Re: obsolete code must die

[Horst von Brand <vonbrand@inf.utfsm.cl>]

Claudio Martins <ctpm@vega.net.dhis.org> said:

[...]

> Anyone with a <486 probably shudders at the space and time requirements
> of compiling modern kernels..
... which they do on their newer machine anyway, as the i386 in the closet
is a router that sports no compiler.

> The older boxes don't support the new hardware anyways..
... so support for old hardware has to stay (unless it hurts somewhere,
which is rather uncommon)

[...]

> But in this age of 'disposability' more and more people just accept they
> have to buy new hardware every 3-5 years.. For those that want to
> maintain Linux on that, so be it..

There are places in the world where a i486sx is viewed with awe... and i386
are commonplace. Don't fall into the trap of "All Linux PCs are in the US".
Besides, if the piece of junk does its job well, why buy an expensive,
completely overqualified new machine for it? Sure would make PC makers
happy, but I'm not _that_ interested in their happiness to tell the truth.

> Maybe we need more Alan Cox's, and then we could have sperate kernel
> trees, Linus's which is the mmmmmother.. (HI MOM!) and the pre-pentium
> tree, the post-pentium tree, the embedded tree etc..

... and a nightmare where nobody knows the state of <favorite driver> in
<targeted kernel tree> and has to cross-port all kind of stuff to get
<random machine> to work.

> But in reality, with all the people contributing now to one tree, there
> is still more work to be done.. Who wants to split those resources?

Bingo!
-- 
Dr. Horst H. von Brand                       mailto:vonbrand@inf.utfsm.cl
Departamento de Informatica                     Fono: +56 32 654431
Universidad Tecnica Federico Santa Maria              +56 32 654239
Casilla 110-V, Valparaiso, Chile                Fax:  +56 32 797513