Dropping pre-i686 from the archive

[Eric Dunbar]

> On 4:07:19 pm 12/01/2006 MadMan2k <madman2k at gmx.de> wrote:
> > to an 64bit system that for a library in the third world to get i686
>
> This is a fair point, however, as people keep saying, it's not just third
> worlders and schools that are running <686 CPUs - anyone running a Via C3
> processor lacks the full i686 instruction set and that is anyone running a
> mini-itx board and probably some others.
>
> This instantly invalidates any arguments about <=586 being so old as not
> worth bothering with, surely? These are CPUs/machines that are on sale
> *now*.

As for the speed argument...

Why doesn't someone with the wherewithal to compile various packages
do a benchmarked comparison of packages compiled for i686 vs i386?

It shouldn't be the hardest thing in the world and it seems that the
web is devoid of meaningful (easy to find, at least ;-) comparisons of
i386 and i686 comparisons. So far, all I've seen in this thread are
anecdotal statements regarding speed (and, given that FireFox and
Epiphany both have been called the faster browser...).

PS Do "Via C3" CPUs not run i686 compiled packages? (I know nothing
about the Via C3) Also, how commonly are these Via C3 CPUs used? Are
they in embedded solutions or in (formerly) high-end machines? If so,
then what's the problem with _not_ supporting them out-of-the-box, IF
i686 packages really do result in a [statistically] SIGNIFICANT and,
more importantly, NOTICEABLE speed up for all i586/i686 compatible
machines (which seem to be nearly all Pentiums faster than 200 MHz)?

Just throwing out some questions.

Eric.