LinuxSA Mailing list archives

  From: Mike Andrew <mikero@norfolk.nf>
  To  : linuxsa@linuxsa.org.au
  Date: Wed, 28 Jun 2000 16:42:54 +1130

Re: More space required

On Wed, 28 Jun 2000, David Drury wrote:
> Hi,
> 
> The point you've missed about the IBM PC architecture, and one that it
> excelled at was that it was cheep. IBM was faced at losing a sizable

[snip]

I agree with your comments, except one. The ibm pc was not cheap,
it was to IBM's chagrin that klones bought the price down.

So that my comments aren't extreme, the design of the PC at the time
was (in my opinion) excellent. Yes, it had flaws, but as a desktop
machine it was overall better than anything else out there.  Perhaps like
you, I had various flavours of MC6802/9/Z80 style desktops at that time, all
were good, but the pc had less overall flaws.

However, as you'd agree, there are remaining errors in the design that make it
a hazard, and yes, these are largely due to being hamstrung by backward
compatibility issues. We could nitpick about whether they should have at
some stage introduced a new serial uart. But the overall problem was 
the  0x86 architecture. Where you could intoduce another type of uart , or
indeed a different hard drive (scsi vs rfm vs ide), replacing a cpu would be no
joke. (I don't know the reasons why the Motorola/IBM cpu never took off, it was
more than capabale of running in Intel mode if it had to. Real pity, it was a
very fine chip, from the beginning. (on that score Apple have a lot to answer
for, their astonishing greed screwed Motorola out of the desktop market)

But, the emphasis now is misdirected (market hype), it is relatively easy to
push the speed barrow because the knowhow is there to endlessly reduce the
micron density for some years (and hence improve speed). But this isn't new
bigger 'n better or improved technology, it's just a non-risky return
on investment. Pity is, the effort hasn't been placed into multi processor, as
in matrix, as in 64 animals at a time. Perhaps as far back as twenty years ago
pc magazines were full of interesting subjects (transmeta matrix signalling,
Mitshubishi try new vectored dma, Amd intoduced Harvard architecture to 2096,
Motorola and Fairlight  attempt anti clock phasing to improve speeds), Now, all
we read about is the breathless excitement of pouring liquid nitrogen over the
cpu.

There's almost complete myopia about improving what's there already. The push
into AGP is inexcusable when you consider just a tiny effort would have removed
the IDE / LBA problem some years ago. But that area isn't glamourous and not as
easy to hype in the shops as an apparently 'better' display. While we now have
Voodoo III's (mistakenly identified as some miracle of the AGP bus), we *still*
cannot get decent ide bandwidth and only now are looking at UDMA66 et al.

The devastation caused by Microsoft to the software landscape, where it was
in nobody's interest to innovate, is mirrored by the Intel monopoly. In both
cases, the effects have been quite subtle because there is always apparent
improvement (faster cpu's, win95/98/2000). But the reality is computing could
have been 10 years further advanced if Wintel were in a level playing field.

The nix landscape has, in my opinion, been equally bleak, because it was
starved of market. It doesn't say much of an operating system that essentially
uses the same interface and commands of 23 years ago. Like how many of us are
fascinated by VT100's ? Most have never seen one.

Fortunately, as you point out, the hardware is cheep, and there more
refugees flooding into this alternative OS because they are starved of the one
thing that motivates them more than a salary :- the chance to
experiment and innovate. Hence, Linux will exponentially improve. If you look
at something like Kdevelop, it's nothing stageringly new, but, it is staggering
that a product of that quality has just 'appeared' out of the blue. More will
follow.

Finally, to get back to the original theme, the idea of even attempting to port
Windoze to alternative platforms is a nightmare. Whereas, Linux from the
beginning shows no favours. It is this, that will eventually lead to other
hardware platforms and alternatives finally taking off.

   -- 
http://linux.nf/stepbystep.htm + mirrors StepByStep submissions:
mikero@norfolk.nf


-- 
LinuxSA WWW: http://www.linuxsa.org.au/  IRC: #linuxsa on irc.linux.org.au
To unsubscribe from the LinuxSA list:
  mail linuxsa-request@linuxsa.org.au with "unsubscribe" as the subject