Re: The Distribution

[George Bonser <grep@oriole.sbay.org>]

On Mon, 19 Jan 1998, Gregory Bell wrote:

> > Because it already exists.  We can START by adopting it and improve it as
> 
>  So does LST and a number of others, it seems to me that you want SEUL to dothe work
> for Debians benifit.

Or LST or a number of others :) Basicly the goal is to have them ALL based
on SEUL.  


> What's my beef, well really it comes in two parts.  If debian is so good, then
> whythere a need for SEUL??  If Seul takes on the programs of the past then it also
> takes on the misconceptions of those programs.  We have a chance to start fresh
> here, why limit ourselves???

Becauese every distribution so far is either difficult to install,
difficult to upgrade or is not well integrated.  Red Hat for example,
their linux ix easy to install but limited in the number of packages that
are provided in the distro and they are often broken.  Packages in the
contrib directory are often not well integrated with each other or the
main distro, their package selector does not inform you of conflicts and
dependancies until AFTER you attempt to install the package. Have you ever
attempted to upgrade a Red Hat system by FTP and had the FTp fail part
way through leaving your system in an unuseable state?  I have.

Debian has its problems too but seem to me to be the easiest to overcome
and are mostly overcome by projects that are already in work. These are a
confusing package selector IF the newbie gets thrown into the conflict
resolution screen without reading the docs.  This is usually caused by
changing too much at once during the initial install. The others problem
is basicly due to being too conservative in its configurations as to make
things so "plain vanilla" as to make it look like you are running on a
Timex Sinclair.


> But when you are attempting something new, relying on the old in many cases
> while it gets the job done, is not the best solution.  So what you end up with is
> a mish-mash of old ways of thinking with new ways appended on.
>

It is probably a need to be expediant.  To write the entire thing from
scratch might result in Linux changing faster than SEUL can be developed.
By adopting someting that already exists and changing it piece-by-piece we
have something that, at the start, is no worse than a good linux
distribution and improves as time goes by.

Debian's tools are pretty good.  As I write this, I am building a custom
boot disk that says SEUL rather than Debian/GNU.  I will then build
another using the 2.0.32 kernel rather than 2.0.29.  I will then build
yet another making even more changes.

 
>  How your horses, there buckroo.  Take a look at your basic system.  A processor(in
> this case an X86 chip), a video card (mostly SVGA), a hard drive (EIDE), memory,
> and a monitor (mostly multisync)  what's so difficult????  Remember this is for a
> first time user whose computer comes from a factory, no modifications.

We are not shipping a distro at this point aimed at computer OEM's Have
you seen Windows' install screen before it goes out hardware hunting?  How
about XF86Setup before it knows what kind of video card you have?  That is
what I am taling about.  A nice graphical INSTALL GUI that works on about
100% of the hardware out there. We will install the accelerated server
later on in the process.  

> 
>   True but unless things have changed, SEUL is not aimed at the millions of PC's
> outthere.  The target audience is, if memory serves the user that is currently
> running windows95
> and wants to try something different (for what ever reason).  Again point to at least
> a 486
> computer

Yup, and there are millions of Win95 users with computers more than 2 or 3
years old.  There are also the Win3x users that have NOT switched to Win95
simply because they could not afford a new computer.  These are great
Linux candidates.

>   And in those three days none other than yourself has said I can't continue UNTIL
> the nextversion of debian comes out.  And we know who your associations are with.

That is because release is due within the next few weeks, most
likely.  WHy devote a lot of time casting things in stone now when the
change to libc6 is so fundamental of a change?  If it was JUST another
release, you would be correct but would you invest a lot of energy in an
a.out release if you knew the world was going ELF within the month?

In this case, Debian has done it right with libc5 AND libc6 programs being
able to co-exist BUT as the libc5 packages are replaced with libc6
packages, the archive is changing daily.  There is just too much turmoil
right now ... that will settle down shortly.

>  Hummm I was never asked about this, neither was at least one other person thatI have
> had contact with.  For all intents and purposes you are attempting to push
> Debian down the throats of everyone on the SEUL list.


No I am not.  It was SEUL's approach to the Debian community that got me
interested in this project. When I read on teh mailing list that they had
decided to become Debian based, I jumped in with both feet. I have used
debian since before it's official release. I went through the ELF upgrade,
the 2.0 upgrade, and now the libc6 upgrade.  I have run several other
distributions of Linux starting with Slackware and including Red Hat,
Debian and two flavors of Caldera.

I would say the Debian is the most easilly configurable and best
integrated of any of the distributions that I have used and is designed
from the start to be the basis for other distributions.


George Bonser 
If NT is the answer, you didn't understand the question. (NOTE: Stolen sig)
http://www.debian.org
Debian/GNU Linux ... the maintainable operating system.