Subject: Re: just a thought... inspired by "CP/M disks for apple IIgs"
Path: lobby!newstf02.news.aol.com!portc01.blue.aol.com!news.maxwell.syr.edu!newsfeed1.uni2.dk!uio.no!news.kth.se!tybalt.admin.kth.se!merope.saaf.se!not-for-mail
From: pausch@saafNOSPAM.se (Paul Schlyter)
Newsgroups: comp.os.cpm,comp.sys.apple2
Date: 24 Sep 2000 09:57:54 +0200
Organization: Svensk Amat|rAstronomisk F|rening (SAAF)
Lines: 55
Message-ID: <8qkc62$dp$1@merope.saaf.se>
References: <39c6cfdc.1625904@news.planetc.com> <39CD5DDF.3A6B2450@acronet.net>
NNTP-Posting-Host: merope.saaf.se
Xref: lobby comp.os.cpm:28457 comp.sys.apple2:106235



In article <39CD5DDF.3A6B2450@acronet.net>,  <dkanecki@acronet.net> wrote:
 
> The 65816 uses a different register set... CP/M was orignially written for
> an 8080 with a register set of PSW/A, B/C, D/E, and H/L or M as the
> combination  8 and 16 bit registers. The problem with 65XX and 65XXX
> series is that they have few registers -- mostly and A and B register and
> an X (index or memory register)
 
This could easily be compensated for by using the 6502 Page Zero as a
set of "extended CPU registers".  After all, the 6502 could access
Page Zero in 3 clock cycles, while the 8080 needed 4 clock cycles to
access a CPU register....
 
The REALLY BIG problem with the 6502 was something else: the CPU stack
which was limited to a mere 256 bytes.  This was a real bottleneck
for all software using more than a very small amount of stack.  And
it's because of this that AppleSoft Basic had much more severe limits
regarding nested subroutine calls, nested parantheses in expressions,
etc than it's "CP/M cousin" M-Basic.
 
> -- So, port CP/M you might want to do want some people did -- write a
> virtual 8080 that run on a 65XX or 65XXX, then you can port the code
> to the system ---
 
That's emulation, not porting....
 
> When the Apple first came out, that was reason for the two basics -- one
> basic used a virtual processor to get around some things with the 65XX
> series processor -- That is what made Woz a hero!
 
I'm sorry, but this is a myth.  Yes, Woz designed and implemented
"Sweet-16 - The Dream Machine" as well as Integer Basic, however Integer
Basic was implemented in 6502 assembly language, not in Sweet-16 code.
Remember that at that time, Integer Basic was faster than most other
Basic interpreters -- if Woz had implemented it in Sweet-16 rather
than 6502 code, Integer Basic would have been some 10 times slower.
 
Another piece of history: Woz implemented Integer Basic directly in
6502 machine language -- there isn't and never was any official
source code for Integer Basic !
 
 
> My first system was an Osborne 1, but I floated between the OS-1 and Apple
> (mainly using logo and Grafforth)...
> 
> I hope this helps...
> 
> D. Kanecki
 
-- 
----------------------------------------------------------------
Paul Schlyter,  Swedish Amateur Astronomer's Society (SAAF)
Grev Turegatan 40,  S-114 38 Stockholm,  SWEDEN
e-mail:  pausch at saaf dot se   or    paul.schlyter at ausys dot se
WWW:     http://hotel04.ausys.se/pausch    http://welcome.to/pausch