Re: [SLUG] Possible Project, Xandros, and Mepis

From: Robin 'Roblimo' Miller (robin@roblimo.com)
Date: Wed Dec 24 2003 - 11:44:33 EST


  IMO, hardware detection scripts
>> ought to be part of the LSB and/or managed by OSDL rather than "secret
>> sauce" for each distro publisher.
>
>
> What other IP to they really have? This is what has always amazed me
> about commercial Linux distributions. The installer is where 90% of
> their support calls come from. The more "secret sauce" they pour into
> the installer, the fewer support calls they will tend to get. Distro
> producers make their money on selling a supported package - if an
> OpenSource project came along that would remove most of the need for
> their valuable support, what differentiates them from their competitors?

What differentiates one car from another if they're both powered by
fuel-injected Otto-cycle gas engines? Why have both Cadillacs and
Chevrolets -- or both pickup trucks and sedans -- made by the same
company? Or more than one car company at all, since all their products
are based on the same engineering?

Let's start with a premise alien to many old-school Linux users and most
programmers, but well-known to auto marketers: The trim and paint are
more important to most consumers (corporate desktop users) than what's
under the hood.

Let's add another thought that goes against the traditional Linux grain:
Since most computer users never change default settings, the defaults
are all-important. (Choice is nice, but probably only 5% of all computer
users exercise it in any meaningful way.)

Combine these two premises, and you end up with the "perfect" user-level
distro being one that offers a number of theme/package/menu options out
of the box, and makes it super-easy to switch between them.

Now let's sell not only premake cute themed package choices that can be
switched with a single click, but sell convenience. Yes, you can usually
upgrade a Debian install with apt and have most things work after you're
done, but it would give 99% of the people out there more confidence in
Linux if they could pay a small sum - I keep coming back to that
$30/year figure - and get their software and updates from a single
company's secure servers, and leave the package testing to the company
that gets the $30, plus emails or other alerts every time there's a
security update or something new/cool available.

No one, I repeat, no one, has ever come out with a total end-user Linux
package as a true consumer product. Michael Robertson has come closest,
but his personality and business methods have screwed it up.

I've been covering the Linux "beat" continuously as a journalist longer
than anyone. I was unquestionably the first non-geek journalist to use
Linux full-time, and I was one of the world's first point-and-click
Linux users. When I say, "This is so," or "That won't work," My
statements are based on long, careful research and observation.

So I will state firmly: There are many ways for Linux distributions to
differentiate themselves from each other that they haven't tried yet
because they have been so wrapped up in duplicating each others' work on
utilities.

I will now bring in another industry analogy: Consumer electronics. Have
you ever looked at how *few* chipsets are used in the majority of TVs,
VCRs, DVD players, and other common devices? Many of them are
essentially identical inside, despite wide variances in external
appearance -- and in price.

Still, there are many companies all selling these things under their own
brand names -- and most of them seem to make money at it most of the time.

Why in the world shouldn't Linux distributions publishers work on
selling a *finished product* along with *value-added services*?

My wife's new Hyundai didn't come with Hyundai-developed tires. It came
with Michelins. Shouldn't Linux distributions all use the same tires, as
it were, and differentiate themselves in other ways?

- Robin

-----------------------------------------------------------------------
This list is provided as an unmoderated internet service by Networked
Knowledge Systems (NKS). Views and opinions expressed in messages
posted are those of the author and do not necessarily reflect the
official policy or position of NKS or any of its employees.



This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 19:56:55 EDT