Portable CP/M
Floppy Software
I am very interested about CP/M, and one thing I want to see is CP/M
running in a modern PC.
I use some emulators, and I have an AMSTRAD PCW too, but what I want
is to run
a modern CP/M in a modern computer.
Some time ago, I discovered Portable CP/M. Sources for BDOS and
utilities (PIP, STAT...) are available, and it is written in C, first
for the 68000 cpu, and then for the Z8000.
It's seems based on CP/M 2, but I think it can be a good start to have
a modern CP/M.
There was any atempt to do this in te past?
How difficult do you think can be to do this?
An Open CP/M sounds very good to me!
Thanks.
Peter Dassow
> I am very interested about CP/M, and one thing I want to see is CP/M
> running in a modern PC.
>
> I use some emulators, and I have an AMSTRAD PCW too, but what I want
> is to run a modern CP/M in a modern computer.
> How difficult do you think can be to do this?
>
> An Open CP/M sounds very good to me!
>
> Thanks.
Are you really interested in having an 8 bit monolithic OS (which was
superseded already by MS-DOS later) on a 64 bit CPU ?
There is already an open source OS which can be stripped down to a lean
alternative ... Linux, and it's modular, multitasking capable etc...
And there are also Z80 emulators which are working great on a modern PC,
e.g. z80pack, SIMH/Altair 8800 simulator or Yaze (s.a.
www.z80.eu/myz80cpm.html ).
So why wasting time in a probable failing project ?
If you want to learn about operating systems, or even if you want to
extend it by yourself, I can recommend Minix 3 ( www.minix3.org ), it's
a very well and clearly documented, modular OS for education purposes -
but really running on a modern PC and they're still working on it ;-)
Regards
Peter
Floppy Software
> Are you really interested in having an 8 bit monolithic OS (which was
> superseded already by MS-DOS later) on a 64 bit CPU ?
Yes, that's what I want. An after that, improve it. Why not?
> There is already an open source OS which can be stripped down to a lean
> alternative ... Linux, and it's modular, multitasking capable etc...
Yes, but I love CP/M. Linux it's not CP/M.
> And there are also Z80 emulators which are working great on a modern PC,
> e.g. z80pack, SIMH/Altair 8800 simulator or Yaze (s.a.www.z80.eu/myz80cpm.html).
Yes, I have some emulators, and they are beautiful things, but it's
not the same.
> So why wasting time in a probable failing project ?
Why not? I want only a CP/M capable of run in a modern computer: only
screen,
keboard and hard drive support. Then, with a running system, it's
possible to improve it.
> If you want to learn about operating systems, or even if you want to
> extend it by yourself, I can recommend Minix 3 (www.minix3.org), it's
> a very well and clearly documented, modular OS for education purposes -
> but really running on a modern PC and they're still working on it ;-)
Again, I want CP/M, not Linux, Minix, ...
I'm not a young boy that want to learn about OSs. I have quite
knowledge about
programming languajes and OSs. My AMSTRAD PCW it's not second hand.
>
> Regards
> Peter
Thanks anyway.
Mike
Roger Ivie
> Some time ago, I discovered Portable CP/M. Sources for BDOS and
> utilities (PIP, STAT...) are available, and it is written in C, first
> for the 68000 cpu, and then for the Z8000.
I'm not familiar with this "Portable CP/M" of which you speak.
However, over at http://cpm.z80.de/download/exchange-01.zip you'll find a
version of CP/M-68K that I poked at enough to make go through gcc. This
version runs as a guest under another OS (I developed on NetBSD), allowing
manipulation of a CP/M disk image.
CP/M-68K is written in ancient dialect of C that differs in some
important respects (most notably surrounding structures and pointers to
structures) from the C we know today.
I also prodded it enough to make it run standalone on a MicroVAX 2000.
In that case, I was running out of an image booted over the Ethernet
that included a disk image. I got to the point where all I needed was to
write disk drivers before World of Warcraft sucked out my soul.
Yes, it can be done. Personally, I'd like to see it run on ARM. But I
can't really think of a good enough reason to motivate me to do it.
--
roger ivie
ri...@ridgenet.net
Mr Emmanuel Roche, France
(Could you find a less ridicule name? The Internet is world-wide.
After using 2 nicknames, I simply could not find anything simpler than
using my real name, with the indication of my country.)
Hum...
As far as I understand the English language, your message mentions 4
distinct subjects:
1) "Portable CP/M"
2) CP/M on "modern" PC
3) CP/M emulators
4) "Open CP/M"
One big problem, to answer you, is that you do not mention the version
of CP/M that you are used to.
For example: here, in the comp.os.cpm Newsgroup, 99.99% of the
Americans are only interested in CP/M 2.2, with ASM and DDT. On the
contrary, 99.99% of the Europeans are only interested in CP/M Plus,
with MAC and SID... This makes for some interesting readings, at time!
Not knowing what is your background (and how long you have read the
comp.os.cpm Newsgroup), I am obliged to make some generalities.
CP/M 2.2 on the IBM Clown: Several years ago, there was one paranoid
American whose name I don't remember, except that he signed his
messages: "Anonymous Guy". As 99.99% of the Americans, he was only
interested in CP/M 2.2, but wanted to keep using it on the wretched
IBM Clown. He had a Web site, at a time, but it disappeared from the
Cyberspace. As far as I know, the only mirror is at:
http://www.nostalgia8.nl/mirrors/cpm86/
CP/M Plus on the IBM Clown: One day, circa 2000, that I mentioned that
I wanted to port CP/M Plus on the IBM Clown, (Englishman) John Elliott
mentioned that one little-known version of CP/M (used by Siemens in
Germany for a line of portables used to program a controller used in
factories) contained, in all the CMD files, the string: "CP/M-86 Plus
Digital Research"!
Needless to say, this message created quite a sensation! However,
despite some hard work by some, no source code and (more important) no
"System Guide" has been found, during the last 10 years. John Elliott,
being more interested in DOS Plus (the version with a MS-DOS Version 2
emulation layer), I worked mainly alone (while not trying to
ressurrect COMAL) on it. Over the years, 4 individuals asked me a copy
of my work: A German, an American, a Swede, and a ? (I forgot).
One little fact that I may mention is that all the versions of CP/M
found so far have their BIOSes written in assembly language, while all
the utilities (more than 20 for CP/M Plus) are written in PL/M. And PL/
M only ran under a DEC VAX... (There was a subset running under ISIS-
II, but it is much more rare than a VAX.)
Finally, in general, CP/M seems to have died in 1983, date of the last
versions. After that, compatibility with MS-DOS became more important,
with the various Concurrent DOS ("Concurrent" was a version of MP/M-86
specially tailored to run on the wretched IBM Clown hardware, with its
memory-mapped video).
Just last week, there was a thread about "FlexOS", a Real-Time version
of Concurrent DOS, whose Copyright is dated 1986. After that date,
impossible to find any mention of Digital Research. FlexOS is still
used, and sold, in 2010, by... IBM!
(All the BIOSes found so far were written for the IBM XT: since the
current hardware is based on the IBM AT, there is some slight
difference. For example: CP/M-86 Plus boots without any problem from a
floppy disk, but not from all hard disks.)
If you still have questions, be more specific.
Yours Sincerely,
Mr. Emmanuel Roche, France
Floppy Software
> On 2010-02-08, Floppy Software <floppysoftw...@gmail.com> wrote:
>
> > Some time ago, I discovered Portable CP/M. Sources for BDOS and
> > utilities (PIP, STAT...) are available, and it is written in C, first
> > for the 68000 cpu, and then for the Z8000.
>
> I'm not familiar with this "Portable CP/M" of which you speak.
>
> version of CP/M-68K that I poked at enough to make go through gcc. This
> version runs as a guest under another OS (I developed on NetBSD), allowing
> manipulation of a CP/M disk image.
>
> CP/M-68K is written in ancient dialect of C that differs in some
> important respects (most notably surrounding structures and pointers to
> structures) from the C we know today.
>
> I also prodded it enough to make it run standalone on a MicroVAX 2000.
> In that case, I was running out of an image booted over the Ethernet
> that included a disk image. I got to the point where all I needed was to
> write disk drivers before World of Warcraft sucked out my soul.
>
> Yes, it can be done. Personally, I'd like to see it run on ARM. But I
> can't really think of a good enough reason to motivate me to do it.
> --
> roger ivie
> ri...@ridgenet.net
Hi Roger!
The source code of CP/M-68K was the start point to CP/M-Z8K.
You can read "PORTABLE CP/M" in the .c files from CP/M-Z8K.
I believe that this was the start point to the source code of CP/
M-16K.
That's why Digital Research named it PORTABLE CP/M.
That's was an attempt to reduce time and efforts to develop CP/M for
each machine, in the way of UNIX.
But I don't know why didn't use PLM, because they needed to translate
source code of utilities as PIP or STAT to C.
Regards,
Mike
Al Kossow
> But I don't know why didn't use PLM, because they needed to translate
> source code of utilities as PIP or STAT to C.
>
The developer of these versions was an experienced UNIX programmer. The
compilers were bought from outside companies for these projects. Since
no PL/M compiers existed, it was faster to translate the utilites. The
Z8000 CP/M is based on the 68K version, and was done because Zilog was
begging them to port CP/M to the Z8000. I don't know if Olivetti was
already lined up as a customer when the project was started.
Floppy Software
wrote:
> Hello, "Floppy Software"!
Hello, Mr. Roche!
> (Could you find a less ridicule name? The Internet is world-wide.
> After using 2 nicknames, I simply could not find anything simpler than
> using my real name, with the indication of my country.)
I believe that it's not the better way to become friends!
In the past, in that group, I used the nick MikePcw. It's better to
you?
> As far as I understand the English language, your message mentions 4
> distinct subjects:
>
> 1) "Portable CP/M"
>
> 2) CP/M on "modern" PC
>
> 3) CP/M emulators
>
> 4) "Open CP/M"
>
> One big problem, to answer you, is that you do not mention the version
> of CP/M that you are used to.
If you know the AMSTRAD PCW I mentioned in my first post in this
thread,
you know too that I run natively CP/M 3.
In emulators, I use CP/M 2, 3, and alikes.
> For example: here, in the comp.os.cpm Newsgroup, 99.99% of the
> Americans are only interested in CP/M 2.2, with ASM and DDT. On the
> contrary, 99.99% of the Europeans are only interested in CP/M Plus,
> with MAC and SID... This makes for some interesting readings, at time!
I am mainly interested in CP/M 3.
> Not knowing what is your background (and how long you have read the
> comp.os.cpm Newsgroup), I am obliged to make some generalities.
You can use the "search" button. FloppySoftware, MikePcw, MESCC, can
be good words to search.
I had a website dedicated to CP/M and AMSTRAD PCW but my free
hosting is dead.
Do you remember MESCC. Yes, my own Small C version for CP/M.
> CP/M 2.2 on the IBM Clown: Several years ago, there was one paranoid
> American whose name I don't remember, except that he signed his
> messages: "Anonymous Guy". As 99.99% of the Americans, he was only
> interested in CP/M 2.2, but wanted to keep using it on the wretched
> IBM Clown. He had a Web site, at a time, but it disappeared from the
> Cyberspace. As far as I know, the only mirror is at:
>
> http://www.nostalgia8.nl/mirrors/cpm86/
Making friends again.
> CP/M Plus on the IBM Clown: One day, circa 2000, that I mentioned that
> I wanted to port CP/M Plus on the IBM Clown, (Englishman) John Elliott
> mentioned that one little-known version of CP/M (used by Siemens in
> Germany for a line of portables used to program a controller used in
> factories) contained, in all the CMD files, the string: "CP/M-86 Plus
> Digital Research"!
From the book THE AMSTRAD CP/M PLUS, by D. Powys-Lybbe and
R.M.Clarke:
"CP/M-86 Plus eventually emerged in 1985, called DOS Plus, two years
later than expected, with added MSDOS emulation."
That was written in April of 1986.
> Needless to say, this message created quite a sensation! However,
> despite some hard work by some, no source code and (more important) no
> "System Guide" has been found, during the last 10 years. John Elliott,
> being more interested in DOS Plus (the version with a MS-DOS Version 2
> emulation layer), I worked mainly alone (while not trying to
> ressurrect COMAL) on it. Over the years, 4 individuals asked me a copy
> of my work: A German, an American, a Swede, and a ? (I forgot).
With your modus operandi, it's not a rare thing.
> One little fact that I may mention is that all the versions of CP/M
> found so far have their BIOSes written in assembly language, while all
> the utilities (more than 20 for CP/M Plus) are written in PL/M. And PL/
> M only ran under a DEC VAX... (There was a subset running under ISIS-
> II, but it is much more rare than a VAX.)
CP/M-86K, CP/M-Z8K and CP/M-16K have all the utilities as PIP or
STAT in C language. Read the source code archives around the web.
> Finally, in general, CP/M seems to have died in 1983, date of the last
> versions. After that, compatibility with MS-DOS became more important,
> with the various Concurrent DOS ("Concurrent" was a version of MP/M-86
> specially tailored to run on the wretched IBM Clown hardware, with its
> memory-mapped video).
We agreed? Oh!
> Just last week, there was a thread about "FlexOS", a Real-Time version
> of Concurrent DOS, whose Copyright is dated 1986. After that date,
> impossible to find any mention of Digital Research. FlexOS is still
> used, and sold, in 2010, by... IBM!
>
> (All the BIOSes found so far were written for the IBM XT: since the
> current hardware is based on the IBM AT, there is some slight
> difference. For example: CP/M-86 Plus boots without any problem from a
> floppy disk, but not from all hard disks.)
Search better.
> If you still have questions, be more specific.
No more questions to you, thanks.
> Yours Sincerely,
> Mr. Emmanuel Roche, France-
Bye,
Mr. Ridiculous Nick Name, Spain.
Bye the way.
The source code for STAT in CP/M-Z8K says:
/* S T A T C O M M A N D
*/
/* -----------------------
*/
/*
*/
/* Common module for MP/M 2.0, MP/M-86 2.0, CP/M-80 2.2, CP/M-86 1.1
*/
/* and Portable CP/M implementation (P-CP/M)
Floppy Software
The source code for the utilities in 68k is in PL/M, but in Z8K, Zilog
translate it
to C. From the STAT source code:
/* Revised: */
/* 20 Jan 80 by Thomas Rolander */
/* 29 July 81 by Doug Huskey (for MP/M 2.0) */
/* 02 Sept 81 (for MP/M-86) */
/* 14 Nov 81 by Doug Huskey (for CP/M-86) */
/* 21 Sept 82 by Zilog (translate to C)
Mr Emmanuel Roche, France
> > (Could you find a less ridicule name? The Internet is world-wide.
> > After using 2 nicknames, I simply could not find anything simpler than
> > using my real name, with the indication of my country.)
>
> I believe that it's not the better way to become friends!
>
> In the past, in that group, I used the nick MikePcw. It's better to you?
Is it you, Miguel? !Hola! ?Que tal, Hombre?
(I still counsel you just to use your real name. Remember: the
Internet is World-Wide, so a few Billions people could read your
message. If you change your "nickname" regularly, people could be
confused.)
> I am mainly interested in CP/M 3.
Me too. CP/M Plus is a total rewrite of CP/M 2.2. It is more logical,
better documented, and the process of producing a SYS file streamlined
by the use of GENCPM. (Too bad the 8086 version is missing...)
> Do you remember MESCC. Yes, my own Small C version for CP/M.
I also remember how the discussion about a "C" for CP/M-86 ended in a
deadway...
> CP/M-86, CP/M-Z8K and CP/M-16K have all the utilities as PIP or
> STAT in C language. Read the source code archives around the web.
Yes, I know. There was a time when, thanks to the arrival of several
Unix Newbies inside DRI, the C language was fashionable and, since
several new CPUs had appeared, at the time, they had the idea of
rewritting CP/M in C, to port it more easily to them.
By the way, except a mention, did you find something new about CP/
M-16K ?
> > (All the BIOSes found so far were written for the IBM XT: since the
> > current hardware is based on the IBM AT, there is some slight
> > difference. For example: CP/M-86 Plus boots without any problem from a
> > floppy disk, but not from all hard disks.)
>
> Search better.
? Are you a specialist of the booting process of hard disks?
> By the way:
>
> The source code for STAT in CP/M-Z8K says:
>
> /* S T A T C O M M A N D
> /* -----------------------
> */
> /* Common module for MP/M 2.0, MP/M-86 2.0, CP/M-80 2.2, CP/M-86 1.1
> */
> /* and Portable CP/M implementation (P-CP/M)
/*
Revised: */
/* 20 Jan 80 by Thomas
Rolander */
/* 29 July 81 by Doug Huskey (for MP/M
2.0) */
/* 02 Sept 81 (for MP/
M-86) */
/* 14 Nov 81 by Doug Huskey (for CP/
M-86) */
/* 21 Sept 82 by Zilog (translate to C) */
If you remember, Doug Huskey once published a message in the
comp.os.cpm Newsgroup, in reaction to a message I sent him about CP/
M-86 Plus. Since he was looking for being paid, he stopped answering
my questions after discovering that we were trying to ressurrect CP/
M-86 Plus just for fun.
(Tom Rolander never answered my messages.)
8-bit CP/M has been placed into the public domain, but everybody
becomes crazy at the idea of MS-DOS compatibility.
Me, I am working on continuing using CP/M Plus, but on the IBM Clown.
That's why I built a custom 400-MHz system. Now, even my full-screen
3D graphics (even written in Dr. Logo, which is much slower than
Mallard BASIC) display instantly onscreen. I would say that 2 things
remain to be done:
1) better booting from hard disks
2) TCP/IP stack for CP/M
s_dub...@yahoo.com
> On 8 feb, 18:20, "Mr Emmanuel Roche, France" <roche...@laposte.net>
[snipped cruft]
>
> I had a website dedicated to CP/M and AMSTRAD PCW but my free
> hosting is dead.
>
> Do you remember MESCC. Yes, my own Small C version for CP/M.
>
There is free hosting here: http://www.000webhost.com/
I use them for my 'small-c with a nasm backend' project:
http://www.project-fbin.hostoi.com/index.htm
Nasm, can be used for cp/m-86. An old way to do it is in:
http://www.nostalgia8.nl/mirrors/cpm86/nasm2hdr.zip
You might consider putting the nasm syntax backend to your MESCC to
move your CP/M in C project along.
My latest version does that to small-c for CP/M-86.
Steve
[snipped cruft]
Steve Nickolas
> "Floppy Software" wrote:
>
>> In the past, in that group, I used the nick MikePcw. It's better to you?
>
> Is it you, Miguel? !Hola! ?Que tal, Hombre?
>
> (I still counsel you just to use your real name. Remember: the
> Internet is World-Wide, so a few Billions people could read your
> message. If you change your "nickname" regularly, people could be
> confused.)
I've used nicks on and off myself. (I mark my messages - even without my
name it's generally obvious who posted them)
>> I am mainly interested in CP/M 3.
>
> Me too. CP/M Plus is a total rewrite of CP/M 2.2. It is more logical,
> better documented, and the process of producing a SYS file streamlined
> by the use of GENCPM. (Too bad the 8086 version is missing...)
My main interest is all versions 2.2 and up which run on stock 8086 and are,
indeed, CP/M under the hood (i.e., not DR DOS).
>> Do you remember MESCC. Yes, my own Small C version for CP/M.
>
> I also remember how the discussion about a "C" for CP/M-86 ended in a
> deadway...
With the proper linker and libc, any DOS compiler could target CP/M-86 (I
never finished my own attempt at this).
> Me, I am working on continuing using CP/M Plus, but on the IBM Clown.
I'd prolly do fine with PCP/M-86 2 (which has the same back end as DOSPLUS
1.2, but without the DOS-ness), personally (pardon the incredibly lame pun).
> 1) better booting from hard disks
>
> 2) TCP/IP stack for CP/M
Even #1 would really be enough. Though #2 would open the door to some nice
stuff too. IRC from CP/M-86, anyone?
*hides*
-uso.
Floppy Software
Thanks, Steve
Floppy Software
wrote:
> Me too. CP/M Plus is a total rewrite of CP/M 2.2. It is more logical,
> better documented, and the process of producing a SYS file streamlined
> by the use of GENCPM. (Too bad the 8086 version is missing...)
CP/M 3 is a very good product, without any doubt.
> Yes, I know. There was a time when, thanks to the arrival of several
> Unix Newbies inside DRI, the C language was fashionable and, since
> several new CPUs had appeared, at the time, they had the idea of
> rewritting CP/M in C, to port it more easily to them.
I belive that DR was very slow working in adapting CP/M to 16 bit
systems.
Unfortunately to they (and to us).
> By the way, except a mention, did you find something new about CP/
> M-16K ?
Not yet.
> 8-bit CP/M has been placed into the public domain, but everybody
> becomes crazy at the idea of MS-DOS compatibility.
CP/M it's not in the public domain. Read again the "license" in Gaby
site.
> Me, I am working on continuing using CP/M Plus, but on the IBM Clown.
> That's why I built a custom 400-MHz system. Now, even my full-screen
> 3D graphics (even written in Dr. Logo, which is much slower than
> Mallard BASIC) display instantly onscreen. I would say that 2 things
> remain to be done:
>
> 1) better booting from hard disks
>
> 2) TCP/IP stack for CP/M
An the CP/M source code.
I am sure that CP/M 16 bit source code exist, but for an or another
reason
(money, I suposse), they (Caldera, etc.) didn't want to release it.
Only 16 bit source code CP/M for Z8K and 68K in C available? All the
other
stuff is lost? I don't believe this.
> Yours Sincerely,
> Mr. Emmanuel Roche, France
See you,
Miguel, Spain
Floppy Software
> >> Do you remember MESCC. Yes, my own Small C version for CP/M.
>
> > I also remember how the discussion about a "C" for CP/M-86 ended in a
> > deadway...
>
> With the proper linker and libc, any DOS compiler could target CP/M-86 (I
> never finished my own attempt at this).
Have a look at www.desmet-c.com.
Version 2.51 (and possibly others) works in CP/M-86. Sources are
available too.
James Moxham (Dr_Acula)
running in a modern PC. "
What an interesting question. I am assuming you want to have *only* CP/
M on the modern PC, ie a new BIOS that reads CP/M off a disk drive and
runs it. That would be difficult. How does it talk to a VGA display?
Does CP/M have drivers for SATA drives or USB?
You can use emulators as mentioned previously. I use the SIMH almost
every day for fast automatic compiling of CP/M programs that I then
download to a real CP/M board as the SIMH emulates a 200Mhz CP/M
machine.
Do you want a modern VGA display and modern keyboard and modern box?
Well, you can driver a modern keyboard from the N8VEM, and you can
drive a modern keyboard and VGA display from the propeller emulation.
But the actual computer need be nothing more than a 10x12cm board with
an sd card and 10 chips. Why use a PC with a 500W power supply when
you can run it on a board drawing 250 milliwatts? http://www.smarthome.viviti.com/propeller
Indeed, check out http://forums.parallax.com/forums/default.aspx?f=25&m=394297&p=7
and if you scroll to a post on the 3rd February there is a photo of a
CP/M computer in a matchbox.And how cool is this, that micro sd card
can hold the entire Walnut Creek CP/M archive.
There are a number of modern single board CP/M computers that run very
well, as indeed, there were many single board CP/M computers back in
the day, running on S100 and other busses and quite a number fitted
into a box much smaller than a modern PC. Perhaps we have gone
backwards with our 500W computers that take up to 10 minutes to boot
up?
re
>Why not? I want only a CP/M capable of run in a modern computer: only
>screen,
>keboard and hard drive support. Then, with a running system, it's
>possible to improve it.
There are a number of old and new designs that can do this. VGA,
standard $5 keyboard and hard drive support. And you can do your hard
drive support on a real hard drive, or a compact flash or even SD
card.
As for improving it - absolutely! What would you like to improve?
Mr Emmanuel Roche, France
> And the CP/M source code.
>
> I am sure that CP/M 16 bit source code exist, but for one
> or another reason (money, I supose), they (Caldera, etc.)
> didn't want to release it.
>
> Only 16-bit source code CP/M for Z8K and 68K in C available?
> All the other stuff is lost? I don't believe this.
Hahaha! I hope that you are not paranoid, Miguel!
There is no conspiracy plot by the Americans (as far as I can
understand) to hide any "legacy" stuff from Digital Research.
If I have well understood (and remember well) what Tim Olmstead wrote
before dying, Caldera released everything that they were given by
Novell, which had previously bought Digital Research. According to
him, many things were lost (like the source code of "Personal CP/M")
by Novell.
So, I really believe that what is available today on the Internet is
everything that was given to Caldera, and that it was Novell which
goofed and lost many interesting things (like the source code of
"Display Manager", still used and sold today, in 2010, by IBM... but
nowhere to be found, like "Access Manager". Both were CP/M 2.2
programs, yet they are nowhere to be found.).
As for "the CP/M source code", this is simple. According to all the
files that I have read, the BIOS was written in assembly language,
while the utilities were written in PL/M. Only CP/M-Z8000 and CP/
M-68000 were programmed in C. As Axel Berger explained, this was the
direct cause of all the bugs found later... So much for the "power of
C"! (Note to pig-headed persons: I am *NOT*, as usual, criticising
"C". See? I merely mention a *HISTORICAL FACT*. Why is the code
written by DRI in C full of bugs, and not the one written in PL/M? I
don't know. But this makes me think, for sure. A thing that C fanatics
cannot do, of course.)
Once you understand this, you can recreate the missing files.
For example, we have only binaries of the CMD files of the 8086
version of CP/M Plus, but we have the source code in PL/M of the
utilities of CP/M Plus...
So, either you recompile them with the option of generating 8086 code,
or simply disassemble the CMD files, using the high-level source code
in PL/M...
Normally, you have:
1) the source code
2) the binary (COM or CMD files)
3) the doc
In our case, for all the versions of CP/M, there are some missing
cases but, by using the surrounding cases, it is possible to recreate,
step by step, some cases.
This is how I recreated the source code of the missing CP/M-86 Plus
utilities, or the "CP/M-86 Plus Programmer's Guide" by removing the MS-
DOS part of the "DOS Plus Programmer's Guide".
It is a way to spend those long Winter nights...
Herbert Johnson
> >"I am very interested about CP/M, and one thing I want to see is CP/M
> > running in a modern PC. "
Aren't there variations of "CP/M-86 for the IBM PC" which will boot up
on fairly modern Intel PC's today? A search of comp.os.cpm on this
subject ought to yield the usual resources. It's a old, old question.
On Feb 9, 8:31 am, "James Moxham (Dr_Acula)" wrote:
>
> What an interesting question. I am assuming you want to have *only* CP/
> M on the modern PC, ie a new BIOS that reads CP/M off a disk drive and
> runs it. That would be difficult. How does it talk to a VGA display?
> Does CP/M have drivers for SATA drives or USB?
CP/M, like MS-DOS, can use BIOS software to perform hardware access.
to the extent the very oldest BIOS interrupts are still supported,
recent Intel systems "ought" to support CP/M-86. If not, use a virtual
machine such as QEMU to emulate what amounts to an older Pentium-based
system. Use that to make changes to support direct operation.
CP/M and USB just don't mix, nor need they. What USB devices, other
than storage, would even be useful to a CP/M program?
> Do you want a modern VGA display and modern keyboard and modern box?
> Well, you can driver a modern keyboard from the N8VEM, and you can
> drive a modern keyboard and VGA display from the propeller emulation.
This person chooses to run on readily-available modern PC's, and
chooses to run 8086 native code and not Z80 code. What's wrong with
that?
> There are a number of modern single board CP/M computers that run very
> well, as indeed, there were many single board CP/M computers back in
> the day, running on S100 and other busses and quite a number fitted
> into a box much smaller than a modern PC. Perhaps we have gone
> backwards with our 500W computers that take up to 10 minutes to boot
> up?
I agree that much of modern personal computing has gone "backwards" as
you suggest. One reason to keep legacy OS's alive, such as CP/M-86, is
to remind us of prior history.
> >Why not? I want only a CP/M capable of run in a modern computer: only
> >screen,
> >keboard and hard drive support. Then, with a running system, it's
> >possible to improve it.
> As for improving it - absolutely! What would you like to improve?
I don't think this person is obliged to have a detailed adjenda. It's
sufficient for this person to get the CP/M-86 they want, on the
computer they want, and then see what it took to do THAT, and how it
performs accordingly. Opportunities for "improvements" may be to
provide better hardware support, if support seems to be lacking or
unresponsive. It may be discovered that USB serial devices, or USB
mass storage devices, could be "added" to CP/M support.
I think this person is simply in a process of discovery. He has set a
goal - run CP/M-86 on a modern PC, look for opportunities for
improvements - and want to see where that takes him. I think that's
fine. Why not find out what HE thinks he needs, or point him to
resources he will LIKELY need, instead of suggesting he follow some
other adjenda?
I believe the resources to run CP/M 86 on modern PC's, have been
established on the Web at a number of Web sites, and for some time.
But it may be that through "Web decay", those resources are not active
or are outdated. I hope "floppy" reports his success in due course,
and shows us the trail he followed to achieve this - so we know the
trail is still there.
Herb Johnson
retrotechnology.com
Herbert Johnson
> If I have well understood (and remember well) what Tim Olmstead wrote
> before dying, Caldera released everything that they were given by
> Novell, which had previously bought Digital Research. According to
> him, many things were lost (like the source code of "Personal CP/M")
> by Novell.
>
> As for "the CP/M source code", this is simple. According to all the
> files that I have read, the BIOS was written in assembly language,
> while the utilities were written in PL/M. Only CP/M-Z8000 and CP/
> M-68000 were programmed in C.
> Once you understand this, you can recreate the missing files.
As usual, Roche talks about everything but provides no specifics.
There's an incomplete collection of CP/M-86 original sources, but any
number of disassemblies of them. The pertinent question is, where are
these sources, objects, and disassemblies? Where are the tools one
might need?
http://www.cpm.z80.de/source.html
Section "CP/M-86" shows sources for CP/M-86, gathered years ago. many
are disassemblies of course in 8086 assembler. On the same Web page,
there are PL/M compilers in Fortran and for VAX. Elsewhere on the site
is a PL/M compiler which runs under Intel's ISIS operating system
(with an ISIS emulator) - to my knowledge it produces 8080 code, there
may be an 8086 version.
CP/M 80 sources may be informative about CP/M-86 sources.
http://www.seasip.demon.co.uk/Cpm/
A long-standing CP/M support site, including support for CP/M-86.
Details of CP/M-86 operation.
http://www.nostalgia8.nl/mirrors/cpm86/patch.htm
A long-standing site which describes how to "patch" CP/M 86 to boot on
a then-modern PC.
http://www.unix4fun.org/z80pack/
This is Udo Munk's site for his z80pack emulator under Linux and
Windows. Under that emulator he has methodically provided all the
early 8080 based DRI products, including sources and tools for working
with them. The point of referencing these regarding CP/M-86, is that
his methods would be the same as for work with CP/M-86 sources. Also
the DRI 8080 products and sources may be informative about their 8086
products, as early versions of CP/M-86 were said to have been
recompiled versions of CP/M-80 products.
http://www.retrotechnology.com/dri/howto_cpm.html
My own, general "how to" notes about running CP/M and what one can run
under CP/M. There are many Web links to sites with CP/M information
and software, including these links. I try to respond to the usual
questions.
Herb Johnson
PS: Miguel, if you produce "new" sources, please do what Roche did
NOT do. 1) Provide them to others. 2) Clearly identify these as
recreations! It's very confusing when Roche changed documents and
source code, but DID NOT IDENTIFY WHAT HE DID. In effect he created
*forgeries* of original work, which may be confused with actual finds
of original work.
Herbert R. Johnson, New Jersey USA
http://www.retrotechnology.com/ retro-technology home pages
-- S-100, CP/M history by "Dr. S-100"
domain mirror: retrotechnology.net
email: hjohnson AAT retrotechnology DOTT com
Mr Emmanuel Roche, France
> As usual, Roche talks about everything but provides no specifics.
and
> PS: Miguel, if you produce "new" sources, please do what Roche did
> NOT do. 1) Provide them to others. 2) Clearly identify these as
> recreations! It's very confusing when Roche changed documents and
> source code, but DID NOT IDENTIFY WHAT HE DID. In effect he created
> *forgeries* of original work, which may be confused with actual finds
> of original work.
Ok. I am taking this opportunity to set straight a few *FACTS*:
1) *FOR YEARS*, I have provided backup copies of my Hard Disk to Herb
Johnson (and a few other regulars of the comp.os.cpm Newsgroup, if
Herb contests this).
2) I am a programmer. Herb once signed a message: "Herb "Not a
Programmer" Johnson"... So, is it surprising that my Hard Disk
contains "works in progress"? Each time I find an interesting DRI
file, I print it. As I explained several times, I only publish files
which no longer evolves, or were sufficiently improved to republish
them. For me, the only reason to use a computer is that a document can
be improved forever. That's why I never put dates in my stuff. When I
no longer use it, I simply publish it. Apparently, some people find
them interesting enough to collect them...
3) I particularly enjoyed the accusation of *forgeries" by Herb...
Several times, I mentioned a fact, only to notice it appearing on
Herb's Web site, a few weeks later... One day, in a private message to
Herb, I mentioned that only Dr. John Torode remained alive, to tell us
where was the first microcomputer which booted CP/M. After that
message, Herb disappeared during a few weeks. Several months later,
while having a look to his Web site, I discovered the end of the
story, that he published only on his Web site, not privately to me, or
publicly to the comp.os.cpm Newsgroup: he corresponded with Dr. Torode
but, as usual (since he is not a programmer), all the stuff that he
got dealt with the hardware (In TCJ, he was known as "Dr. S-100"). Me,
when I surf his Web site, I see countless mention of stuff that I gave
him, that he had never mentioned before me.
4) Notice that all the files that I write start by a name and, when I
retype an article, always contains the full references... This must be
*forgeries* ?
Faced to all those *FACTS*, how do you "identify" Herb Johnson?
Maybe Herb Johnson is "confused" with "original work", but me I am not
confused by his behavior!
James Moxham (Dr_Acula)
M
running in a modern PC."
I'm still a bit unclear about the question but I don't think you can
boot up CP/M on a modern PC. But if this is possible, I'd be
interested to see how because I have some old PCs lying around.
Herb says "What USB devices, other than storage, would even be useful
to a CP/M program?"
I think the most obvious one would be a serial port or two. Those are
fairly fundamental to a CP/M program, unless you direct Conin and
Conout somewhere else. The problem is that modern PCs don't have
serial ports any more, so you have buy a USB to serial port adaptor.
No problem on a PC, because the adaptor is worth $3 and you plug it in
and Windows names a new COM port. But if you were running CP/M on a PC
you need that software in the background to run the USB host.
And hard drives, well it is getting hard now to find a modern PC that
has IDE drives. Everything changes so rapidly, and if you want a PC
with an IDE drive and a real serial port, you have to search really
hard or get one from a second hand store.
So the practical solution is to get a modern PC, run a modern
operating system on it and then run an emulation of some kind. I'm
fairly sure most emulations can attach CP/M devices like Conin and
Punch to real world ports on a PC (like COM1, or even COM5 where COM5
is actually a USB to serial device).
Re "I use some emulators, and I have an AMSTRAD PCW too, but what I
want is to run a modern CP/M in a modern computer."
If you don't want to run an emulator, why not run on a real board?
Either a new board or a pre-loved one. Herb Johnson sells real boards
that will run CP/M. You can always put that board inside a PC box with
a sticker "CP/M Inside" and no-one will know the difference!
Freek
"Mr Emmanuel Roche, France" <roch...@laposte.net> wrote in message
news:248cc82b-40b7-484e...@l26g2000yqd.googlegroups.com...
> As far as I understand the English language....
> ...
> Several years ago, there was one paranoid
> American whose name I don't remember, except that he signed his
> messages: "Anonymous Guy".
>
Emmanuel,
Several years ago, I had lots of very pleasant email exchanges with this
same person (Kirk Lawrence) and it never occurred to me that he might be
"paranoid" in any way. Why are you using this rather denigrating
qualification?
Or maybe you just misunderstand the English word "paranoid"?
Regards,
Freek.
Axel Berger
>Unfortunately to they
I find this extremely funny. In school I learnt correct English as:
"Who wants another slice of cake?" "Me."
While different from most other languages, that was, what correct
English required, what I learnt, and what I still use. Later some self
declared "educated" English people decided, they had to use the
nominative too, where everyone else did. So English got a bit more
bland and less beautiful, bad enough. But as these people were and are
totally clueless about language and its rules, they were incapable of
changing the nominative alone. All they could manage was a "me bad, I
good; him bad, he good", like the sheep in 1984.
So now it is:
"Who's here today?" "I"
"Whose is this purse?" "I"
"Whom did I hand my purse?" "I"
"Whom did I offend?" "I"
instead of I, mine, me, me, as would be grammatically correct.
P.S: Sorry, Miguel. After writing all this I came to your signature and
noticed, you're not a native English speaker. Still, as you only copy
what many of them do here, I'll leave it.
Floppy Software
wrote:
> Hello, Miguel!
>
> > And the CP/M source code.
>
> > I am sure that CP/M 16 bit source code exist, but for one
> > or another reason (money, I supose), they (Caldera, etc.)
> > didn't want to release it.
>
> > Only 16-bit source code CP/M for Z8K and 68K in C available?
> > All the other stuff is lost? I don't believe this.
>
> Hahaha! I hope that you are not paranoid, Miguel!
>
> There is no conspiracy plot by the Americans (as far as I can
> understand) to hide any "legacy" stuff from Digital Research.
>
> If I have well understood (and remember well) what Tim Olmstead wrote
> before dying, Caldera released everything that they were given by
> Novell, which had previously bought Digital Research. According to
> him, many things were lost (like the source code of "Personal CP/M")
> by Novell.
>
> So, I really believe that what is available today on the Internet is
> everything that was given to Caldera, and that it was Novell which
> goofed and lost many interesting things (like the source code of
> "Display Manager", still used and sold today, in 2010, by IBM... but
> nowhere to be found, like "Access Manager". Both were CP/M 2.2
> programs, yet they are nowhere to be found.).
Yes, "Caldera released everything that they were given by Novell" as
you say.
But, Novel gave all the stuff they had? That's the important thing.
Do you know that Novell marketed Novell Dos (really DR DOS) when
they bought Digital Research?
An then, they surprisingly finished developing and shelling it?
Why Novell bought Digital Research, then?
They gave to Caldera very old stuff, but they not found more recent
sources?
Well, I'm not a paranoid man, but I don't believe that.
> As for "the CP/M source code", this is simple. According to all the
> files that I have read, the BIOS was written in assembly language,
> while the utilities were written in PL/M. Only CP/M-Z8000 and CP/
> M-68000 were programmed in C. As Axel Berger explained, this was the
> direct cause of all the bugs found later...
They translated the sources for PIP, STAT and ED from PL/M to C.
The bugs came from that.
Floppy Software
Yes, you're right in three things:
1. I'm not a native English speaker.
2. "Unfortunately to them" is the right sentence.
3. I copy that I want.
Apart from that, do you find something of interest in my previous
msgs?
Related to CP/M, of course.
Floppy Software
Hi Herbert,
My interest in CP/M came from the mids 80s when I bought
my first computer: an Amstrad PCW with CP/M Plus.
I enjoyed a lot using SID to discover what the "dark" CP/M functions
did...
In that time, I didn't have programmers manual for CP/M.
> My own, general "how to" notes about running CP/M and what one can run
> under CP/M. There are many Web links to sites with CP/M information
> and software, including these links. I try to respond to the usual
> questions.
I visit your website frequently and it is very very useful, thanks.
> PS: Miguel, if you produce "new" sources, please ... Provide them to others.
Of course. I had a website with some of my programs. I am looking for
a
new hosting. My small-c version for CP/M-80 (MESCC) still somewhere on
the net.
Thanks Herbert.
Floppy Software
wrote:
> Re "I am very interested about CP/M, and one thing I want to see is CP/
> M
> running in a modern PC."
>
> I'm still a bit unclear about the question but I don't think you can
> boot up CP/M on a modern PC.
Well, for sure I explained my interest very bad.
That I want is a CP/M with source code that can be compiled and used
today in
computers of today, not a Z80 emulator:
I only ask if the sources from Portable CP/M (68K, Z8K, 16K) written
in C,
can be used for that.
I did ask same question in 2004 and in 2007, and I can see that things
remains the same: not much interest in that thing.
Well, I will try to compile PIP source code, and run in native and
emulated CP/M.
pbetti
Just a curiosity. You mention Portable CP/M for a 16K implementation.
Where is it? I look on www.cpm.z80.de and didn't find the sources.
Am i missing something?
More in general: where to find the most complete source set for
Portable CP/M (clearly in C)?
Piergiorgio Betti
glen herrmannsfeldt
(snip)
> Well, for sure I explained my interest very bad.
> That I want is a CP/M with source code that can be compiled
> and used today in computers of today, not a Z80 emulator:
Since the 8086 instruction set was designed to be source compatible
with the 8080 instructions, it should be possible to directly
translate the 8080 source to 8086 source, assemble, and run it.
That may have been the beginning of CP/M-86
-- glen
Steve Nickolas
> "Who wants another slice of cake?" "Me."
Actually, I learned that that was, while almost exclusively what was
actually used, not the proper way (which was "I do"). There are a lot of
cases where I was taught one thing but everyone, including myself, generally
says another.
For the record (moving on!) - I wonder how hard it would be to take a C
userland for CP/M, and get it working on CP/M-80 and -86 apart from the
original userland.
-uso.
Herbert Johnson
revealing about his own attitudes about his "work" and what others do
with it, and about how he works with others.
Essentially, he says that his documents are never done, so he can't
really publish them. This makes it convenient for him to claim I've
used his "work", when I actually DO something about something he said.
Or, more often, when I write and post on my site, what OTHERS have
done with the history of early CP/M. And he's not very happy about
what is written, in any event, a "sour grapes" response.
I welcome anyone to visit my site and judge what is there on its own
merits. Meanwhile, it's convenient for Roche to point to his
unpublished work, or a few comments of his, and then claim I "scooped"
him.
Also, he makes some claims about sending out CD-ROMs, or the
"collection" of things he posts in comp.os.cpm, posting which he calls
"publication". The facts are, that he refuses to create or use a Web
site, or work with CP/M archiving sites, which are a reasonable venue
for his printed and programming work on CP/M. Instead he posts
documents here. I happen to have first-hand knowledge of all this.
He sent me, unsolicited, CD-ROM backups of his hard drive - but
strictly NOT for distribution, only for his backup purposes. Later, he
sent one with a "Last Will" ("published" also in comp.os.cpm) that it
not be released until his death. A death, he defines as some period of
time when he does not post in comp.os.cpm! At least one other person
who got his "last will" CD, posted that they discarded it. I've posted
that I reject being his "executor". Neither of these represent a true
distribution of Roche's work.
As for "collecting" what he's posted - I was the one to try to do
that, via my Web site a few years ago. I took some of his posted
documents and provided them as documents to CP/M archiving sites. They
worked with me. Then he insulted me once too often (the "not a
programmer" remark), and I stopped that effort.
Finally, my remark about "forgeries". I stand behind my previous post,
that Roche is not cautious about recreating code and documents and
identifying them as not original. If such items are well-identified, I
have no issue with them. Period.
The bottom line in all this, is that Roche insists on using
comp.os.cpm for his own purposes, NOT using other venues, and that his
posting behavior has become more rude, hostile and off-topic. This is
simply undesirable behavior, but discussing and arguing about it at
length is counter-productive. That's part of the nature of dealing
with a "troll". My remarks about trolls - and the risks of responding
to them - is on this Web page.
http://www.retrotechnology.com/herbs_stuff/offtopic.html
Herb Johnson
retrotechnology.com
s_dub...@yahoo.com
> Floppy Software <floppysoftw...@gmail.com> wrote:
>
> (snip)
>
> > Well, for sure I explained my interest very bad.
> > That I want is a CP/M with source code that can be compiled
> > and used today in computers of today, not a Z80 emulator:
>
32bit or 64bit stuff, it would hardly be recongnizable as cp/m.
If you want to use CD-Rom, you would need to support a hierarchical
file system. A cd bootable floppy image needs to be in fat form. So
you looking at CP/M on fat.
Considering all the odd devices, and the capabilities of the hardware,
you would probably want to relook at the bdos as a 'driver', just one
of some number of drivers operating in separate tasks.
-Or maybe you mean you want something like CP/M-86 operating off a
floppy? -only different?
> Since the 8086 instruction set was designed to be source compatible
> with the 8080 instructions, it should be possible to directly
> translate the 8080 source to 8086 source, assemble, and run it.
>
> That may have been the beginning of CP/M-86
>
> -- glen
Hi Glen, imho, that _was_ the beginnings of CP/M-86.
If you take the DRI xlt translation program, on an emulator, to the cp/
m-80 v2.2 bdos & ccp sources, you end up with cp/m-86 v1.0 less the
memory management stuff, and a few other things needed for the 8086.
So doing it gets you about 85% of the way to cp/m-86 bdos&ccp. The
footprints of DRI doing that is in the Loader code..
see: LDBDOS.A86
; title 'Bdos Interface, Bdos, Version 2.2 Feb, 1980'
in dsk 04 of MPM862SR.ZIP
Steve
All...@localhost.net
<roch...@laposte.net> wrote:
>If I have well understood (and remember well) what Tim Olmstead wrote
>before dying, Caldera released everything that they were given by
>Novell, which had previously bought Digital Research. According to
>him, many things were lost (like the source code of "Personal CP/M")
>by Novell.
Go read again, You forget Lineo!
As one of Tim's peronal frieds and one of the the holders of Tim's
archive set (6 CDs) as well as Gaby we were asked to insure
there would always be an on line archive of that work before he left
the world in 2001.
>As for "the CP/M source code", this is simple. According to all the
>files that I have read, the BIOS was written in assembly language,
>while the utilities were written in PL/M. Only CP/M-Z8000 and CP/
Not totlaly true. The early versions (V1 - V1.4) bios was written in
PL/M and later translated to ASM as that was part of the supplied
utilities (MDS800 BIOS source too).
>M-68000 were programmed in C. As Axel Berger explained, this was the
>direct cause of all the bugs found later... So much for the "power of
>C"! (Note to pig-headed persons: I am *NOT*, as usual, criticising
>"C". See? I merely mention a *HISTORICAL FACT*. Why is the code
>written by DRI in C full of bugs, and not the one written in PL/M? I
>don't know. But this makes me think, for sure. A thing that C fanatics
>cannot do, of course.)
The modern C programming community does this all the time
a few binutes browsing sourceforge and other repositories
might be a learning experience if you able to understand how
it works and what CVS does.
It wasn't C, it was the incomplete sources, unknown revisions
and lost pieces. You cant build from incomplete or multiple
revision code. Additionally you need a 1982 vintage compliler
as C like most languages currrently in use has evolved and
the sematics of early C had variations that were incompatable
not unlike the multitue of BASIC flavors.
The PL/M version was buggy too.
As to portable C thats teh responseability of the programmer
and if done correctly C is as portable a language as possible.
However, the programmer must also allow for anything that
is machine dependent or the will not work or at best be buggy.
Doesn't help that the C code was created by creating a PL/M
to C tanslator that was buggy at best.
>Once you understand this, you can recreate the missing files.
If you do not know the function or content of the missing files
your only guessting lacking specification.
>For example, we have only binaries of the CMD files of the 8086
>version of CP/M Plus, but we have the source code in PL/M of the
>utilities of CP/M Plus...
>
>So, either you recompile them with the option of generating 8086 code,
>or simply disassemble the CMD files, using the high-level source code
>in PL/M...
Only if you have the original PL/M compiler and envronment thankfully
I believe Odo put serious work into regreating that.
>Normally, you have:
>
>1) the source code
But it must be complete, any recreaton of missing peices
are only guesses.
>2) the binary (COM or CMD files)
Handy to see if what you produced it the same.
>3) the doc
However the really needed DOC may not exist and that would be
the internal to DRC specification of what CP/M is and how it should go
about doing it.
>
>In our case, for all the versions of CP/M, there are some missing
>cases but, by using the surrounding cases, it is possible to recreate,
>step by step, some cases.
Some cases, not all and not identically.
>This is how I recreated the source code of the missing CP/M-86 Plus
>utilities, or the "CP/M-86 Plus Programmer's Guide" by removing the MS-
>DOS part of the "DOS Plus Programmer's Guide".
Reverse engineering is an exercise often for no good reason other than
you can or outright theft. However you can never claim that code is
the original DRC source, only an approximation that compiles to the
same binary.
Allison
BruceMcF
> By the time you make all the allowances for memory management, and
> 32bit or 64bit stuff, it would hardly be recongnizable as cp/m.
Unless the memory management was hidden within BDOS as RAM disks ...
in which case maybe CP/M Plus is a better case, but either way it'd be
quite clearly recognizable as CP/M.
> If you want to use CD-Rom, you would need to support a hierarchical
> file system. A cd bootable floppy image needs to be in fat form. So
> you looking at CP/M on fat.
A bootable floppy (USB key etc.) needs to be in fat form, but a CP/M
image could well be an image file in the fat system - so again the
bdos acquires more of the capabilities of a CP/M emulator, but running
as the host system.
And as in an emulator, if there is a default image that is the
original "drive" when a new USB/SD/CD-ROM etc. is started, then if
that default image contains a utility that can mount a new image, the
utility can handle all the specification of the hierarchical file
system.
> Considering all the odd devices, and the capabilities of the hardware,
> you would probably want to relook at the bdos as a 'driver', just one
> of some number of drivers operating in separate tasks.
Ah, but if you want separate tasks, then its MP/M. Start with CP/M and
work up.
s_dub...@yahoo.com
> On Feb 10, 3:10 pm, s_dubrov...@yahoo.com wrote:
>
> > By the time you make all the allowances for memory management, and
> > 32bit or 64bit stuff, it would hardly be recongnizable as cp/m.
>
As to your following comments.. the memory management for CP/M-86 is
hidden in the BDOS with very disappointing results, badly considered
algorithms, even mp/m-86 and ccp/m-86 did things differently, tho with
some backward support to mem fns for cp/m-86 programs. CP/M Plus
probably would be a better choice, except, even less is known about CP/
M-86 Plus - it looks to me to be a deprecated ccp/m. It would've been
better to abstract the functions for the bdos and push off into the
bios the implementation mapping.
> Unless the memory management was hidden within BDOS as RAM disks ...
> in which case maybe CP/M Plus is a better case, but either way it'd be
> quite clearly recognizable as CP/M.
>
Well, there is the rub for me. How do you define, or what is, 'quite
clearly recognizable as CP/M'? This probably should be looked at
early, while considering such a task.
a. Is it the A> prompt and commands?
b. Is it something more abstract such as {'utilitarian, straight
foreward, understandable by one, simple-ist workhorse there is'}??
c. ??
I tend to agree with what Allison has said in the past, that cp/m only
makes sense, or mostly makes sense for the 8080cpu. (I was so
impressed by her observation, I made a copy of it here)..
http://www.project-fbin.hostoi.com/cpmlinks/SoftwareArcheology/CPM80vs.txt.htm
Her observation (or argument) holds water...
Unless we are talking about (b.).
I've a feeling that 'what is cp/m' is in the eyes of the beholder,
still the question needs asking before undertaking such a task. For
MikePCW, or anyone.
> > If you want to use CD-Rom, you would need to support a hierarchical
> > file system. A cd bootable floppy image needs to be in fat form. So
> > you looking at CP/M on fat.
>
> A bootable floppy (USB key etc.) needs to be in fat form, but a CP/M
> image could well be an image file in the fat system - so again the
> bdos acquires more of the capabilities of a CP/M emulator, but running
> as the host system.
>
Yes.
> And as in an emulator, if there is a default image that is the
> original "drive" when a new USB/SD/CD-ROM etc. is started, then if
> that default image contains a utility that can mount a new image, the
> utility can handle all the specification of the hierarchical file
> system.
Yes.
> > Considering all the odd devices, and the capabilities of the hardware,
> > you would probably want to relook at the bdos as a 'driver', just one
> > of some number of drivers operating in separate tasks.
>
> Ah, but if you want separate tasks, then its MP/M. Start with CP/M and
> work up.
Yes.
The bootstrap of a mp/m or ccp/m bdos|bios required a functioning cp/m
and ddt to load it and debug it. That's where it stands. The way
foreward from here would be.. b.?
Steve
All...@localhost.net
<agi...@netscape.net> wrote:
>On Feb 10, 3:10�pm, s_dubrov...@yahoo.com wrote:
>> By the time you make all the allowances for memory management, and
>> 32bit or 64bit stuff, it would hardly be recongnizable as cp/m.
>
>Unless the memory management was hidden within BDOS as RAM disks ...
>in which case maybe CP/M Plus is a better case, but either way it'd be
>quite clearly recognizable as CP/M.
Roger Ivie ran CP/M 68K sources on VAX cpu and thats 32bit!
CP/M plus would be a better choice as the hooks are there for a MMU
but if running on a machine with more than 16bit addressing you
already get a larger space. But the only choice you have for porting
to a differnt platform is the CP/M68K as the C source is more likely
to port than 8080 or 8088 assembler.
The memeory management is and should be hidden save from the
applications programmer. That person needs to know how to aquire
and allocate/deallocate space for his application. Not a big task
but an important one.
>> If you want to use CD-Rom, you would need to support a hierarchical
>> file system. �A cd bootable floppy image needs to be in fat form. �So
>> you looking at CP/M on fat.
>
>A bootable floppy (USB key etc.) needs to be in fat form, but a CP/M
>image could well be an image file in the fat system - so again the
>bdos acquires more of the capabilities of a CP/M emulator, but running
>as the host system.
That is a limit as CP/M has some hard size limits that most CF or USB
memory sticks easily exceed and most large cpus (32bit or larger)
have more addressable ram than 32MB! The problem there that
the extent based FS doesn't have enough bits to address it.
>And as in an emulator, if there is a default image that is the
>original "drive" when a new USB/SD/CD-ROM etc. is started, then if
>that default image contains a utility that can mount a new image, the
>utility can handle all the specification of the hierarchical file
>system.
Emulators hide the hardware and create directly or indirectly
a vertual machine and hardware.
>
>> Considering all the odd devices, and the capabilities of the hardware,
>> you would probably want to relook at the bdos as a 'driver', just one
>> of some number of drivers operating in separate tasks.
>
>Ah, but if you want separate tasks, then its MP/M. Start with CP/M and
>work up.
The BIOS is a driver. The BDOS incurrent parlance is the API.
However, with a tiny OS there is nothing to have an overlay os
running multiple instances of CP/M as individual tasks. That does
solve the problem nicely of the BDOS (other than MPM) lacking
the ability to support recusion of itself. If you apportion 60K
to each instance you have as many as 16 tasks in a megabyte
with overlying OS and driver space. That would be assuming Z80
with MMU and 1mb but that is a implmentation limitation. For larger
CPUs you can grow from their.
Allison
Steve Nickolas
> On Wed, 10 Feb 2010 14:12:56 -0800 (PST), BruceMcF
> <agi...@netscape.net> wrote:
>
>>On Feb 10, 3:10�pm, s_dubrov...@yahoo.com wrote:
>>> By the time you make all the allowances for memory management, and
>>> 32bit or 64bit stuff, it would hardly be recongnizable as cp/m.
>>
>>Unless the memory management was hidden within BDOS as RAM disks ...
>>in which case maybe CP/M Plus is a better case, but either way it'd be
>>quite clearly recognizable as CP/M.
>
> Roger Ivie ran CP/M 68K sources on VAX cpu and thats 32bit!
So's the 68000 ~.^
>>> If you want to use CD-Rom, you would need to support a hierarchical
>>> file system. �A cd bootable floppy image needs to be in fat form.
�So
>>> you looking at CP/M on fat.
>>
>>A bootable floppy (USB key etc.) needs to be in fat form, but a CP/M
>>image could well be an image file in the fat system - so again the
>>bdos acquires more of the capabilities of a CP/M emulator, but running
>>as the host system.
>
> That is a limit as CP/M has some hard size limits that most CF or USB
> memory sticks easily exceed and most large cpus (32bit or larger)
> have more addressable ram than 32MB! The problem there that
> the extent based FS doesn't have enough bits to address it.
DOSPLUS does FAT16; if necessary one could implement FAT and just overlay
the same APIs DOSPLUS used. Though to what extent the system would be CP/M
is debatable.
-uso.
BruceMcF
> >A bootable floppy (USB key etc.) needs to be in fat form, but a CP/M
> >image could well be an image file in the fat system - so again the
> >bdos acquires more of the capabilities of a CP/M emulator, but running
> >as the host system.
> That is a limit as CP/M has some hard size limits that most CF or USB
> memory sticks easily exceed and most large cpus (32bit or larger)
> have more addressable ram than 32MB! The problem there that
> the extent based FS doesn't have enough bits to address it.
Yes, that's why:
> >A bootable floppy (USB key etc.) needs to be in fat form, but a CP/M
> >image could well be an image file in the fat system
The BIOS serves CP/M sectors from inside a FAT *file*. The FAT bootup
routine finds the default file image. The minimum information needed
to find the information in the directory entry required to access the
file image is stored in a safe place for the BIOS to find and serve CP/
M sectors to the BDOS. A utility that knows where the file image
bootstrap information and can navigate the FAT filesystem would be
able to change disk images if required.
> >And as in an emulator, if there is a default image that is the
> >original "drive" when a new USB/SD/CD-ROM etc. is started, then if
> >that default image contains a utility that can mount a new image, the
> >utility can handle all the specification of the hierarchical file
> >system.
> Emulators hide the hardware and create directly or indirectly
> a virtual machine and hardware.
Precisely the difference. This does not create a virtual machine, but
rather a BIOS that serves CP/M sectors from within a FAT file.
> >Ah, but if you want separate tasks, then its MP/M. Start with CP/M and
> >work up.
> The BIOS is a driver.
Yes, I glitched and wrote BDOS everywhere I meant BIOS.
> However, with a tiny OS there is nothing to have an overlay os
> running multiple instances of CP/M as individual tasks. That does
> solve the problem nicely of the BDOS (other than MPM) lacking
> the ability to support recusion of itself. If you apportion 60K
> to each instance you have as many as 16 tasks in a megabyte
> with overlying OS and driver space. That would be assuming Z80
> with MMU and 1mb but that is a implmentation limitation. For larger
> CPUs you can grow from their.
When they implement microkernal OS's, the "micro" has all gone away by
the time a POSIX system has been built on top of it. So if one were
cheating, a timeslice microkernal OS that ran instances of the BIOS
could indeed support multiple instances of CP/M BDOS and application
space.
But more in the uber-retro spirit to start out with a single tasking
BIOS running bare on metal and hosting a single tasking BDOS, and
building up to MP/M to get to multi-tasking.
John Elliott
: CP/M Plus probably would be a better choice, except, even less is known
: about CP/M-86 Plus - it looks to me to be a deprecated ccp/m. It
: would've been better to abstract the functions for the bdos and push off
: into the bios the implementation mapping.
I've got a copy for the Apricot PC/Xi:
<http://www.seasip.info/Unix/QDAE/screens.html>
The BDOS is just the same as in Personal CP/M-86 1.x, minus a couple of
bugfixes.
--
John Elliott
Thinks: This is what a nice clean life leads to. Hmm, why did I ever lead one?
-- Bluebottle, in the Goon Show
Bill Buckels
> It is pointless to argue with Roche.
And also with you, and for that matter with me:)
> Finally, my remark about "forgeries".
Nonsense! Generally a "forgery" would be considered a criminal
activity and to suggest that Roche is a criminal is a daffy comment,
so I conclude that you are joking. Repackaging, revising or re-
releasing Roche's own programs by himself is not forgery.
> The bottom line in all this, is that Roche insists on using comp.os.cpm for his own purposes
As do you and as do I...
>The facts are, that he refuses to create or use a Web site, or work with CP/M archiving sites, which are a reasonable venue for his printed and programming work on CP/M. Instead he posts documents here.
Absolute Bullshit (not "facts" surely) from my personal experience
with Roche! A case in point:
http://www.cpm8680.com/turbodos/Tdos1.htm
Welcome to the Wonderfully Ancient World of TurboDOS
A Multiprocessor Operating System for Z-80-based computers
http://www.cpm8680.com/turbodos/index.htm#intro
http://www.cpm8680.com/turbodos/index.htm#history
Website History
In 2009, Mr. Emmanuel Roche, a regular of the comp.os.cpm Newsgroup,
decided to search for this "Lost DOS" (TurboDOS) and found a PDF of a
TurboDOS Newsletter which mentioned Dr. R. Roger Breton (below).
http://www.cpm8680.com/turbodos/Roger.jpg
Dr. Rene Roger Breton (above), now a patent paralegal, has an
extensive computer engineering background. Between 1986 and 1994, he
worked as a computer programmer, technical writer, and engineer in
various jobs. Dr. Breton is a published author of TurboDOS Z80
Assembly Language Programming and TurboDOS 8086-80186-80286 Assembly
Language Programming.
E. Roche managed to find him, and Roger agreed to release his 500-page
book about TurboDOS; Z-80 ASSEMBLY LANGUAGE UNDER TurboDOS.
M. Roche hopes that "this masterwork will lead to the ressurrection of
TurboDOS." My hopes are slightly less ambitious; I hope you enjoy your
visit to the TurboDOS Museum. Roche forwarded the initial material for
this website and Roger forwarded his book. I rummaged around a found
some other bits and pieces including Piergiorgio Betti's Personal
Site. Since this website is a Work in Progress it will be expanded
over time.
...
Roche has said that unlike others who are "selling stuff" he doesn't
have the time and/or money to set up a Web site.
I also agree with Roche that "The last place on Earth where CP/M is
talked about is the comp.os.cpm Newsgroup. So, where else could I
publish the result of my work?".
Keep in mind, Herb, that your hammering-on about your perception of
what this ng is all about is only your opinion just as Roche's
opinions are his opinions and I don't know who is the bigger bully...
you or Roche... or maybe me:) However, I have graciously ignored your
big opinion of what is appropriate for this ng even when your comments
were directed at me, but believe me when I say that there are many
appropriate uses for an online community and not just the ones that
Herb Johnson advocates.
Roche contends that as far as "work with CP/M archiving sites" that it
is you, Herb, who erased his files on your Web site! As far as we
know, BitSavers.Org and Wikipedia mention some of his stuff, and 2 or
3 Web sites host some of his files -- without him asking them to host
them! I find this argument quite agreeable and evidently consistent
with the bullying that goes on between the two of you.
Anyway, both of you should cut-out this arguing at least for now.
And don't go jumping all over me thanks (no petulant ot otherwise
defensive responses please). I have your links from my website and I
am not playing favourites. I don't always like the way Roche pokes
fun, especially when I am a target... but I also think that sometimes
he is having health problesm as are several others in these online
forums and I try to cut slack for that reason alone when I am feeling
good myself.
I like to think that both of you are my friends as are many others in
this ng and other groups. Dassow and many others come to mind as well.
Enough crap about Roche!
Regards,
Bill Buckels
This website is a Work In Progress dedicated to CP/M (Control Program
for Microcomputers) which was (and is) an operating system similar to
MS-DOS but which predated MS-DOS, and which is in fact the operating
system that MS-DOS descended from. But CP/M is much more than simply
an early version of MS-DOS and CP/M is very much alive in online
communities of Enthusiasts and Others who still use CP/M in Old (and
sometimes newer) Computers and Emulators.
http://en.wikipedia.org/wiki/User:Bill_Buckels
http://www.appleoldies.ca/
http://www.c64classics.ca/
http://www.clipshop.ca/c86/
http://www.clipshop.ca/
http://www.teacherschoice.ca/
pbetti
I agree 100%. And is not possible to say Roche is not working with
archiving sites.
If you look here: http://z80cpu.eu/content/view/39/55/ you'll find a
place available to Emmanuel to publish what he want.
Yes, it's true his attitude to go off topic sometimes, but normally
this is not (and should be not) a matter, unless original thread
subject is not sufficient to mantain the discussion opened.
So since THIS thread is really ineteresting please avoid personal
flames...
Piergiorgio Betti
Henk Siewert
First, try to understand that although this is a group where the
English language is the premier language, it is an International
forum.
The only reason for using English is that most Americans cannot use
another language than their dialect of the English language.
Sometimes the use of English is not as you are accustomed to. So there
will be misunderstandings. Sometimes it is not so easy to say what you
mean. And sometimes you are not meaning what you say. If you can
follow me.
Be aware that you would me missing a lot of interesting stuff when we
Europeans would not take the trouble to talk to you in a language you
can understand.
We Europeans can talk to each other because we learn each other’s
language in school.
So I can read Dutch, German, French, English and speak all but French
fluently (I think).
While learning a language one learns a lot of things about the other’s
culture. So we learn to be very lenient about the cultural
differences.
And that is the way it should be in an International forum. Especially
on the Net.
Although the American act as if they invented the WorldwideWeb they
did not.
We Europeans did. Oh yea, did you know that the official spelling of
the WWW = WorldWideWeb. Look it up…
Maybe the Internet is American. But they did not develop it alone. As
an Air Force officer (Royal Netherlands Air Force) I was the computer
man who helped to establish the first, via satellite, computer link
between Europe and NORAD. In 1968! We had a lot to do before we could
establish the link. And a lot of European expertise wends in to the
net.
Why am I writing this? Because I am feeling there is a tension between
Europeans and Americans on the web. A lot of Europeans are not amused
with the way the USA is doing as if the web is theirs. And thinking
English is the natural way of communicating on the web.
So, please let’s be friends, and stop hackling at each other.
If Mr. Emmanuel Roche is doing the things his way, let him. If you are
doing your things your way, and asking money for it (how American),
it’s OK with me.
In this group it is about CP/M. So where else should one place its
work?
Greeting from Holland,
Henk Siewert, Maj. Royal Netherlands Air Force, Ret.
Axel Berger
>Nonsense! Generally a "forgery" would be considered a criminal activity
>and to suggest that Roche is a criminal is a daffy comment,
The English Wikipedia seems to agree with you, perhaps influenced by a
totally lawyer-dominated America. The German "F�lschung" has a wider
meaning with "F�lschung in der Wissenschaft" linked to the bland and
imprecise "scientific misconduct".
To my German-influenced mind forgery is a valid generic term
encompassing falsification, fabrication, misattribution and others in a
factual, not legal, sense.
BruceMcF
> *Bill Buckels* wrote on Thu, 10-02-11 10:29:
> >Nonsense! Generally a "forgery" would be considered a criminal activity
> >and to suggest that Roche is a criminal is a daffy comment,
> The English Wikipedia seems to agree with you, perhaps influenced by a
> totally lawyer-dominated America.
This was the english meaning of the term well before any issue of
America being "lawyer dominated" ever emerged. Forgery is an effort to
deceive with the intent to defraud.
Fabricated would be an example of a broader term without the
connotation of being illegal by design.
BruceMcF
> On Feb 10, 4:12 pm, BruceMcF <agil...@netscape.net> wrote:> On Feb 10, 3:10 pm, s_dubrov...@yahoo.com wrote:
> > > By the time you make all the allowances for memory management, and
> > > 32bit or 64bit stuff, it would hardly be recongnizable as cp/m.
> As to your following comments.. the memory management for CP/M-86 is
> hidden in the BDOS with very disappointing results, badly considered
> algorithms, even mp/m-86 and ccp/m-86 did things differently, tho with
> some backward support to mem fns for cp/m-86 programs.
I glitched there, I should have written BIOS. The BIOS serves expanded
memory as a RAMdisk.
> CP/M Plus probably would be a better choice, except, even less is
> known about CP/M-86 Plus - it looks to me to be a deprecated ccp/m.
From what I understand about the PCP/M concept, what's critical here
is the API, not the code. As long as the PCP/M BIOS when compiled for
a 8080 or Z80 system supports a CP/M 2.2 BDOS, and the PCP/M BDOS when
compiled for an 8080 or Z80 can use an existing CP/M 2.2 system BIOS,
then the "portability" is that they can also both be compiled and run
on a processor that has never before hosted CP/M.
An interesting application would be a microcontroller descended from
the Z80 with four of its I/O lines dedicated to an SD socket accessed
via the SPI protocol. The firmware boot-up would find default CPM
image file, load the BIOS and BDOS, and the BIOS would take over from
there.
MikeS
Goodness, you're starting to sound like our French Luser... who are
these "self declared "educated" English people" who say those things?
Do you have any references to that usage?
FWIW I was taught that "me" in the first example is most definitely
technically incorrect, although it is indeed often used and generally
accepted because the single sound "I" is confusing and sounds very
awkward as an answer to a question; "I do/am/will,etc." are usually
preferred.
As to the rest of your examples I have never heard anyone use "I" as
the answer, although people will sometimes misuse "I" instead of "me"
when it is combined with another object (i.e., "She accused he and I
of stealing), probably as a result of being taught that using "me" in
a sentence where you should use "I" is a sign of a poor education.
BTW, the original sentence should have read "Unfortunately *for*
them".
Having said all that, I am always astounded and impressed by the
excellent quality of the English used by non-English-speaking folks
who contribute posts to groups like this, often superior to native
speakers.
My (virtual) hat's off to all of you!
mike
MikeS
If you are really serious about "being friends and stop[ping]
hackling," then perhaps you could start by refraining from these
judgemental generalizations about "Europeans" and "Americans".
Granted, Americans, Canadians, Australians, NewZealanders, English,
Irish and Scots, and many others have an advantage by speaking English
natively and as I say elsewhere most of us appreciate and respect the
fact that so many non-English speaker use English internationally, and
as well as they often do, even if it was unfortunately largely an
economic and practical necessity.
I don't think that all Europeans are as multi-lingual as you claim to
be, and I suspect that in many cases English is not only used to
communicate with those unilingual "Americans," (as opposed to
Australians, Canadians, etc.) but is often also used as a lingua
franca among non-English speaking people of different languages.
Still, while I don't have to look very far through the archive to find
insulting, disparaging, judgemental and even racist remarks directed
at "Americans" by "Europeans," I don't find any by Americans directed
at non-Americans except a few in response to some of M. Roche's more
outrageous insults.
And if we're going to judge people on the basis of the continent of
their birth, I suppose we could also look at Europe's continuing
history of intolerance and judgemental nationalism among its own
nations, not just towards the big bad 'USA'
What does all this have to do with our common hobby of old computers
and CP/M? I don't expect M Roche to cease his abusive invective, but
it would be nice if those not suffering from his problems would
refrain from judgementally talking about "Americans," or any other
nationality, race, etc.
FWIW, I'm not American and in fact "European " by birth.
mike
MikeS
It seems to me that words like forgery originated in England, so I
don't see what American lawyers have to do with anything, nor where
Wikipedia's definition states that it is specifically an illegal
"criminal" activity:
"Forgery is the process of making, adapting, or imitating objects,
statistics, or documents (see false document), with the intent to
deceive."
But whether it's justified or not, I think what Herb meant was
"plagiarism."
BruceMcF
> It seems to me that words like forgery originated in England, so I
> don't see what American lawyers have to do with anything, nor where
> Wikipedia's definition states that it is specifically an illegal
> "criminal" activity:
It certainly carries the connotation of being a criminal act. Remember
that wikipedia is not an authoritative source. A better source for
definitions would be, for example, Merriam Webster online:
1 archaic : invention
2 : something forged
3 : an act of forging; especially : the crime of falsely and
fraudulently making or altering a document (as a check)
Bill Marcum
> From what I understand about the PCP/M concept, what's critical here
> is the API, not the code. As long as the PCP/M BIOS when compiled for
> a 8080 or Z80 system supports a CP/M 2.2 BDOS, and the PCP/M BDOS when
> compiled for an 8080 or Z80 can use an existing CP/M 2.2 system BIOS,
> then the "portability" is that they can also both be compiled and run
> on a processor that has never before hosted CP/M.
>
> An interesting application would be a microcontroller descended from
> the Z80 with four of its I/O lines dedicated to an SD socket accessed
> via the SPI protocol. The firmware boot-up would find default CPM
> image file, load the BIOS and BDOS, and the BIOS would take over from
> there.
I subscribe to the Club 100 mailing list, dedicated to the TRS-80 Model
100 and related computers. One of the users there created a device called
the NADSbox, which uses a microcontroller to connect an SD card to an
RS-232 port, emulating the Tandy portable disk drive. Details can be
found at http://club100.org or http://www.kenpettit.com/nadsbox.html
BruceMcF
> I subscribe to the Club 100 mailing list, dedicated to the TRS-80 Model
> 100 and related computers. One of the users there created a device called
> the NADSbox, which uses a microcontroller to connect an SD card to an
> RS-232 port, emulating the Tandy portable disk drive. Details can be
> found athttp://club100.orgorhttp://www.kenpettit.com/nadsbox.html
There's a similar device that allows a SD card to emulate a 1541 drive
for a Commodore 64.
In this application, rather than the microcontroller serving the SD
card to an external device, its using the SD card for its own purposes.
James Moxham (Dr_Acula)
>An interesting application would be a microcontroller descended from
>the Z80 with four of its I/O lines dedicated to an SD socket accessed
>via the SPI protocol. The firmware boot-up would find default CPM
>image file, load the BIOS and BDOS, and the BIOS would take over from
>there.
There are a handful of board designs that can do SD card access. There
are not very many probably because SD code is quite complicated. The
N8VEM group has one person who managed to write such code and the
assembly source is available on the website, but that person has
disappeared and no-one else has managed to replicate its brilliant
simplicity.
For CP/M applications, SD cards are as fast as hard drives and much
faster than floppy drives. And you can get gigabytes of storage on an
SD card, so the size of hard drives is limited by CP/M. I've got a
little micro SD card the size of my thumbnail that cost $7 and has
eight 8mb CP/M hard drive images on it.
So - if you want to use an SD card the first task is to find some
driver software. The next task is to understand how to use that
software! The only example I've found where it was possible to do both
is the Propeller chip, and indeed it ended up being easier to use a
Z80 software emulation than it did to write SD driver code for a real
Z80. http://www.smarthome.viviti.com/propeller Does exactly what
BruceMcF suggests - it finds the drive image A> on the SD card, reads
CP/M off that image into memory and runs it.
But I have no idea how you would do that with the hardware of a modern
PC, even allowing for the fact that a modern PC has all the hardware
you need. I'm intrigued by the discussion earlier about 8086 opcodes
and CP/M. My understanding is that pretty much all CP/M programs are
8080 opcodes only (with exceptions like BBC Basic), and that even when
the 8080 opcodes are written with Z80 mnemonics, they compile to 8080
opcodes. And the 8080 branched into two - into Z80 and 8086. Does that
mean that 8080 opcodes are buried inside x86 processors? Like the old
ones, and even the modern ones?
BruceMcF
<moxh...@internode.on.net> wrote:
> There are a handful of board designs that can do SD card access. There
> are not very many probably because SD code is quite complicated.
As long as you access it in SPI mode, SD access seems to be
straightforward. I'd suppose the dedicated SD modes are more
complex, but I've never worried about them.
> So - if you want to use an SD card the first task is to find some
> driver software. The next task is to understand how to use that
> software! The only example I've found where it was possible to do both
> is the Propeller chip, and indeed it ended up being easier to use a
> Z80 software emulation than it did to write SD driver code for a real
> Z80.http://www.smarthome.viviti.com/propeller Does exactly what
> BruceMcF suggests - it finds the drive image A> on the SD card, reads
> CP/M off that image into memory and runs it.
> But I have no idea how you would do that with the hardware of a modern
> PC, even allowing for the fact that a modern PC has all the hardware
> you need.
A PCP/M BIOS for a modern PC would use the hardware resources on the
motherboard - including the motherboard BIOS.
> I'm intrigued by the discussion earlier about 8086 opcodes and CP/M.
> My understanding is that pretty much all CP/M programs are
> 8080 opcodes only (with exceptions like BBC Basic), and that even when
> the 8080 opcodes are written with Z80 mnemonics, they compile to 8080
> opcodes. And the 8080 branched into two - into Z80 and 8086. Does that
> mean that 8080 opcodes are buried inside x86 processors? Like the old
> ones, and even the modern ones?
The 8086 descended from the 8080, and it is assembly language
compatible but not identical binary opcodes. With the right 8086
assembler, 8080 assembly language programs can be assembled, but they
won't result in identical binary files.
A lot of the more capable raplcements for the CP/M BDOS were Z80 based.
Bill Marcum
> But I have no idea how you would do that with the hardware of a modern
> PC, even allowing for the fact that a modern PC has all the hardware
> you need. I'm intrigued by the discussion earlier about 8086 opcodes
> and CP/M. My understanding is that pretty much all CP/M programs are
> 8080 opcodes only (with exceptions like BBC Basic), and that even when
> the 8080 opcodes are written with Z80 mnemonics, they compile to 8080
> opcodes. And the 8080 branched into two - into Z80 and 8086. Does that
> mean that 8080 opcodes are buried inside x86 processors? Like the old
> ones, and even the modern ones?
>
emulation mode, but modern x86 processors don't have that. Still, they
can run emulators much faster than an original 8080 or Z80.
s_dub...@yahoo.com
Aha! One sighted in the wild!
So then Personal CP/M-86 is the place to start.
Steve
s_dub...@yahoo.com
> On Feb 10, 8:41 pm, s_dubrov...@yahoo.com wrote:
>
> > On Feb 10, 4:12 pm, BruceMcF <agil...@netscape.net> wrote:> On Feb 10, 3:10 pm, s_dubrov...@yahoo.com wrote:
> > > > By the time you make all the allowances for memory management, and
> > > > 32bit or 64bit stuff, it would hardly be recongnizable as cp/m.
> > As to your following comments.. the memory management for CP/M-86 is
> > hidden in the BDOS with very disappointing results, badly considered
> > algorithms, even mp/m-86 and ccp/m-86 did things differently, tho with
> > some backward support to mem fns for cp/m-86 programs.
>
> I glitched there, I should have written BIOS. The BIOS serves expanded
> memory as a RAMdisk.
>
> > CP/M Plus probably would be a better choice, except, even less is
> > known about CP/M-86 Plus - it looks to me to be a deprecated ccp/m.
>
> From what I understand about the PCP/M concept, what's critical here
> is the API, not the code. As long as the PCP/M BIOS when compiled for
> a 8080 or Z80 system supports a CP/M 2.2 BDOS, and the PCP/M BDOS when
> compiled for an 8080 or Z80 can use an existing CP/M 2.2 system BIOS,
> then the "portability" is that they can also both be compiled and run
> on a processor that has never before hosted CP/M.
>
I guess, well the API shifts alittle even among cp/m-80 versions.
> An interesting application would be a microcontroller descended from
> the Z80 with four of its I/O lines dedicated to an SD socket accessed
> via the SPI protocol. The firmware boot-up would find default CPM
> image file, load the BIOS and BDOS, and the BIOS would take over from
> there.
Yes, but the OP framed his need around a modern PC.
Steve
BruceMcF
> > An interesting application would be a microcontroller descended from
> > the Z80 with four of its I/O lines dedicated to an SD socket accessed
> > via the SPI protocol. The firmware boot-up would find default CPM
> > image file, load the BIOS and BDOS, and the BIOS would take over from
> > there.
> Yes, but the OP framed his need around a modern PC.
Its probably easier to start with a modern PC, but that does not imply
its necessary to stop there.
s_dub...@yahoo.com
<moxh...@internode.on.net> wrote:
> Re Bruce McF
[snip]
>
> But I have no idea how you would do that with the hardware of a modern
> PC, even allowing for the fact that a modern PC has all the hardware
> you need. I'm intrigued by the discussion earlier about 8086 opcodes
> and CP/M. My understanding is that pretty much all CP/M programs are
> 8080 opcodes only (with exceptions like BBC Basic), and that even when
> the 8080 opcodes are written with Z80 mnemonics, they compile to 8080
> opcodes. And the 8080 branched into two - into Z80 and 8086. Does that
> mean that 8080 opcodes are buried inside x86 processors? Like the old
> ones, and even the modern ones?
No, the opcodes are not identical. 8080 NOP is 00h, 8086 NOP is 90h,
for example.
But there is close functional equivalents, 8080 NOP -> 8086 NOP, so
that .asm code can be translated into .a86 code. DRI made such a
translator: XLT86.COM which runs under CP/M-80 to emit ASM86 assembler
syntax from ASM syntax, that is then assembled under cp/m-86.
(actually there was a cp/m-80 hosted asm86 assembler too.)
This was practical as CP/M-86 and CP/M-80 used the same file system on
diskette.
Make sure you get the DRI XLT86, there is a clone version which is
somewhat buggy which runs on pcdos. I believe it is on Gaby's site.
There is a very fine DRI document for XLT86 which covers the details
of the translation strategy.
There is one processor which has 8080 machine code capability and that
is the NEC V20 cpu, which is otherwise a clone of the ia186 cpu.
There is a machine code instruction to 'shift' into 8080 mode. - but
not z80.
Steve
MikeS
Well, I *was* replying to the comment that Wikipedia, no doubt
influenced by American lawyers, defined forgery as specifically
a criminal act; of course you'll find various other definitions if
you
look for them. FWIW, none of my print dictionaries including
the New Webster's and two versions of the Oxford mention
"crime" at all in their definitions; the "crime" would be fraud.
Like most words it has several meanings and most people would
be able to determine the correct one from the context, but if the
objective is to challenge and attack Herb instead of taking his
intended meaning then of course you'll choose the connotation
that best suits your purpose, no matter how inappropriate.
I'd say it's purposely a real stretch to interpret Herb's remark
as calling M Roche a "criminal."
Interesting that M Roche's offensive insults and name-calling are
considered as just "poking fun," and acceptable expressions
of his "opinions"; I suspect that Allison, his favourite target, is
not
amused except perhaps by his silliness. But when someone chides
him for his self-serving rudeness, cluttering up the list with page
after page of listings and documents that would be less annoying
and more useful posted on a web site somewhere, etc., that does
merit people jumping on Herb for expressing *his* opinions...
Enough indeed about Roche and Herb, but it has been an interesting
insight into some of the attitudes on this list...
'nuff said.
Roger Ivie
> As long as you access it in SPI mode, SD access seems to be
> straightforward. I'd suppose the dedicated SD modes are more
> complex, but I've never worried about them.
I'll second that. I've rolled my own code that can access SD and SDHC
cards on an ARM; it's not hard.
--
roger ivie
ri...@ridgenet.net
Rod Pemberton
news:94935b8f-a084-4d5e...@g27g2000yqh.googlegroups.com...
>
> There is one processor which has 8080 machine code capability and that
> is the NEC V20 cpu, which is otherwise a clone of the ia186 cpu.
> There is a machine code instruction to 'shift' into 8080 mode. - but
> not z80.
>
Ah...
I didn't know that (or it was long forgotten...). It seems RBIL's
86BUGS.LST documents this. You could use both CP/M OS and code and DOS
development tools on the same machine. I'd think it'd be very good for you
to have one of them (the NEC cpu). After checking the old posts, it seems
you may have one. Yes? Congrats! I bet that makes your life a bit easier.
Hmm, it seems the V20/V30 uses bit 15 in eflags... I wonder if that might
help explain the oddity of x86's bit 15 in eflags. I'll have to look at
where the extra V20 instructions fit into the post 186, x86 map. Nah, it
looks like they (V20/V30 specific x86 mode instructions) collide with
various generations of x86 instructions. Although, it does seem that many
didn't collide until SSE or SSE3.
Rod Pemberton
PS. I noticed another version of SmallC from you. Unfortunately, I've got
so many personal projects, the one I'm working on (with the help of your
versions) has been on the back burner for a while... So, it's probably all
up to you to figure out that parser. Sorry. If the parser wasn't somewhat
compact, I'd probably just scrap it - because I'd understand whatever I
wrote better... Maybe you could ask around if someone understands SmallC's
recursive descent parser. You just need someone to figure out how to add-in
structs, a couple logical operators, and what else?
Peter Dassow
>> PS: Miguel, if you produce "new" sources, please ... Provide them to others.
>
> Of course. I had a website with some of my programs. I am looking for
> a new hosting. My small-c version for CP/M-80 (MESCC) still somewhere on
> the net.
Yes. I mirrored it at http://www.z80.eu/c-compiler.html ...
Regards
Peter
Floppy Software
Yes, I know, Peter. Thank you very much!!!
Floppy Software
> From what I understand about the PCP/M concept, what's critical here
> is the API, not the code. As long as the PCP/M BIOS when compiled for
> a 8080 or Z80 system supports a CP/M 2.2 BDOS, and the PCP/M BDOS when
> compiled for an 8080 or Z80 can use an existing CP/M 2.2 system BIOS,
> then the "portability" is that they can also both be compiled and run
> on a processor that has never before hosted CP/M.
I agree with that.
That's why I asked about Portable CP/M sources in C.
I will try to modify the C source code of CP/M 68K & Z8K to be
ANSI C.
As I told, I'll start with PIP program.
If I have success with PIP, then I'll try with STAT, ED and BDOS.
s_dub...@yahoo.com
> <s_dubrov...@yahoo.com> wrote in message
>
> news:94935b8f-a084-4d5e...@g27g2000yqh.googlegroups.com...
>
>
>
> > There is one processor which has 8080 machine code capability and that
> > is the NEC V20 cpu, which is otherwise a clone of the ia186 cpu.
> > There is a machine code instruction to 'shift' into 8080 mode. - but
> > not z80.
>
> Ah...
>
> I didn't know that (or it was long forgotten...). It seems RBIL's
> 86BUGS.LST documents this. You could use both CP/M OS and code and DOS
> development tools on the same machine. I'd think it'd be very good for you
> to have one of them (the NEC cpu). After checking the old posts, it seems
> you may have one. Yes? Congrats! I bet that makes your life a bit easier.
>
wrapper coding. I do use both DOS and CP/M-86 oses on it and its hard
drive.
> Hmm, it seems the V20/V30 uses bit 15 in eflags... I wonder if that might
> help explain the oddity of x86's bit 15 in eflags. I'll have to look at
> where the extra V20 instructions fit into the post 186, x86 map. Nah, it
> looks like they (V20/V30 specific x86 mode instructions) collide with
> various generations of x86 instructions. Although, it does seem that many
> didn't collide until SSE or SSE3.
>
> Rod Pemberton
> PS. I noticed another version of SmallC from you. Unfortunately, I've got
> so many personal projects, the one I'm working on (with the help of your
> versions) has been on the back burner for a while... So, it's probably all
> up to you to figure out that parser. Sorry. If the parser wasn't somewhat
> compact, I'd probably just scrap it - because I'd understand whatever I
> wrote better... Maybe you could ask around if someone understands SmallC's
> recursive descent parser. You just need someone to figure out how to add-in
> structs, a couple logical operators, and what else?
alot more..
AIR you were looking for C syntax tests. There are syntax test
programs found in:
cpm68k1.zip \v102a\al40\test test.c,test.h
'Beginning 68000 C Compiler Test'
The function form is K&R but otherwise it pretty far advanced. It's
not your fathers' simple c anymore.
..the latest version I posted of the small-c for nasm is derived from
the CP/M-86 version, which does two main things, separates code and
data, and expands the lval[] array by 1 element. I've made the
adjustments for nasm & self compile, added stub prolog,epilog
%includes, changed the literals output from db decimals to db 'string
lits',0 so the nasm output is alot more readable. Otherwise, the
parser isn't improved much, I've been gathering ideas and slowly
sifting thru them.
Steve
BruceMcF
> Well, I *was* replying to the comment that Wikipedia, no doubt
> influenced by American lawyers, defined forgery as specifically
> a criminal act;
Aha, "seems to suggest" as equal to "defined as".
Yes, it is *only* suggested and *not* stated in the Wikipedia,
but then Wikipedia is not an actual dictionary. Its stated as
one of the senses of the word in any reasonably complete
English language dictionary.
MikeS
Well, since none of my dictionaries define forgery as a "crime"
and yours does, I guess as usual in this sort of discussion
your definition of a "reasonably complete .. dictionary" is one
that supports your viewpoint...
BTW, I just read the Wikipedia entry and it actually does *not*
"seem to suggest" at all that forgery is a crime; in fact it states
quite explicitly that the "crime" is "fraud." Of course forgery could
be used to commit fraud, as could lying, misrepresentation, and
in fact any kind of deception, but that does not by definition make
them criminal acts; they could be hoaxes, jokes, tests, or other
non-crimes.
But if you're truly determined and really *want* to believe that *in
this context* Herb was calling M Roche a "criminal," i.e. someone
the Surete would be interested in, then I doubt that I'll be able to
dissuade you and I'll have to leave you with that belief; I'll choose
to believe that he meant misrepresentation and perhaps plagiarism.
Apologies for the tedious OT digression; now back to more
productive discussions...
mike
BruceMcF
> But if you're truly determined and really *want* to believe that *in
> this context* Herb was calling M Roche a "criminal," i.e. someone
> the Surete would be interested in, then I doubt that I'll be able to
> dissuade you and I'll have to leave you with that belief; I'll choose
> to believe that he meant misrepresentation and perhaps plagiarism.
I never said anything of the sort. I find it far more likely that
Herb was speaking loosely than that he actually intended
to convey what it sounded like.
Bill Buckels
If he was speaking loosely (as in verbal diarrhea [from the Greek,
"diarrhoia" meaning "a flowing through"]) in criticism of M.Roche's
alleged questionable publication techniques and we have all heard
Herb's rants about Roche before, then perhaps Herb was effectively
indulging in his own form of "forgery" since he seems to repeat what
he has already said over the years about Roche so Herb is also
creating new copies of the same-old ad nauseum. However, Roche seems
to usually stick to topics related somewhat to his use of CP/M whereas
Herb Johnson rants about Roche which is woefully off-topic in this ng
AFAIK and when he points a finger at Roche the other 3 are pointing
back at himself (providing of course that he has not had the other 3
already chopped-off from finger pointing - pinocchio's nose is not the
only pointed object to be avoided, metaphorically speaking of course.)
> On Feb 10, 12:27 pm, Herbert Johnson <herbrjohn...@gmail.com> wrote:
>
> > It is pointless to argue with Roche.
Then perhaps we should all quit pointing.
Now as far as this discussion of "criminality" as related to
forgeries, my comment in response to Herb's remark clearly concludes
that Herb is joking... and I don't believe that connotations of
criminality were ever an issue on the part of matter of the two grumpy
old men in question... or perhaps all the grumpy old men, me
included:)
On Feb 11, 3:29 am, Bill Buckels <bbuck...@escape.ca> wrote:
> > Finally, my remark about "forgeries".
>
> Nonsense! Generally a "forgery" would be considered a criminal
> activity and to suggest that Roche is a criminal is a daffy comment,
> so I conclude that you are joking. Repackaging, revising or re-
> releasing Roche's own programs by himself is not forgery.
> Enough crap about Roche!
And enough crap about Herb!
Regards,
Bill Buckels
http://en.wikipedia.org/wiki/User:Bill_Buckels
http://www.clipshop.ca/c86/
http://www.clipshop.ca/
http://www.teacherschoice.ca/
http://www.grindstoneharbour.com/
http://www.appleoldies.ca/
http://www.c64classics.ca/
http://www.cpm8680.com/
"In 1974, Dr. Gary A. Kildall, while working for Intel Corporation,
created CP/M as the first operating system for the new microprocessor.
By 1977, CP/M had become the most popular operating system (OS) in the
fledgling microcomputer (PC) industry. The largest Digital Research
licensee of CP/M was a small company which had started life as Traf-0-
Data, and is now known as Microsoft. In 1981, Microsoft paid Seattle
Software Works for an unauthorized clone of CP/M, and Microsoft
licensed this clone to IBM which marketed it as PC-DOS on the first
IBM PC in 1981, and Microsoft marketed it to all other PC OEMs as MS-
DOS." (The First PC Operating System)
In 1982, Bill Gates said in an interview with PC Magazine "the
greatest installed base of CP/M-80 machines are the users of Microsoft
softcards which plug into Apple computers" (The Man Behind The
Machine).
Today, over a quarter century later and although they are not
plentiful, new and used Softcards and Softcard clones, and Apple II
computers are still used, bought and sold.
MikeS
<snippage>
> Now as far as this discussion of "criminality" as related to
> forgeries, my comment in response to Herb's remark clearly concludes
> that Herb is joking... and I don't believe that connotations of
> criminality were ever an issue on the part of matter of the two grumpy
> old men in question... or perhaps all the grumpy old men, me
> included:)
<more snippage>
-------------
That's exactly my point; whatever we may think of Herb's opinions
I don't think he *was* joking, but you are (IMHO) unfairly choosing
an interpretation that is obviously ridiculous in this context in
order
to make fun of and therefore dismiss what he said. Frankly I don't
see much difference between (or point in) Herb's rant about Roche
and your rants about Herb...
For the benefit of people like Axel, in a different context "forgery"
might well be an example of fraud, but strictly speaking it does
indeed
mean essentially what your German translation suggests; "Faelschung"
is probably a crime if it involves currency, for example, but that
does not
make it a criminal act per se (American lawyers notwithstanding).
Wikipedia and my dictionaries suggest, "Forgery is the process of
making, adapting, or imitating objects, statistics, or documents ...
with
the intent to deceive," which IMHO is exactly what Herb meant; it
does
not specifically mean "with the intent to *defraud*," as Bill and
Bruce
suggest, although of course forgeries are often used to defraud
someone.
Having said all that (and more), I agree that Herb could have used a
better
word; I would also like to suggest to Herb that he take his own advice
and
try to curb the urge to respond to Roche's "troll"s; ironic that this
lengthy,
not to be taken *too* seriously and hopefuly friendly OT discussion
grew
out of that.
If you (Bill) or Bruce want to continue this, maybe it'd be better off-
list...
mike
Floppy Software
No interest in CP/M, then?
Herbert Johnson
"forgery". I regret using the word. There seems to be more interest in
some issue about intent to create a forgery, than the point I made
about confusing copies with originals, irrespective of intentions -
and a side issue, anyway. Otherwise, I see as a common theme,
"intentions" versus "consequences". Responsible people consider the
consequences of their behavior, versus whatever their intentions are;
and they generally have the capacity to anticipate consequences.
> I would also like to suggest to Herb that he take his own advice
> and try to curb the urge to respond to Roche's "troll"s; ironic that this
> lengthy, not to be taken *too* seriously and hopefully friendly OT discussion
> grew out of that.
I agree about taking my own advice. I've posted no follow-ups since my
last two posts. I made my points, and mostly they have not been
disputed, and I'll simply acknowledge the disputes are mostly
reasonable. I note on my site, and often when I respond to Roche, the
risk of responding to a troll as such is to become one, and to attract
others. I apologize for my part, in contributing to an unproductive
and unhappy distraction from an otherwise productive discussion
thread. That was a "consequence" of my "intentions" to discourage
other rude and inappropriate postings. A specific apology to "Floppy
Software", who commented in effect this side discussion was
inappropriate; please note the bulk of the thread was useful and on-
topic.
Herb Johnson
retrotechnology.com
MikeS
<snip>
> I apologize for my part, in contributing to an unproductive
> and unhappy distraction from an otherwise productive discussion
> thread. That was a "consequence" of my "intentions" to discourage
> other rude and inappropriate postings. A specific apology to "Floppy
> Software", who commented in effect this side discussion was
> inappropriate; please note the bulk of the thread was useful and on-
> topic.
>
> Herb Johnson
> retrotechnology.com
And I too apologize to "Floppy Software" and others for my part in
dragging it on.
mike
Bill Buckels
> And I too apologize to "Floppy Software" and others for my part in
> dragging it on.
>
Hi Mike,
I apologize to M. Roche and Herb Johnson for ridiculing their
behaviour. I agree with Herb that the bulk of this thread was on-topic
and disagree with Floppy that it was "to convert this thread in a
shit" ... this struggle between Herb and Roche needs to stop. I was
saddened when Jack Crenshaw left this list and I hope that
moderatoritis will not cause M. Roche to withdraw.
If you feel the further need to go on and on, feel free to discuss
"the pot calling the kettle black" off-list...
Bill
Axel Berger
>Do you have any references to that usage?
Not ones I've retained to refer back to as proof, but things like "he
invited my wife and I to dinner" are seen more and more frequently.
Some years ago it used to be the other way round, and "of course it was
me who had to do the work" was wrong in all languages but English. It
seems to me, the frowning on "me" is American, not English, in origin
but I'm not sure.
>probably as a result of being taught that using "me" in a sentence where
>you should use "I" is a sign of a poor education.
Yes, that's exactly my impression.
Perhaps having learnt Latin and loved it makes me more sensitive to
these things than some.
--
Tsch� wa
Axel
James Moxham (Dr_Acula)
>Not ones I've retained to refer back to as proof, but things like "he
>invited my wife and I to dinner" are seen more and more frequently.
All I can say is that English is totally illogical.
There is the proper "Queen's English". But only some people talk like
that.
Here in Australia you might hear
"he invited my wife and I to dinner"
or
"he invited me wife and I to dinner"
or
"he invited me and me wife to dinner"
or even
"youse lot wanna come round for dinner?" [you singular=you plural. So
people invented 'youse' for many people]
If you spoke like that someone who is good with accents could not only
work out what city you are from, but even what suburb in that city. My
dad met someone in Cambridge, UK, who was able to do that.
In Australia some of that incorrect English comes from Cockney in
London, and see if you can work out any logic at all in the way they
speak http://www.bbc.co.uk/dna/h2g2/alabaster/A649
The above becomes;
"he invited me and the trouble and strife to dinner"
And if that doesn't make sense, see what the Chinese are doing to
Engrish http://www.engrish.com/
I agree. Bring back Latin. At least all the rules for endings on words
were logical and consistent.
Back on topic...
Re Floppy Software
>Thank you very much to convert this thread in a shit.
>No interest in CP/M, then?
Lots of interest. It is quite a challenge to think of CP/M on a modern
PC without being allowed to use an emulator. I did some research last
night on the BIOS commands for a typical new motherboard. A lot of the
hard work is in the BIOS - not in an 'operating system'. eg you need a
USB driver in BIOS otherwise you can't talk to a USB mouse and if you
can't use the mouse it is hard to install an operating system. Ditto
drivers for CD rom, hard drivers and keyboard and display.
Somewhere in the BIOS I think it goes and looks for a 'boot device'
and I think you can set the order eg, cd drive, hard drive 1, hard
drive 2 etc. Next thing is to look for something on that device. I saw
a comment somewhere that the first operating system to do this was....
CP/M. So a little tick there for CP/M. So - at the very simplest
level, set a hard drive as the boot device, find some bytes on a hard
drive and execute them. What you do next is the complicated bit - do
you load in something in C or do you try to run just 8080 opcodes.
Oh, and you will probably be into "plug and pray" mode with interrupts
and devices - even those on the motherboard, as windows handles all
the potential interrupt clashes http://en.wikipedia.org/wiki/BIOS_interrupt_call
I found this with a list of bios calls
ftp://ftp.embeddedarm.com/old/saved-downloads-manuals/EBIOS-UM.PDF
The propeller emulation uses a similar concept with Out and In to
various ports which are trapped and go and do things like get a
keyboard character from a PS2 keyboard or put a character on the
screen.
I'm still not sure why an emulator is not allowed. I could set up an
autoexec in most versions of windows that would make the screen go
blue with white text as soon as you boot up, make the dos window full
screen, then run the Altair SIMH with a whole lot of settings that
attach the SIMH to all the ports on a PC, and boot into CP/M directly
and you would never know it was windows behind the scenes. It would
look and behave like a 200Mhz CP/M machine. If you stripped down
Windows so nothing was running in the background (ctrl-alt-del to
check), I think windows would be very stable.
But I am very interested if someone has got Portable CP/M working.
BruceMcF
Its allowed, its just (1) already solved and (2) not the same thing.
Its like asking why you are not allowed to turn over while swimming
the backstroke. You are allowed to turn over, but if you do, you are
not swimming the backstroke anymore and have to compete in a different
event.
glen herrmannsfeldt
> On Feb 15, 11:44?pm, "James Moxham (Dr_Acula)"
> <moxh...@internode.on.net> wrote:
>> I'm still not sure why an emulator is not allowed.
> Its allowed, its just (1) already solved and (2) not the same thing.
I understand the restriction, but it isn't always so convincing.
Well, maybe in the case of the 8080, but for larger processors
that were traditionally microprogrammed you have to explain why
the microprogrammed machine is not an emulator. Does it matter
who built the machine? Is the IBM P/370 a real IBM 370 or is
it IBM's emulation of a 370?
Is an FPGA based hardware implementation emulation or not?
> Its like asking why you are not allowed to turn over while swimming
> the backstroke. You are allowed to turn over, but if you do, you are
> not swimming the backstroke anymore and have to compete in a different
> event.
-- glen
Mr Emmanuel Roche, France
> All I can say is that English is totally illogical.
>
> There is the proper "Queen's English".
> But only some people talk like that.
Hahaha! The 28 August 2009, I wrote to an Englishman, on the
comp.sys.asmstrad.8bit Newsgroup, who happens to be member of a group
interested in the "right pronunciation" of the English language... I
wrote him, saying:
1) it is the people speaking it who define the pronunciation of the
language. The more foreigners will speak it, the more different it
will be.
2) and, anyway, English is one of the worst language on Earth since
(contrary to French) it is not a phonetic language. When you read
English, there is no way to know how to pronunce it, and even less
"rightly" or "correctly"! I gave as an example a famous poem, that
99.99% of native English speakers are unable to pronunce correctly:
http://www.spellingsociety.org/journals/j17/caos.php
Notice the words in Italics: *EACH ONE* needs to be pronunced
differently than the way it appears, printed. Have a try. It is
estimated that only a handful of native English Phonetics Professors
are able to pronunce this poem without making a mistake...
> I agree. Bring back Latin.
> At least all the rules for endings on
> words were logical and consistent.
YES! Since Europe is trying to find common laws for its 500-Million
citisens, I am of the opinion that it should go back to its roots:
Greece and Roma. The French, Spanish, and Italian languages are direct
descendants from the Latin language. (The oldest laws in France are
descending directly from the Latin.) Personally, I advocate going back
to Latin as the official language of Europe. (Today, only the Vatican
still use Latin as its official language.)
Yours Sincerely,
Mr. Emmanuel Roche, France
Peter Dassow
>> I agree. Bring back Latin.
>> At least all the rules for endings on
>> words were logical and consistent.
>
> YES! Since Europe is trying to find common laws for its 500-Million
> citisens, I am of the opinion that it should go back to its roots:
> Greece and Roma. The French, Spanish, and Italian languages are direct
> descendants from the Latin language. (The oldest laws in France are
> descending directly from the Latin.) Personally, I advocate going back
> to Latin as the official language of Europe. (Today, only the Vatican
> still use Latin as its official language.)
Even if I'm german I have to notice the most common *living* language is
English, Spanish or may be Chinese ;-) French is far away being spoken
so widely (and so german also). And Latin is really dead.
So what's the point ?
This is comp.os.cpm, not alt.linguistic or similar...
Regards
Peter
Mr Emmanuel Roche, France
> And Latin is really dead.
http://la.wikipedia.org/wiki/Pagina_prima
(Since this is the comp.os.cpm Newsgroup, notice that there are 36,000
articles available, today, in 2010, in the Latin language on this
Latin version of Wikipedia... Now, could you tell me how many CP/M
articles there are, on a CP/M-centric version of Wikipedia?)
Steve Nickolas
So it would be best to say not that Latin is dead, but that Latin is
*undead*.
-uso.
Peter Dassow
>
> http://la.wikipedia.org/wiki/Pagina_prima
>
> (Since this is the comp.os.cpm Newsgroup, notice that there are 36,000
> articles available, today, in 2010, in the Latin language on this
> Latin version of Wikipedia... Now, could you tell me how many CP/M
> articles there are, on a CP/M-centric version of Wikipedia?)
Do you really tested a few of the "36000" articles there ?
I typed in "Windows" (because of the popularity of this term) and I got
only a skeleton article about Microsoft Windows with a few headlines.
If all articles are of this quality, you can forget la.wikipedia.org.
And of course latin is still used, but it's a dead language because
there is no further development of the language, and nobody really
speaks latin in his normal course of life.
So it's ridiculous to talk further about latin here.
Regards
Peter
Axel Berger
>but it's a dead language because there is no further development of the
>language,
If that's your criterion, then Latin was dead and artificial at the
time of Gaius J. Caesar, who wrote an accomplished Latin grammar while
on his campaigns. It is this artificiality eschewing the depradations
of native speakers ("Handy", etwas erinnern, wegen wem, he invited my
wife and I) that makes Latin so valuable and logical.
Axel Berger
>When you read English, there is no way to know how to pronunce it
Come on now, I've never found it particularly difficult and neither
have you, if you're honest, as your flawless spelling proves. Yes, a
vowel may be pronounced diffently when being part of different
characteristic letter groups, but the same goes for French and they
have four different accents as well, instead of to the German one.
Then there is question of saying numbers. French shopkeepers tend to be
completely stumped by the simple and polite request "en chiffre, s'il
vous plait" and can't name the price of anything intelligbly.
English is known to be the easiest of all foreign languages to learn
(except, perhaps, esperanto) and a very good choice for a common
language, even if not chosen rationally but through American military
dominance.
Peter Dassow
>
> If that's your criterion, then Latin was dead and artificial at the
> time of Gaius J. Caesar, who wrote an accomplished Latin grammar while
> on his campaigns. It is this artificiality eschewing the depradations
> of native speakers ("Handy", etwas erinnern, wegen wem, he invited my
> wife and I) that makes Latin so valuable and logical.
Remember this is already offtopic for this newsgroup.
Regards
Peter
BruceMcF
> I understand the restriction, but it isn't always so convincing. Well, maybe in the case of the 8080, but for larger processors that were traditionally microprogrammed you have to explain why the microprogrammed machine is not an emulator. Does it matter who built the machine? Is the IBM P/370 a real IBM 370 or is it IBM's emulation of a 370?
This is like the distinction between bridge and layer in networking.
If you have one network layer running on one data transport, and a
different network layer running on a second data transport, and there
is a connection between the two that translates from one to the other,
you have a bridge. If on the other hand, you implement the first
network layer on the second data layer, and use the original network
layer running on both data transports to connect them, the network is
running native on both data transports.
If the CP/M BIOS is operating by calling the services of another
operating system, its an emulation of the CP/M operating system in
another operating system. If the CP/M BIOS is operating by calling a
motherboard BIOS directly, then BDOS running on top of it is running
native, since CP/M is the primary operating system.
> Is an FPGA based hardware implementation emulation or not?
The processor core is a hardware emulation of the original processor.
If it were a copy of the original processor design, you'd have to pay
the IP owner of the processor design.
BruceMcF
> Come on now, I've never found it particularly difficult and neither
> have you, if you're honest, as your flawless spelling proves.
Spelling proves nothing regarding pronunciation. I had a large number
of Chinese and Thai students in Australia who were excellent spellers
but had difficulty saying numbers of distinct words so that they could
be distinguished one from the other.
And of course English can be pronounced phonetically, one simply needs
to know which of several incompatible pronunciation systems a
particular word falls into. And it cannot always be inferred from the
spelling which system the word falls in:
The lieutenant at the bow performed a flawless bow when he saw the
captain's wife walking toward him.
> English is known to be the easiest of all foreign languages to learn
By who, precisely? I have frequently heard the opposite.
Steve Nickolas
> English is known to be the easiest of all foreign languages to learn
> (except, perhaps, esperanto) and a very good choice for a common
> language, even if not chosen rationally but through American military
> dominance.
That's funny, I've heard it's the absolute hardest, except possibly for
Japanese or Chinese.
-uso.
Mr Emmanuel Roche, France
Hahaha! It is funny how the subject of the English language can be
interpreted so differently!
> Come on now, I've never found it particularly difficult
Are you sure that you do not forget that English is, technically, a
branch of the German language?
If you had studied foreign languages at University, you would have
learned that English and French are the two most different languages
in Europe: that's why we have so much trouble speaking our opposite
language correctly (not to mention the fact that 1/3 of the English
words are of French origin, but are either pronunced differently, or
have another meaning, now).
By the way, have you tried to read aloud the poem whose link I gave?
If you managed to read it without one single error, stop what you are
doing, and become a Professor of the English language. There are maybe
less than 10 native English Phonetic Profs able to do this. Each time
I showed this poem to a professional translator (at a time, I knew
several), he was stopping somewhere, wondering how to pronunce this
word. And, at the time, the text was not showing, in Italics, the
difficult words. In fact, I was told that the BBC was using this poem
to test its "speakers".
Finally: so, the source of all those problems is that English is not a
phonetic language. But, interestingly, Sanskrit, one of the oldest
known written language happens to be precisely a phonetic language!
Incredible! We know, without the slightest doubt, how people were
pronuncing Sanskrit 5,000 years ago! (Scolars spend their time
discussing how the Romans were pronuncing famous sentences, 2,000
years ago...) According to Linguists, the reason is that they were
recording sacred texts, 5,000 years ago. So, they studied their own
language, and invented a written phonetic language that was unmatched
during 5,000 years, until the Occident reinvented Linguistics.
(Another fascinating fact of the Sanskrit language is that, at school,
we learn to write our words *ON* a line. Sanskrit write its words
*UNDER* a line... It is really a fascinating language!)
BruceMcF
> If you had studied foreign languages at University, you would have
> learned that English and French are the two most different languages
> in Europe:
Wouldn't that be something and Finnish or something and Baaque?
English is a language uilt on a French vocabulary and a German grammar
and then allowing them to fight a bloody indecisive war for dominance.
glen herrmannsfeldt
(snip, I wrote)
>> Is an FPGA based hardware implementation emulation or not?
> The processor core is a hardware emulation of the original processor.
> If it were a copy of the original processor design, you'd have to pay
> the IP owner of the processor design.
So if I run CP/M on an AMD 9080A that counts as emulation?
-- glen
Herbert Johnson
> I apologize to M. Roche and Herb Johnson for ridiculing their
> behaviour. I agree with Herb that the bulk of this thread was on-topic
> and disagree with Floppy that it was "to convert this thread in a
> shit" ... this struggle between Herb and Roche needs to stop. I was
> saddened when Jack Crenshaw left this list and I hope that
> moderatoritis will not cause M. Roche to withdraw.
>
> If you feel the further need to go on and on, feel free to discuss
> "the pot calling the kettle black" off-list...
Bill, I'm done, no need to "moderate" me further, if I'm the "you" so
addressed.
You might consider moderating the off-topic "language" discussion in
progress, however, and already retagged "Latin". That ain't about me,
Bill.
Herb Johnson
pot or kettle, you decide
retrotechnology.com
Holger Petersen
>completely stumped by the simple and polite request "en chiffre, s'il
>vous plait" and can't name the price of anything intelligbly.
En mill neuf cent soicente dix J'etais a Carcassonne
In 1970 I was in Carcassonne...
SCNR, Holger
BruceMcF
Are you wanting "emulation" to be a generic word that ignores what is
being emulated? Intrinsically, it can't be.
If you run CP/M on an AMD9080A that is running another operating
system, and the CP/M BIOS operates by making requests of that other
operating system, then that counts as emulation of the CP/M operating
system.
If you run CP/M on an AMD9080A with a BIOS that is accessing the
hardware without the mediation of another operating system, then that
counts as running the CP/M natively on that hardware.
How is it possible to complicate such a straightforward distinction?
Mr Emmanuel Roche, France
Well... The subject of languages seems popular...
> > If you had studied foreign languages at University, you would
> > have learned that English and French are the two most different
> > languages in Europe:
>
> Wouldn't that be something and Finnish, or something and Basque?
No: I confirm: the English and the French languages are the two most
different languages in Europe, yet they are only 75 kilometers apart
(the "Channel", which was an effective border).
I don't know, of course, what you are thinking by "something", but for
the Basque language, I happen to know quite a lot, since my Great-
Father, despite being born here, in Tonnay-Charente, chose to spend
his retirement (about 20 years) in Bayonne, "ïbaï-ona" in Basque:
"Clear Waters". As a result, I went to Bayonne during 20 years.
("Bayonet" is a word named after this city.)
As I wrote several times, we, Europeans, are the descendants of a
people coming from Africa, 100,000 years ago. However, when they
arrived in Europe, there were already some peoples living there. The
Australopitheque disappeared, and nobody knows why.
Apparently, the Basques are the remnants of a local population, that
was/is living in the mountains to the West of the above-mentionned
Bayonne. (Bayonne is in France, Basques are predominently in Spain.)
Genetic studies have shown that they did not move during the, at
least, last 20,000 years. This fact also shows in their language (one
characteristic of languages is that, the older it is, the more
complicated it is. The French Language, rooted in Latin, has 6 endings
for its verbs. (The Latin had 8.)). The Basque language is, by far,
the most difficult language of Europe, since you cannot re-use the
rules of another European language to understand it. For example,
several times I was told by Spaniards that they understand without
difficulties Italians speaking. A French simply do not understand one
Englishman speaking: the two languages are too different.
I could keep on telling you several other things on the Basques and
their languages, but I counsel you to have a look to:
http://en.wikipedia.org/wiki/Basque_people
http://en.wikipedia.org/wiki/Basque_language
I have read them: I saw no obvious error. Among all this stuff, note:
"The Basque language is thought to be a genetic language isolate. Thus
Basque contrasts with other European languages, almost all of which
belong to the large Indo-European language family. Another peculiarity
of Basque is that it has been spoken continuously in situ, in and
around its present territorial location, for longer than other modern
European languages, which have all been introduced in historical or
prehistorical times through population migrations or other processes
of cultural transmission."
and:
"It is spoken by 25% of Basques in all territories (...) Of these,
614,000 live in the Spanish part of the Basque country and the
remaining 51,800 live in the French part."
To finish, after World War One, in the Army Regiment were my Great-
Father served, there was only one man, from the mountains, near the
Spanish border, who was not bilingual. All the others were speaking
French fluently, even long before the Television.
After the 1960s, the "Lefties" decided to create a simpler version
(before, each valley had its own version, and they were so different
that, usually, they could not understand themselves -- just to give
you an idea how difficult this language is). It is this one that is
taught today, in some schools. ("A standardized form of the Basque
language, called Batua, was developed by the Basque Language Academy
in the late 1960s. Batua is mainly used in the Spanish Basque
Country.")
Since the Basques have no written History, little is known about them.
If you are American, maybe you could be interested in knowing that it
is believed that Basque fishermen crossed the Atlantic Ocean several
centuries before Christoph Colomb, but they never established any
colonies on the land: they were just interested in the fish.
Everything known about them is from European Historians.
> English is a language built on a French vocabulary and a German
> grammar and then allowing them to fight a bloody indecisive war
> for dominance.
Hahaha! Actually, it is a little bit more complicated than this...
But this message is already long enough.
glen herrmannsfeldt
(snip)
>> So if I run CP/M on an AMD 9080A that counts as emulation?
> Are you wanting "emulation" to be a generic word that ignores what is
> being emulated? Intrinsically, it can't be.
> If you run CP/M on an AMD9080A that is running another operating
> system, and the CP/M BIOS operates by making requests of that other
> operating system, then that counts as emulation of the CP/M operating
> system.
> If you run CP/M on an AMD9080A with a BIOS that is accessing the
> hardware without the mediation of another operating system, then that
> counts as running the CP/M natively on that hardware.
The 9080A is not an Intel 8080A. Some might consider that an
emulation of the Intel chip. The previous post seemed to make
a distinction based on proper licensing of Intel technology.
> How is it possible to complicate such a straightforward distinction?
There is always a gray area.
-- glen
J. Clarke
>
> Well... The subject of languages seems popular...
>
>>> If you had studied foreign languages at University, you would
>>> have learned that English and French are the two most different
>>> languages in Europe:
>>
>> Wouldn't that be something and Finnish, or something and Basque?
>
> No: I confirm: the English and the French languages are the two most
> different languages in Europe, yet they are only 75 kilometers apart
> (the "Channel", which was an effective border).
>
> I don't know, of course, what you are thinking by "something", but for
> the Basque language, I happen to know quite a lot, since my Great-
> Father, despite being born here, in Tonnay-Charente, chose to spend
> "Clear Waters". As a result, I went to Bayonne during 20 years.
> ("Bayonet" is a word named after this city.)
So you assert that Basque and Finnish are more alike than French and English
despite Basque not even belonging to the same family of languages as
Finnish.
Would you care to share with us the definition of "different" that you are
using?
<long discourse on Basque that does not appear to suppor the argument you
made snipped>
BruceMcF
> The 9080A is not an Intel 8080A. Some might consider that an
> emulation of the Intel chip. The previous post seemed to make
> a distinction based on proper licensing of Intel technology.
But the question was not whether it was an emulation of the 8080
processor, the question was whether it was an emulation of CP/M, an
operating system.
Its an emulation of an operating system if there is some program
running on some other operating system that makes a CP/M program think
its running on the CP/M OS. If there is no other OS there, then CP/M
is running natively.
BruceMcF
wrote:
> No: I confirm: the English and the French languages are the two most
> different languages in Europe, yet they are only 75 kilometers apart
> (the "Channel", which was an effective border).
But they are both Indo-European languages.
Finnish is not an Indo-European language at all, it is a Uralic family
language, while Basque is an isolate language. So French is more
different from Finnish and Basque than it is from English, and English
is more different from Finnish and Basque than it is from French, so
French and English cannot be the two most different European
languages since they are each more different from one isolate language
and one language family than they are from each other.
I prefer the Basque comparison, since Basque country runs into France
and France and England are the first two countries directly north of
Spanish Basque country. Plus I cheer for Euskaltel-Euskadi to do well
in the mountain stages of the TdF.