On Fri, 16 Apr 2004 11:40:19 +0200 "Simmel" <[EMAIL PROTECTED]> wrote: > > So why not think about using a strategy that almost every company uses > (although Debian isn't one), e.g. Redhat, SuSe, even > Microdoft........... For me as a user and systems administrator > something like this would be much much better. > > Why not do it this way? > > enterprise - this is for servers only - not much GUI/ focused on > servers/ networking,routing/ multiple cpus/driver support and so on > > workstation - this is for home users and workplaces - not much server > stuff here/ focused on multimedia/ x-server/ openoffice and so on > > sandbox - (I like that word, Monique :-) this can stay the same and is > meant for people who would like to help the Debian project with further > releases, simply a sandbox to play with to find and report bugs..... > (maybe there should be two then, something like E-sandbox, for the > enterprise stuff, and W-sandbox for the workstation part)
[ snip ]
In making suggestions like this, and like many others in this thread,
the implicit assumption is that the reason the three distros (stable,
testing and unstable) exist is so that users have a choice of distros,
and can then choose the one that suits their needs best. With that
assumption, it is of course very important that the explanations be as
clear to users as they possibly can; and it would make sense to even
consider structuring the contents (and thus titles) of the distros
differently, so that they'd better achieve that goal of giving users
choices.
But this assumption is wrong. The purpose of the existence of testing
and unstable is *not* to give users choices. It may also be true that
their existence gives users choices; but that's not what they're
fundamentally for. The purpose of their existence is to facilitate the
development process that produces stable releases. Users may decide to
track unstable or testing (and many of us do); but the existence of
those distros is to help the developers do what needs to be done to get
packages into good shape and get releases out. Period. And thus,
the most important thing is that the descriptions of these distros
be clear to developers, and that their functions be useful for
developers. "Re-branding" the distros, and changing their descriptions,
isn't sensible: testing and unstable don't cleanly fall into categories
that are sensible for users, and trying to label them that way is (as
Monique said) trying to assign characteristics that don't exist. But
that's not a bug; that's a feature. It's intentional. Their purpose
is to facilitate the job of the developers.
Looked at this way, the problem with your suggestion above is that
it doesn't accomplish the goal of facilitating the next release. It
tears down some of the infrastructure the developers use without
replacing it with something that helps them do their job as well or
better.
"Well, OK," you may be saying, "but what's wrong with giving users
choices as well? Instead of having unstable, testing, and stable, why
not have unstable and testing to help developers (and helpful users),
and also distros like `server' and `workstation' and so on?" But this
is a false dichotomy: users already have the choices that other
distributions provide with such focussed releases. Put another way,
what people are concerned about in this thread is getting more recent
versions of packages than stable provides; creating "server" and
"workstation" releases of Debian like other distributions do wouldn't
solve this. "Server" and "workstation" releases of other Linux
distributions don't typically differ in the versions of the packages
they provide. Rather, they differ in *which* packages are provided.
The "server" release may include apache but not frozen-bubble; while
the "workstation" release may include frozen-bubble but not apache.
But they'll both typically have the same version of X (provided the
server has X at all, of course). So functional releases wouldn't
obviously address the thing that people have been concerned about
in this thread -- more current versions of software.
Furthermore, functional releases would go against Debian's philosophy.
Perhaps you want a machine with one or the other, or perhaps *both*, or
perhaps *neither*. Rather than decide for you what you'll need, Debian
lets you decide. The purpose of tasksel is to make that a little less
onerous, so that you don't have build your system up entirely from
scratch; you can select "Web server" or "Java development" and get a
bunch of packages relevant to whatever you need. But you can install
whatever you want. Want server packages? Install them. Want
workstation packages? Install them. It's your machine; do what you
want with it.
> P.S.: And while I'm on it, pleeeeez enhance the installation routine,
> something like a graphical interface. This takes the fear off most
> users. Take a look at SuSe and Redaht and you'll know what I mean. I
> know that there are also a lot of small things which aren't good, like
> the package selection, those are far better in Debian. But the "blue
> screen" :-P is really annoying and confusing. My first installtion were
> more like 3 1/2 installations, if you catch my drift.
Sarge will have a new installer. But that said, most of the developers
have traditionally found having a GUI installer to be very low priority,
for the simple reason that with Debian, you should only have to install
once per machine, ever. The painfulness of the Debian install process
just isn't something the typical user has to face very often.
-c
--
Chris Metzler [EMAIL PROTECTED]
(remove "snip-me." to email)
"As a child I understood how to give; I have forgotten this grace since I
have become civilized." - Chief Luther Standing Bear
pgp00000.pgp
Description: PGP signature
