Date   : Sun, 24 Jul 2011 19:37:16 +0200
From   : rick@... (Rick Murray)
Subject: Risc PC (Was 'Minitel in France')

On 24/07/2011 09:41, Tom Walker wrote:

> Don't forget the stupid memory restrictions.

Ah, the MEMC's 4Mb limitation. It wasn't so big a deal when it was 
designed, but it was painful towards the end of its life.

Thankfully it was quite viable to use a fully-blown desktop publisher on 
a 2Mb machine with only a floppy drive... Okay, it would be problematic 
if pictures were required, but hey, try getting that far on a PC. ;-)


> It was a bit better than an ARM3'ed A300, but not that much better.

I remember my A3000. Floppy access and high-res display, the screen 
would blank during disc accesses. Hehehe...


> Most games failed.

Figures. Most of the software I used worked just fine between OS 
versions. Ovation, the only time that fell over was the protection 
system on the installer. But then that did odd things. Ovation itself 
worked fine.
Likewise, OvPro has worked for me on A5000, RiscPC, and RedSquirrel. 
Ditto for a lot of other stuff.

Games, on the other hand, take a lot of icky tricks to eke out maximum 
performance.


> Acorn never really gave that much attention to backwards compatibility.

Really? I think Acorn bent over backwards to emphasise USE LEGAL 
INTERFACES. Hell, BBC BASIC V even picks up and works with CALL &FFFx 
for OSBYTE, OSWORD, etc.

If, on the other hand, you start coding making assumptions, or bypassing 
the legal interfaces, then you're on your own.

Example. For my PtrIIC module, I tap into the parallel port and timer 
(Timer2?). I use some minimal checks (Parallel_ReadAddress or somesuch), 
and then I bash the hardware directly. However it is on the 
understanding that it may not work. But, actually it works on both A5000 
and RiscPC (different OS, different IOMD/IOC, similar combi-IO chip), 
but not others because of technical limitations (A3000, logic-chip 
parallel port; Iyonix, none...).


> Then of course you've got the endless compatibility problems when the
> StrongARM came in.

Again, these were, I think (distant memory) self-modifying code and the 
changes within the processor. How much "legal" code would have worked, 
and how many trickery-things would have failed?

Look at the change from 26 bit PC+PSR to 32 bit PC and PSR. When you 
look at it, the change was not really hard, just fiddly and very very 
tedious.


> and more the reality of them not being ARM's biggest customer anymore

Or perhaps ARM also seeing the future better than Acorn could.


> (and of course ARM screwing up the ARM8).

Howso? [just asking...]


> Exactly the sort of thing the OS should have trapped and worked
> around, given this was hardly an unknown technique.

Hardly an unknown technique, but never a legal method. I, personally, 
agree with you, given that it often stopped the machine dead in its 
tracks, that it should have in this instance trapped it.

That said, writing rubbish to &0 could mess things up - and how long was 
it before that was write protected from user mode software?


Best wishes,

Rick.

-- 
Rick Murray, eeePC901 & ADSL WiFI'd into it, all ETLAs!
BBC B: DNFS, 2 x 5.25" floppies, EPROM prog, Acorn TTX
E01S FileStore, A3000/A5000/RiscPC/various PCs/blahblah...