About RomWBW

[ladislau szilagyi]

Hi Wayne,

Yesterday, during a conversation on the "vintage computer federation forums", I received a reply that got me thinking.

I stated:

"There is a system software for Z80 computers named RomWBW ( https://github.com/wwarthen/RomWBW ).
If installed, regardless of the hardware configuration, for all kinds of 512KB (or more) RAM hardware configurations, it boots a CP/M where you can access the "extra" RAM using the same function calls. This way, my software that uses "extra" RAM (a text editor, the enhanced HiTech C compiler, the Cowgol compiler, etc) can be executed on these "RomWBW" machines without any problem, regardless the particular hardware on which CP/M runs."

I received the following response:

"Sure. For a certain limited set of hardware configuration. I suggest you count up the total number of CP/M vendors, the number of vendors that released systems capable of bank switching, and the number of vendors that use RomWBW. I think you'll find that the last is a small fraction of all the different CP/M systems available out there."

Leaving aside the word 'vendors', which in my opinion is used inappropriately here (we are talking about "retro-computing"...), I was struck by the statement related to the "coverage" that RomWBW offers for Z80 systems provided with 512KB (or more) RAM.

It's a shame that there are still people spreading such fake news...

I am convinced that the truth is different, but I couldn't answer anything because I don't have any statistics at hand.

Wayne, can you help me present some data that proves the widespread use of RomWBW?

thanks,
Ladislau

[Doug Jackson]

Just have a look at the configuration files for RomWBW.  It details a wide variety of systems that support RomWBW.

I have even used it for my own custom STD bus based cards. 

--
You received this message because you are subscribed to the Google Groups "retro-comp" group.
To unsubscribe from this group and stop receiving emails from it, send an email to retro-comp+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/retro-comp/f99e38dc-a0d0-4217-a720-42574ce8eb69n%40googlegroups.com.

[Alan Cox]

On Tue, 10 Jun 2025 at 14:08, 'Doug Jackson' via retro-comp <retro...@googlegroups.com> wrote:

Just have a look at the configuration files for RomWBW.  It details a wide variety of systems that support RomWBW.

Most of which exist in the tens, hundreds or sometimes even ones.

There were 8 million Amstrad PCW machines shipped (which btw you could make run RomWBW if you wanted to win the argument by cheating a bit).  Even in the pure retro space there are probably several ZX Spectrum Next's for every RC2014 system capable of running ROMWBW.

It's a tiny little corner of an obscure corner of computing, and really who cares anyway about volumes. However the other party is almost certainly right by any reasonable way of looking at it.

This btw is why I never use weird BIOS and platform specific hackery for code but use standard interfaces like ramdisc ones - they work everywhere.

Alan

 

[Peter Onion]

On Tue, 2025-06-10 at 14:52 +0100, Alan Cox wrote:
>
>
> On Tue, 10 Jun 2025 at 14:08, 'Doug Jackson' via retro-comp
> <retro...@googlegroups.com> wrote:
> > Just have a look at the configuration files for RomWBW.  It details a wide variety of
> > systems that support RomWBW.
> >
>
>
> Most of which exist in the tens, hundreds or sometimes even ones.

Can Stephen (Small Computer Central) and Spencer (Z80KITS) give some idea of now many
systems they have sold ?

Does WayneW have any count of RomWBW disk images downloads from github ?

> There were 8 million Amstrad PCW machines shipped 

Which is rather moot since there aren't 8 million of them still running to day.


PeterO

[Alan Cox]

> There were 8 million Amstrad PCW machines shipped 

Which is rather moot since there aren't 8 million of them still running to day.

I'd bet there are rather more of them in runnable condition than there are RC2014 machines. It doesn't really matter anyway, and the system A v system B nonsense was tiresome enough when it was C64 v Spectrum.

If you are having fun who cares.

[Wayne Warthen]

It is a sincere compliment that folks are so supportive of RomWBW.  Thank you!

I really don't have any idea how many RomWBW users are out there.  Even if I did, I couldn't guess how many are "active".  GitHub provides no download stats (I wish they did).  Just based on what I know of the number of RomWBW systems distributed by Spencer Owens and Stephen Cousins, I do believe that it is in the thousands.  

I suggest that no one engages in debating the merits of RomWBW vs anything else.  It does not aim to compete with actual 40 year old systems.  It is apples and oranges.  The hardware and software that are part of the RomWBW ecosystem are amazing.  The people that have contributed to RomWBW should be proud of themselves and have my gratitude.

Alan hit the nail on the head.  We do this because it is fun and we want to.  Those who see a benefit to using RomWBW will do so.  The rest will not.  I couldn't be happier with that outcome!

Thanks! Wayne

[Steve Cousins]

My sales of RomWBW capable kits through TIndie are:

608 x SC126 (Z180 CPU)

481 x SC131 (Z180 CPU)

200 x other (Z180 CPU)

100 x other (Z80 CPU)

I also sell in other places but in lower volume so add about 10 to 15% to the above.

In addition I sell PCBs (not in kits). Total guess say another 10 to 15% to the above.

My designs are open source with Gerbers etc. freely available. I don't know how many systems have been created from these.

As you can see the vast majority of my RomWBW capable kits are Z180 CPU based. I think Spencer picks up nearly all the sales of Z80 based systems.

In total I'd say my designs account for less than 2000 RomWBW systems in the wild. How many are actually in use is a different question.

Steve

[Steve Cousins]

Oops forgot to include SC130 (my lowest cost RomWBW system)
246 x SC131 (Z180 CPU)

[Bill Shen]

I remember during the height of COVID, Spencer mentioned that he has sold 5000 RC2014.  I’m speculating that half of RC2014 is RomWBW capable.  So truly many thousands of units are running RomWBW.

Bill

[Doug Jackson]

Wayne said.

"I suggest that no one engages in debating the merits of RomWBW vs anything else.  It does not aim to compete with actual 40 year old systems.  It is apples and oranges.  The hardware and software that are part of the RomWBW ecosystem are amazing.  The people that have contributed to RomWBW should be proud of themselves and have my gratitude."

Wayne, I have had a couple of small Z80 machines allegedly with Hard disk support if you had the correct disk adapter cards and disks.  All of which are unobtanium.

With RomWBW, it's possible to spin up a system and have support for a heap of large disks, something that was impossible in the mid 80s unless you were cashed up.

You can support SD cards, CF cards, floppies with a common format, various RTCs and other hardware because it's an ecosystem of people who are far greater experts than I am.  

I was able to modify my system to support XON/XOFFB and the use of CONIO to direct output.   Byt disks were impossible.

Thank you for the environment.  I use it in many systems, and it made CP/M come back to life.

Systems I own.  N8VEM Zeta, N8 Home computer, a couple of RCBUS systems with random z80 or z180 CPUs, and my own STD bus based systems.  I also have a Pulsar Little Big Board, but thats a snapshot in time.  I will eventually give it a hard disk when I get my head around the BIOS changes. AIt only has 64k of RAM.

Doug.

--

You received this message because you are subscribed to the Google Groups "retro-comp" group.
To unsubscribe from this group and stop receiving emails from it, send an email to retro-comp+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/retro-comp/824e129b-291e-46fd-b510-f409fe1b8f58n%40googlegroups.com.

[ladislau szilagyi]

Hi Doug,

This is your good part of the story. Mine too.

Let me tell you what the bad part might look like.

I mentioned above about a discussion I had on the "vintage computer federation forums".
I said there about a possibility to improve the performance of some CP/M programs given that we have a Z80 computer equipped with 128/512KB RAM, possibly running RomWBW.
I gave as an example the use of my enhanced HiTech C compiler to process very large C files.

I don't want to tell you what comments I received in response... but, I would like to mention a few statements.

First, I was told that what I was running was not 'real' CP/M.
That, running CP/M on RomWBW, is not actually CP/M, but a weirdo.
Then came the statement that, in fact, RomWBW users are only an insignificant part.
Then, I was faced with the following tough question: How can I use expanded memory on an S-100 equipped with a CROMEMCO 512 board (we are dealing with 4 x 64KB RAM, so RomWBW is not an answer here...).

I gave as an example the solution I found for SC108, which has 2 x 64KB RAM.
It's about using routines stored in both 64K banks, up in the BIOS (the so called "shadow routines", used to move bytes between the two 64KB banks).

When I mentioned that I had to modify the BIOS, making it a little "slimmer", to fit ~150 bytes containing the "shadow routines", I was met with distrust and rejection.

I included in the answers code fragments from my modified BIOS (not a big deal, simply I replaced some fragments written in "8080" with shorter Z80 code), I even added a file containing an extended memory allocation driver, customized for the S-100 with CROMEMCO 512 board, which I made in a hurry, adapting some already existing code.

Well, they were downright horrified.
What kind of CP/M is this? Did you dare to modify the BIOS? This is not CP/M anymore, real CP/M only runs on 64KB.

I ended up being accused of being impertinent, of insulting the "retro-computing" community, and so on...

So, to ease my frustration, I went back to my humble RC2104s and installed the "Colossal Game" on one of them.

I copied via XMODEM the "adv.c" single-source C file (3800 lines), I compiled-it,
(by the way, it took me a quarter of hour to compile-it, the assembler file built by OPTIM has ~13500 lines !)
...but in the end, I had my happy hour of fun, playing Adventure, on my humble RC2014... which does not run the "real CP/M", according to some!

regards,

Ladislau

[Doug Jackson]

Oh.   That's horrific. Some people need a significant attitude adjustment.  

Of course it's CP/M .  

One of the negatives of the community is that there are some that are set in their ways.  

But one advantage of there being a community is the large number of people, some of which have briliant ideas.

Its as important today as it was 20 years ago to have a kill file, so you can idnotmre the haters. 

To view this discussion visit https://groups.google.com/d/msgid/retro-comp/8e01bdc7-79f8-42e9-b4ca-0cd1b6ae8f55n%40googlegroups.com.

[Peter Onion]

To some people the preservation of a system is the most important part of what they do.  They will see any modifications or use of non-original hardware or software as destroying the authenticity (and hence the value) of an object.

To other people the enjoyment comes from being able to do useful or interesting or amazing  stuff with a computer that has only a small fraction of the power of the phone in your pocket. 

So it's not surprising that when these two world views collide over something like using RomWBW on older hardware there might be some who don't understand the motivations of the other group.

"If you are having fun who cares" Alan Cox

PeterO

[Tom Storey]

Too many people get caught up in trying to define what something is. And some people can't accept any definition other than their own.

Also, oppinions are like butt holes, everyones got one. The internet is not short of pedantry either. And we're all probably guilty of it at some point and to some degree, and often it "hurts the ego" to be on the receiving end of it. Sometimes it can be a good thing to get that "reality check" because it forces you to re-think your own position and that of others.

I'd also suggest that forums tend to encourage herd mentality sometimes, possibly because they are largely anonymous places (I tend to use my real name because I like to think it helps keep me honest 😄). It only takes the "wrong" person to respond to a conversation with a negative thought or idea, and that can encourage others to do the same. Forums just also tend to have their resident grouches.

You've done a lot of good work, as have a lot of other people, making retro computing more accessible and fun in these modern times and all should be proud of this. Stick to what you do best, and to hell with the rest.

To view this discussion visit https://groups.google.com/d/msgid/retro-comp/8e01bdc7-79f8-42e9-b4ca-0cd1b6ae8f55n%40googlegroups.com.

[David Reese]

I'm Sorry, Folks, But It Needs Be Said,

Here's the wind-up, and here's the pitch:

I've read most of this thread, and I'm forced to take some of my scant downtime to wonder why are we bothering to prove anything?  I think all these CP/M curmudgeons are just jealous because the folks here, the ones who produce the hardware and firmware we love, have beaten them at what they thought was their own game, and they know it.

I cut my teeth on CP/M 2.2 in the late seventies, and I can tell anyone who will listen that I vastly prefer ROMWBW to anything produced by OSM/Zeus, Cromemco, Actrix, Kaypro, Compaq, Televideo, or any other vendor I've used.  My SC126 in its little tigerwood case runs for over 100 hours on a 10,000 mAh cell phone battery, is miles per second faster than anything I ever ran in the seventies or early eighties, has (what?) eight times the RAM and insanely more abundant, faster, more reliable, and cheaper mass storage than any of those old buggy buggies from the bad old days, for a mere fraction of the cost (not to mention the mass).  Some folks would grumble if you hung 'em with a new rope.

These folks that say, "that's not CP/M" couldn't produce anything better on a good day.  Why should any of us care one whit what they think?  Fooey!

Oh, and "Mr. Statistics", who is living so far in the past he's eating dinosaur eggs for breakfast?  Go and ask him how many of his precious thousands of old CP/M machines are still soldiering on, and just how easy it is to get parts to support them.  (I do computer repair, by the way.  A+ since 2003 and all that.)  "A small fraction of all the different CP/M systems available out there..." ???  What?  Did he fall asleep in 1973 and wake up yesterday?  Somebody get that man a cup of coffee -- hot, black and sweet, 'cause he needs it.

I've got perhaps $150.00 US in my SC126, and that includes the exotic wood case, the battery, cabling, SD cards -- the whole schmeer, including the terminal I'm using with it.
You cannot touch a complete working CP/M system on eBay for less than three times that much.  A portable?  That'll set ya back even more.  Go price an Actrix:  I dare ya.

Oh, and in the same search, there's an offer for this:

SC126 Z180 SBC for RCBus (RC2014/Z80) CP/M Assembled+case

Opens in a new window or tab

$135.00

Buy It Now

+$8.50 delivery

Okay... rant over, I think, except for one more grain of salt to rub in this guy's wounds:  "There are three kinds of lies:  lies, damned lies, and statistics." -- variously attributed to Benjamin Disraeli and Samuel Clemens (Mark Twain).

Now, go forth, conquer, and have fun all you maniacs -- that's an order.  After all, that's what this thing of ours is all about.

Later On,
David

--
You received this message because you are subscribed to the Google Groups "retro-comp" group.
To unsubscribe from this group and stop receiving emails from it, send an email to retro-comp+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/retro-comp/f99e38dc-a0d0-4217-a720-42574ce8eb69n%40googlegroups.com.

[Peter Onion]

If people want to make up their own minds, the thread is at https://forum.vcfed.org/index.php?threads/writing-assemblers-what-should-a-good-assembler-do.1245894/page-14 onwards.

PeterO

[ladislau szilagyi]

Yes, there you can find the whole discussion.
Towards the end, I lost my patience and probably didn't respond with the proper neutrality...that's it.
But, I tried to represent the interests of real users and not those who sit on the sidelines, criticize and comment on terminology.
Ladislau

[ladislau szilagyi]

Note: not all the conversations from  "vintage computer federation forums" are conducted in this chaotic and biased way...

Here's another conversation that is conducted in the most decent and correct way possible: https://forum.vcfed.org/index.php?threads/which-c-compiler-do-you-prefer-for-cp-m.1253284/

Ultimately, I think what I experienced is an isolated case, after all.

Ladislau

[Sergey Kiselev]

The thread doesn't look too bad, chaotic or biased. It's a good idea to limit the drama and focus on the constructive feedback. Assume the best intent. Avoid trying to prove that someone is wrong. Also try to be open to other people's opinions... Remember that we do retro/vintage computing mostly for fun.

• The people in the thread are not wrong about CP/M 2.2 not providing memory bank switching API. Any software that supports more than 64 KiB on 8080/Z80 will be system or perhaps firmware/BIOS specific.
  • Alan has a good point about using standard interfaces, e.g., block devices such as a RAM disk. Of course that would result in significantly slower performance than accessing the memory directly using bank switching. So there is a tradeoff of speed vs. compatibility.
• The number currently useable systems with RomWBW firmware vs. number of other currently useable CP/M systems is up to debate.
  • But the total number of CP/M capable systems made far exceeds the number of the systems with RomWBW firmware...
• RomWBW is likely the best thing that happened to the retro CP/M community and I truly appreciate Wayne's effort and dedication. Without RomWBW, we all would have to hack our own BIOSes, reinvent multiple wheels (not that it is not fun).

All the best,

Sergey

[Alan Cox]

• • Alan has a good point about using standard interfaces, e.g., block devices such as a RAM disk. Of course that would result in significantly slower performance than accessing the memory directly using bank switching. So there is a tradeoff of speed vs. compatibility.

For a lot of uses the difference is tiny as you are doing block copies. For many machines you also didn't have much choice as the only way to add ramdiscs was generally as an I/O interface unless you wanted to do major scary stuff on your system like stacking RAM chips and attaching the CS pins to floating 74138s. At that point you could only have a block I/O access model really.

[ladislau szilagyi]

Things are starting to move (I think in the right direction).
People are starting to ask questions about RomWBW...
See https://forum.vcfed.org/index.php?threads/romwbw-on-s-100.1253333/ 

[john]

What I would really like to see is a drop in RomWBW card for existing S-100 systems.  No CPU thank you.

For example, replacing a disk controller card in an IMSAI with a RomWBW card.

I don't mind patching CP/M BIOS, done it many times before.

Sadly, it seems the vintage systems are disappearing.  I get the impression most of the thrust these days is building new CP/M systems from the ground up.

-J

[john]

Yes, I saw the response on VCF too.

Let me say there are 2 conflicting views here:

1) Purists with the goal of preserving history

2) Hobbyists that love to tinker and improve their systems

I understand both sides.  To that end, I maintain some systems in 'vintage state' with era appropriate hardware/software.

Others, are essentially 'experiment boxes'.

And I should note that before preserving computing history was a thing, hobbyists were the founders of personal computing as we know it.

I'm running RomWBW in a Heathkit H8 and love it.

-J

On Tuesday, June 10, 2025 at 7:04:12 AM UTC-5 ladislau...@gmail.com wrote:

[ladislau szilagyi]

Yes John,

you're right, there are two camps: the purists and the hobbyists.

I was a purist myself in the 80's. I still have Sir Clive Sinclair's ZX81. Unfortunately, it does not work anymore, an I have not the expertise to fix-it.
But, in 2021, I discovered RC2014, and became a hobbyist, a tinkerer. I learnt the hard way how to build an RC2014, starting from boards, components and chips. (I don't know how I would have managed this quest without Steve Cousins' help...).

Now, if I think about it, I also respect the effort of some people who try to keep computers manufactured 50 years ago in working condition.
What I can't really understand is the mentality of some purists, regarding, for example, the topic of what is and what is not "real CP/M".

This fixation on not accepting technological progress and rejecting some hardware configurations, which in the opinion of some purists, contradict the notion of "true CP/M", leaves me downright perplexed.

I had the impression that, in this movement called "retro computing", you have to be very open in accepting new, even innovative, alternatives, in order to preserve, maintain, modify these little machines built around old CPU's. And, finally, to have fun while tinkering with hardware or software.

Ladislau

[Peter Onion]

On Thu, 2025-06-12 at 23:27 -0700, ladislau szilagyi wrote:
> Yes John,
> you're right, there are two camps: the purists and the hobbyists.
>
> I was a purist myself in the 80's. I still have Sir Clive Sinclair's ZX81.
> Unfortunately, it does not work anymore, an I have not the expertise to fix-it.
> But, in 2021, I discovered RC2014, and became a hobbyist, a tinkerer. I learnt the hard
> way how to build an RC2014, starting from boards, components and chips. (I don't know
> how I would have managed this quest without Steve Cousins' help...).
>
> Now, if I think about it, I also respect the effort of some people who try to keep
> computers manufactured 50 years ago in working condition.
> What I can't really understand is the mentality of some purists, regarding, for example,
> the topic of what is and what is not "real CP/M".

It's all about preservation. As I said earlier they consider using any non-original part
(be that hardware or software) reduces the value (or maybe worthyness) of a machine.

> This fixation on not accepting technological progress and rejecting some hardware
> configurations, which in the opinion of some purists, contradict the notion of "true
> CP/M", leaves me downright perplexed.
>
> I had the impression that, in this movement called "retro computing", you have to be
> very open in accepting new, even innovative, alternatives, in order to preserve,
> maintain, modify these little machines built around old CPU's. And, finally, to have fun
> while tinkering with hardware or software.

While that is true, in the preservation arena I think it only applies once you do not have
the ability to maintain a machine using original parts. When the Z80 becomes unobtainable
then other solutions will be found (FPGAs?) but while they are still available using such
a part on an original S100 CPU board would be considered wrong. And since software
doesn't become unobtainable in the same way as hardware does, they will see using RomWBW
in the same was as they see using a FPGA-Z80.

PeterO

[Tadeusz Pycio]

I completely understand the different approaches to retro-computers, some people keep their 40-year-old computers in working order, others focus their activity on recreating old machines using modern components, but there are also those who want to feel a breath of fresh air in forgotten integrated circuits. I think everyone can find their place among the options available. Personally, I have walked each of these roads, looking for the place that suits me best and I am where I am ;).
There was a time when I got greedy with the novelties, MCS-51, AVR, ARM-Cortex, now a bit RISC-V, but still 8-bit is my hobby. I'm well aware of how high the entry threshold is for new designs lacking a unified BIOS, I've been developing my simple IPS boot loader for a few years now and I don't think it will ever be suitable to make it public. That's why I think RomWBW and the concept of simple modules is the best thing that could have happened to people like me. I appreciate the effort Wayne put into developing Z80-compatible systems, I think without his work a lot of new computers would not have been built. It seems that some people have forgotten that the most important thing in all this is to have fun and not to cross out solutions that will not work on their computers, a common occurrence with operating systems that did not define the hardware model.

[ladislau szilagyi]

It's not just about hardware. For software, I think there's the same problem: purists versus hobbyists.

I can't understand the attitude of some who reject any code change in BIOS. What can we say about the use of complex systems, like RomWBW, created to make life easier for CP/M users? Reject all these software tools because they do not belong to the "real CP/M"? (I would also like to know what the definition of "true CP/M" might be. CP/M for 8080? Perhaps, also Z80? But how about Z180? Why not 6502, like David Given shows it's quite possible...Come on, we will fail into ridicule...)

Why it is harmful to use all the hardware resources? For example, why you shouldn't use extra RAM, if it is accessible, to be able to edit larger files or to be able to compile or assemble larger files? Of course, this use of additional resources must be well regulated and configured. If the software is labeled to be used only for 512KB RAM configurations, I do not see any problem (who will try to use-it on a 64KB RAM computer, knowing that that software is made only for 512KB RAM ?).

From my point of view, it's about the same as rejecting the use of hard disks, CF, SD, etc., for the sole reason that at the beginning, CP/M only used floppy disks (by the way, I remember that I started by using 8-inch drives, then I migrated to 5-inch drives, etc.).

But, that's it, I'm starting to get used to this reality (purists vs. hobbyists)

Ladislau

[7alken]

hi all, ya, purists, umm, so, as I had last week refocused on SW again, because of also learning about m68k deeper, this thread probably speeds up publishing of my notes/updates here;
I don't understand wars; btw, condolences to all affected (not only) in UK by India plane crash :-(
Petr

[David Reese]

It's okay to lose one's patience from time to time.  Case in point:  my-own-self when confronted with... well, let's not get into that again.

A point I haven't seen mentioned yet:  in its early days, CP/M was rather a spartan OS that did not support any actual "native" hardware (except, maybe the Z80 and 8080 CPUs).  Digital Research did that for a reason.  Vendors had to come up with all kinds of hacks to make their systems run, which is why each system's firmware was different.  That's what that fellow said that really smacked me upside the head  - it's like he was living in a bubble or something.

I've mentioned the Actrix DS before.  It was rather a latecomer to the CP/M scene.  It's a "portable", at least by the Army definition ("If a Regiment can move it then it's portable.")  

33 lbs (just shy of 15 kg), with a 4 MHz Z80A, 64 kb of RAM, two DS/DD 5.25" floppy drives (with an option for a whopping 10 MB Xebec hard drive on a special card, and another option for an 8088 card to allow it to run MS-DOS), an alcove for storing a small number of floppy disks, a 5" amber flat-CRT display that was actually readable, a 300 baud acoustic coupler modem and a built-in Epson MX-80 printer (I added a tractor feed to mine).  Oh, and a partridge in a pear tree somewhere in there, I'm sure.  It shipped with an office suite from Perfect Software that included bound manuals, and a few other things, notably MBASIC 5.22 (which I've not seen anywhere else.)

The firmware for this beast was intense.  I think there was something like 350 of these produced under the Access Matrix and then the Actrix brand names.  Nice machine, but came to market way too late (1983 is when I bought mine...)

Kaypro CP/M would NOT run on it (I tried), nor would Kaypro Wordstar.  (I did copy the files over to Actrix-formatted disks, hacked around on them for a while, then gave up and "run what ya brung" as the drag racers used to say.)  If I'd had access to Lifeboat CP/M back then, I might've tried it.

 

It was left up to system vendors to adapt "generic" CP/M to their individual systems.  Somewhere there's a publication that gets into how to do that...  ah, here it is:  The CP/M Operating System Manual

copyright 1976 by Digital Research, Inc.  It's starts in Chapter 5 (CP/M 2 System Interface), and continues in Chapter 6 (CP/M Alteration), so, here we see that DRI intended for vendors to alter CP/M to fit their hardware offerings.  That's one of the reasons CP/M was such a successful OS for its day.  I like to think ROMWBW continues that grand tradition.

[Dave White]

Yes, you alter CP/M to fit your system, but only the BIOS. The BDOS and CCP remain the same.  So not being able to run Kaypro CP/M on an Actrix is no surprise at all. The CP/M part would have been compatible, but the BIOS was tied to the Kaypro. 

That was the whole point, of CP/M. Manufacturers of hardware only needed to modify a relatively small BIOS to handle a few hardware interface calls to make them compatible with their hardware, bolt on the standard BDOS and CCP, and CP/M would "just work". That is why CP/M was so successful - only a small BIOS rewrite was needed to allow it to run on unlimited hardware.

[ladislau szilagyi]

There is an interesting conversation here: https://forum.vcfed.org/index.php?threads/what-makes-a-cp-m-program-a-cp-m-program-and-not-just-say-a-z80-program.1253341/

Ladislau