RE: SEUL: About Independence, LaetOs, RedHat and Debian

[Al Hudson <eah106@york.ac.uk>]

On Fri, 9 Apr 1999, Bill wrote:

> Has anyone noticed how much money Redhat is getting in the form of
> investment from traditional Windows industry and various closed-source
> players?  The WinTel marketplace has one thing that people are seeking.
> MONEY!  Now before you start throwing things at me, listen.  Money corrupts
> yes, but it also empowers.  It pays developers, it pays for marketing, it
> pays for government lobbiests, it pays for advertising during the SuperBowl,
> and it tells regular peaple that something is serious.

Hmmm, but you have to ask yourself why they got there - a lot of people
would have said (years ago) that Lasermoon were going to produce 'the'
distribution - all the stuff about POSIX compliance, etc., which noone
else had. RH seem to be doing well because they're offering a set of
standards (e.g., rpm) - a lot of the rest of the linux world is trying to
build powerful systems based on diversity - i.e., by getting something to
run on many systems & many platforms you can be fairly sure the code
you've written is pretty good, or flexible at least. RH seem more
'Windows' in that respect than the rest of the linux community, and I
think there are two fundamentally different approaches there. Should the
argument be, therefore, something more along the lines of should we build
for X distro versus should we build for all distro's? I would have thought
that it would not be a massive hurdle to overcome all the differences, and
producing software that 'unifies' (in some respect) a front end to all
distros (such as alien, if you see what I mean) would possibly be just as
useful as something that made installing / updating a system a breeze?
It's one of those tools that's reducing the learning time, especially in a
multi-distro environment (fairly unlikely, tho', I would have though ;)

One last thing about RH - I think I read in various places, probably /.,
that they're opposed to LSB - is this a good thing? I've never looked into
LSB personally, but I would suggest that (if my understanding of LSB is
correct) a standard set of libraries would be a good thing? Not
necessarily installed, but if a user has downloaded foomachine.tgz, or
something equally imaginative, and uncompacted it & stuff, instead of it
saying 'I need nastylibrary.so.1', or something equally terse, it should
say 'Pop the CD in and I'll sort it out while you have a cuppa'. Being
able to assume things about systems must make things easier?

It's not necessarily just a case of making this distro easy to use;
really, this type of software should be being written anyway, distro or no
distro. I mean; GNOME doesn't replace window managers or X, but you never
hear people saying 'I'm running GNOME/Enlightment' or something like that,
they just say 'I'm running GNOME'. They know that having GNOME installed
entitles them to access a certain toolkit; perhaps that's we seek to
achieve here? If someone says 'I have [such-and-such easy distro]
installed on my computer', the base doesn't necessarily matter insomuch
that the finished result would be intimately tied to any distro? 

My $0.02's worth anyway ;))

Cheers,

Alex.