Re: What are your obstacles to running CP/M?
- From: Herb Johnson <herbrjohnson@xxxxxxxxx>
- Date: Fri, 4 Apr 2008 10:05:45 -0700 (PDT)
On Wed, 2 Apr 2008 17:44:07 -0700 (PDT), Bill H <b...@xxxxxxxxx>On Apr 3, 9:01 am, no.s...@xxxxxxxxxxxxxxxxxxxxxxx wrote:
wrote:
I think one of the biggest issues would be with hardware. If a "noob"
comes accross a complete CP/M system they would probably fall into the
boot disc catagory. But if they just have some parts and want to put
together a complete system, the "noob" would need help knowing what
parts go with what [memory card, CPU, etc.]....
With the PC over the bast 20 years you could basically take any PC
card and plug it in your system and have a 95% chance of it working
with little more than putting in the driver disc....
I think a great resource for CP/M would be a large cross referenced
database that shows what cards work with what hardware with what
jumper settings and / or bios changes etc.
I'd agree save for it's near impossible to build. There is a lack of
information and lack of resources to test it.
The matrix for the table would have to include:
[followed by a page of features of computers of the 70's: case,
power supply, CPU, memory, I/O types, file storage.]
For S100 system those are the major elements an integrator would
grapple with. For all of those there are compatability,
interoperability and revisions. [Plus age of components, plus lack
of available documentation.]
These posts make the point, that many CP/M vintage systems are not
really "systems" as is thought today. They are collections of
hardware, cards and chips. In the S-100 world, anyone could pile any
set of cards together - dealers or end users. That was the point of
S-100! It was difficult but possible to make some sets work, some not.
CP/M (and S-100) were the "glue" to make it possible, and even then it
may not have been enough. Hey, some of those boards did not work, even
when first sold! But many did. Today, a "nubie" may get a set of cards
with no knowlege, by buyer OR seller, if they will work together, if
they work at all.
The problem of making these work today is compounded, as Allison says,
by age of components and lack of docs. Well, most of the docs ARE
available - I should know, I have them, I make them available. Others
do too, for free on their Web sites.
But there are more subtle issues which are not outright documented in
the manuals; those were discussed in the magazines and books of the
time, or later in email discussions. Dynamic RAM boards were not "good
players" in the S-100 world, many designs worked poorly, some failed
when combined with other hardware. And age makes memory chips
"forgetful", RAMs lose bits. Even ROMS lose bits after a few decades.
Certainly, one issue for "nubies" is that it's a lot of work to dig up
documented information. But the Web and time has mostly solved that.
Another issue is the technical challenge to understand basic CP/M
operation and requirements. But a deeper challenge is to understand
chip-level designs, and to know why some combination of hardware may
have problems - NOT because the manuals says so, but because the
digital logic between the chips or boards is flawed. Timing errors,
signal delays, and incompatible signals between S-100 cards through
the bus. There are similar problems between floppy controllers and
floppy drives.
One can avoid some of these issues by only working with single-card CP/
M hardware; or use a well-established set of S-100 cards (from
Northstar, Compupro, etc). Or, as Udo Munk provides, use an emulator
and avoid old hardware all together. But some people insist that they
want to make a "fresh" or "from scratch" hardware CP/M system.
In that case, Allison and Bill have outlined that class of issues. But
there is no general "fix" for those problems, no list of "here is what
will work", because there are just too many ways cards and hardware
can be combined. There is no manual or list, and there never, ever
will be.
One has to do what they did in the 1970's when that situation was new.
They have to know the hardware, down to the chip and bus. It's a
considerable challenge, for some. For others, it's one they don't care
for. A "nubie" needs to know this is what they are in for, so they can
make a decision about these challenges.
But I see some people, which seem to believe they can be talked
through these technical issues. If they ask enough people, get enough
discussion, they will get "help" and they can make their stuff "work".
What that does, is pass these problems on to someone with more
knowledge - they have to face those same issues. With luck, or with
some changes to hardware used, there is some success that way. I
suppose an obstacle for those nubies is "finding the right people to
help you". One can argue where the line is, between "help you" and
"doing it for you". But as Allison and Bill note, some critical
knowlege is from experience, and sometimes you just have to ask around
to find out.
Thanks for the comments.
Herb Johnson
retrotechnology.com
.
- Follow-Ups:
- Re: What are your obstacles to running CP/M?
- From: no . spam
- Re: What are your obstacles to running CP/M?
- References:
- What are your obstacles to running CP/M?
- From: Herb Johnson
- Re: What are your obstacles to running CP/M?
- From: Bill H
- Re: What are your obstacles to running CP/M?
- From: no . spam
- What are your obstacles to running CP/M?
- Prev by Date: Re: What are your obstacles to running CP/M?
- Next by Date: Re: How do you grab memory dynamically in CP/M?
- Previous by thread: Re: What are your obstacles to running CP/M?
- Next by thread: Re: What are your obstacles to running CP/M?
- Index(es):
Relevant Pages
|
