problems can be alleviated by not releasing some of the
lesser ports as often as, and not at the same time, as the major ports are
released, which is what's being discussed. What isn't being discussed is the
underlying issue that's created the current problem in the first place:
Debian trying to be everything to everyone only to end up dissatisfying many.
Ubuntu is proof Debian, in its current form, can't be everything to everyone,
and I think we generally agree with that, however....
> It may well be that there is room for two systems:
> 1. Debian, as the "grand collector of package updates," and
> 2. Ubuntu, as the folks that actually create release candidates on
> some reasonably regular schedule.
That is unacceptable to me for the same reason you chose Debian for your
Server needs. Ubuntu != Debian. That's why there's still people like me
hanging around with Debian even though we'd probably be better served in a
technical sense by moving to Ubuntu's platform. You want Debian on the
Server, we want Debian on the Desktop. Its beginning to look like Debian
clearly can't do both to both our satisfaction. Just as the Linux world
found out that Linus "couldn't scale", the Debian community I think is
beginning to see that Debian, as a single entity, can't scale either. So
telling one group to "go away" is not an option, both groups are a stubborn
bunch. :) Telling one group to go with a commercial, profit-driven company
when most of them came to Debian precisely because it was not one of those,
is also not an option. Solution: 2 Debians. Brother Server Debian, and
Sister Desktop Debian. :)
Let the supporting companies financially support whichever group (or both)
that serves their purpose the most, SPI would support both groups equally as
the umbrella organization, let the DDs drift to whichever group is doing what
interests them and/or keep a foot in both organizations if they wish, let the
2 groups go in different directions while still sharing in common what can be
shared (underlying base system, dpkg, apt, DD membership process, DFSG, etc),
let each group take the lead for the software most crucial to them with the
other group altering or extending the other's work according to their
specific need (just like Ubuntu is doing with Debian's base now), but let
each group make the different choices they need to make to satisfy the
diverging needs of their constituents. Debian is dead, long live The
(Note: I'm not accusing specifically you above of telling anyone to "go
away", I'm not accusing anyone of anything actually, its just an allusion to
the larger problem of the Server people not being interested in the concerns
of the Desktop people, and vice-versa.)
(Damn, sorry this got so long again)