Hi Greg,

  > I've done a fair bit with PICs....

They look interesting, but I've never done more than skim through the data
sheets.

  >....The PIC bank selects regions of RAM, though any ram location
  >can be used in any instruction- is the Z8 arranged similarly?

No, the Z8 treats memory as a single contiguous 64K block. It's been a few
years (5 or so) since I last touched the Z8 but, unless my memory is
seriously failing me, the instruction set was tailored towards being
register oriented (you had 16 "banks" of 16 8-bit registers (or was that 8
of 8?)) so most operations had to be performed on registers.

  > As I recall, the only thing I didn't like much about the 6502 was
  >the small # of registers....

Yes, coming from the Z80 that was the main thing I disliked too, but it
wasn't long until I discovered why "zero page" addressing was invented....
;-)

  >....I really like the indexed addressing- but the hardwired stack
  >is somewhat objectionable.

I must admit the fixed location of the stack never really bothered me that
much. It was strange, but kind of made sense in the overall scheme of
things....

  > OK, what is IYKWIM?  ;)

If You Know What I Mean.

  >....the x86 indexed addressing, though half the trouble was
  >Intel's rather loose syntax....

Absolutely, I've always found Intel's ideas about assembly language rather
difficult to get to grips with.

A prime example is i8080 vs Z80. Even though they're VERY similar beasts
from the programmer's perspective, I found Z80 assembly language
CONSIDERABLY easier to learn. Zilogs' mnemonics and syntax struck me as much
more logical than Intels'.

  >....funny, they even added a "bugs" option which caused their
  >assembler to duplicate masm's problems...

Yes, I remember hearing about that a few years ago....brought a smile to my
face.

  > I wonder about the "savings" and "simplification" gained by the
  >use of C/C++ over assembly in embedded work sometimes....
  >....doing the nasty critical routines in C/C++ can be asking for
  >trouble. Using C/C++ immediately requires more/faster hardware....

Absolutely, one of the companies local to me now use 68K family processors
where before they were using 8031 family chips. Another is using i486s and
although the system is simpler and more powerful than the old one again
they're doing all the low level stuff in "C".

  >....I think the hidden cost is that the more plentiful C/C++
  >programmers also don't know very much about the realities of
  >making hardware go....

Although that's a sweeping generalisation it's actually pretty true in my
experience. This isn't their fault though, rather the fault of their
teachers.

When I went to polytechnic in 1983 we were taught COBOL, PASCAL and
microprocessor systems architecture - with an emphasis on 6802 and Z80
systems. These days, as I understand it, there is little importance placed
on the systems architecture part of such courses.

  > Nothing like holidays for soaking up time better spent
  >elsewhere... ;)

Heh, heh....

  TTFN - Pete.

--
Spam deflector in effect, edit reply address - it's obvious!

**********
Join "vectrex_dev" today, the new mailing list for Vectrex developers.
Go to to join.
**********

Hardware & Software Engineer. Sound Engineer.
Collector of Arcade Machines, Games Consoles & Obsolete Computers (esp DEC)
--

I guess a more highly orthaognal architecture with fewer protruding
uglies would be in better taste- but with decent compilers the need
for it is somewhat less.  Frankly, most of the x86's richness is
wasted on me at least- to be sure theres all these wonderful
instructions to do all sorts of esoterica, but I always seem to be
playing around getting data into the right registers.  I do like the
architectures with lots of registers, even on RISC style hardware, I
think it could frequently save code space simply because there are
fewer constraints on the dataflows.

<<shudder>>, I'm doing some 8051 assembly now plotting points on a LCD
graphic panel, pixels packed into bytes, mapped to the display
controller.  Then throw in the FP to get decent diagonal lines.  Eww.
Its a home project so I don't have a decent compiler, just assembly
and Basic on a 12mc 8051AH.  My current nightmare is the text mode
output, which is supposed to be independent of the dot graphics is
indeed printing properly, but the text data is appearing as garble in
the graphics display.  Something about the address space assignments
of the text and graphics memory banks I'm pretty sure- but where where
where...

But the 8051 has external ram, our device has only internal RAM and
EPROM- so we're stuck at 8k code and 300 bytes or so RAM.  Don't some
of the 8051 clones/derivatives make it up to > 40mc or so?

Well, when the Emacs people start talking about the OS acting as their
device driver, there must be prodigious wonders under the hood...  One
of these days, I'll run into somebody who does AutoLisp programming
(AutoCAD's Lisp implementation)- I'm really curious about the design
approaches taken to get Lisp code to draw stuff.

But surely all those new features make it ever so much better, right??
;)

Its a 2 pass macro assembler, the first generates assembly and the 2nd
feeds it into an assember.  An optimizer could remove the redundant
bank selects, and the assembler would fix up the references, but the
damnable PIC architecture will cause trouble.  It divides its 8k
address space into 4 2k pages, but code cannot flow over the page
division- the boundry must be jumped over.  The 1st assembler pass
watches for this, bombing the compilation if the IP passes over a page
boundry.  Thus I can't generate assembly output that exceeds its
address space, and then be optimized back into place.  So, I think I'm
stuck.  The assember people (Microengineering Labs) are very
responsive, I asked them about it and they are very aware of the
efficiency issues.  As I'm sure you imagine, they have a list of
things to deal with in each release, but they say the efficiency will
improve (it has to some degree the past several releases).

So roomy, and comfy!  So what do you need all that space for, a
postscript interpreter or something?  ;)

Thats such a tough issue- I think its lots tougher than the same
scenario with regular PC software; its so much harder to understand
whats happening inside the processor, you often can't be sure if this
chunk of odd code is actually doing something important- either for
the app itself or the processor or the peripheral chips or....  Then
you go clean up the code- make it more efficient you know, and bring
the whole thing to a crashing wreck.  I'm glad I won't be the one
dealing with that little nightmare.

Gregm

I didnt know that. Ive had my account with Ebay for quite awhile so I never
had to do that.  It is a great way to reduce the scammers!

Don

...

Hi Kevin,

Do you have permission to post the entire contents of every issue of
Creative Computing, or just selected articles or sections? Do you have any
idea how copyrights last on magazine articles? It might be near the end of
some of the earlier issues anyways.

One possibility for those magazines that you were unable to contact: Scan
them and post them anonymously in binary news groups and place them on free
web servers.

-William R. Cousert
wrcous@yahoo.com

I like that "Mostly no one cares."  What about the ones that do care?  I
have around 2000 disks full of software that use to be sold for around
$9 per disk (in the day).  All of the disks came from Lotsabytes
Software.  The guy use to sale with a ad in Antic.  I bought the
software and have been think about putting all of the software on a CD
ROM or 2 whatever it takes.  A collection like this shouldnt be kept by
just one person.  I feel I should get this software to everyone that
would like it but, getting the okay from all the programmers would be
impossible to do.  So do I just make the CD(s) and hope nobody ever says
anything.  And what does one do when the copyright holder says
something?  Can they sue you?  What legal action can they take?

Don

On Fri, 14 Jan 2000 07:55:56 -0000, "Richard Cortese"

I would bet there was no legal action to "force" the website to take
down the Ultima IV image, if that is what you are saying. I think
TxResumes is correct that the copyright holder must defend the
copyright, in other words Atari 8bit copyright holders can't come back
after <ahem> 20 years of letting those rights gather dust and just
pounce on fans who kept those titles alive on their web-sites. Not
that anyone is doing that.

I can't. I think we are talking about American copyright law for many
of these 8bit titles.

Did he get them? I recall reading that many early Tonight Show with
Johnny Carson tapes are gone forever because the network taped over
them.

In light of that, the activities of 8bit enthusiasts posting disk
images etc. safeguards the programs against being lost forever. To
prove that I would remind you that Phillip Price who authored
Alternate Reality once posted on this news-group asking someone to
give him a cracked copy of his game.

No, you are the one who is wrong. B&C do not offer Alternate Reality
or AR the Dungeon. They have a long text of titles but it is unclear
if the titles have the box and docs, or are they just pulling disks
from a stack of old floppies. They appear to be offering second hand
software which would have a whole different set of copyright
implications I should think!

Yeah, but all you have to do is demonstrate a reasonable effort to
protect your copyrights over time. Then you are covered.

VS
To email me just remove the Xs and underline marks from my email address.

My glass typewriter shows Velcro SPY typing...
[Snip]

        No, that is a trademark.

--

Fortran Dragon -==(UDIC)==-
Hidalgo Trading Company: <>
rgcud FAQ: <>

I suggest you spend a bit more time learning English grammar and spelling
and a bit less time whining because the users in this group are not
prepared to send you "sympathy" computers.

jkp

Use search then scroll down to BY ITEM NUMBER input number the click on
enter.

Don

thats true..

a peson i knew few years back (konp) from polin wrote me that
he is working on an 8bit ver...

i also saw a doom like demo few years in one of the demos party
collections....

nir

...