Hi everyone.

Did not find a better place to post this - sorry in advance if I step on protocols.

I joined this group because of the fantastic work that has been done on the CMM2.  I wanted to find an education computer system so that I can help my grand-kids discover the world or programming the way I did in the 70's/80's and if you try to find an "educational computer" now, you can only find school laptops.

I spent days working with AI and searching and I could not find a better alternative.  Something that makes it easy for a novice to start.  Something that a child doesn't need an account for.  Something that doesn't require learning OS commands and concepts.  Something that allows interactions with the real world like LEDs, motors, sensors.

While I understand and admire the passion of people that preserve the old hardware (I cut my teeth on a System 80); I like the modern peripherals that are dirt cheap, and modern MCUs that are so powerful, rather than trying to keep floppy drives working and playing with cassette tapes.  

The CMM2 ticks so many boxes and the only sad thing is that it appears to have been sabotaged by the gods of time that took the wind out of its sails through the chip shortage era.  It is now very hard to find the hardware.  I took it as a sign when I found a Gen1 on eBay and I immediately snapped it up.  I have four grand-kids so that means I need MORE.  

In the end, I settled on doing a JLPCB production run for five Gen2 boards that will be my spare-time project for a little while.  When you do this you see that even a few years of elapsed time the rot sets in and key parts are no longer available.  It won't be much longer when the sdram is off the shelf as they are running out remaining supplies now in some places I looked. I thank the community for providing the build kit that allowed me to do this.

I really see that preserving the function and experience that the CMM2-style system provides is a worthwhile thing - even if it never makes anyone rich.  Having read a bit about the history, I see that a lot of passion and dedication went into its creation.  For that I am grateful.  And, hopefully, my grand-kids will be grateful.

So now, I will stumble around getting to know the intricacies of the ecosystem while I build up some hardware.  Maybe I will need some help on yet-to-be discovered technicalities but time will tell.  In parallel, I am working on getting a reference design into KiCAD 9 and reacquainting myself with the STM32.  I think preservation is important as we are currently losing the battle on building capable minds and becoming AI junkies.  Maybe I can assist with preservation at some stage.

Just a bit about me -  I taught myself to program on a System 80 (TRS80 clone) using Basic and Z80 assembly.  I then stepped in C and eventually C++.  I became a "Digital Systems" engineer through RMIT and spent years in software and hardware in such areas as data-acquisition system, system emulators and assemblers.  I spent some time in projects such as corporate PKI and high-security networks. At one stage I designed and built a modern replacement for a Perkins Elmer mini-computer system that replaced a national maintenance management system for Defence.

I retired early and live on acreage with my family nearby.  I have built off-grid solar systems and play with ESP32 embedded systems including designing and building my own boards for such things like my current project - a reverse osmosis bore water harvesting system.  I work with ESPHome, Node-Red and Home Assistant a lot to automate many of the mundane tasks - particularly around prioritising and maximising solar energy as we cut the cord to the utility company.

I read Geoff's history of MMBasic and how he built it from scratch using the exact approach I have done a few times and knew that the system would be maintainable - and I think history has proven this.  Hat-tip to Geoff and the others who's names I don't yet know.

BTW, I love that forums haven't gone away totally into Facebook and the like.  There's a lot to like about the way we did things in the good ole days.
Edited 2026-02-09 07:38 by rrroonn

Welcome to the forum. I share many of your experiences in the PC world--CP/M system with Z80 and onward and upward from there.

The other name you should know in addition to Geoff's is Peter Mather. He ported MMBasic to the CMM2 and guided the hardware development, as he has also done with the PicoMite.

MMBasic on this hardware is a powerful system, and you're not likely ever to experience one where the developer is so forward-looking and at the same time responsive to user requests. (My all-time record for request made to inclusion in a release: under 2 hours--lucky timing.)

Thank you lizby.  

I liked the z80. Never understood the love for the 6502 but I guess it was a price-point thing.


Shout out to Peter Mather @matherp too. Great work. Thank you.

  rrroonn said  

I liked the z80. Never understood the love for the 6502 but I guess it was a price-point thing.

I loved the Z80.   But the 6502 had the Zero Page addressing in Pre-indexed and Post-Index modes.    That made big programs (like BASIC interpreters) able to very efficiently navigate tables and index into things.

A 1 MHz. Apple could keep up to a 4 MHz. Z80 doing MBASIC.

  EDNEDN said  

A 1 MHz. Apple could keep up to a 4 MHz. Z80 doing MBASIC.

Please don't bring up that old annoying Z80 vs. 6502 debate again, when everyone knows that the 6809 is the champion.
rrroonn, Welcome to the club.
Cheers
Martin

I liked the Z80 RST commands. The Nascom computers put them to great use in their monitors. Relative Calls for relocatable routines? Output a string to stdout using a single command? Nice. :)

Welcome, rrroonn.  :)

The CMM2 design is fairly old now. For something readily available it might be worth looking at the PicoMite family of MMBasic machines as well as the CMM2. They aren't in the same league performance wise, but as learning systems there isn't a lot to choose between them and the CMM2. MMBasic is very similar on both platforms. Systems based on the RP2350, as used in the Pico 2, can drive HDMI displays in several modes and we can now connect a passive USB hub to allow the use of USB keyboards, mice and some game controllers.

Thanks for the welcomes.

I did look at the picomites and they have some appeal.
When I decided to bite the bullet, only the cmm2 seemed available as a finished product whereas Pico's were a diy thing and every example I saw was "untidy".

There is a saying that good enough is good enough but that never appealed to me.  If it can run at 480Mhz, then it should. If it can have a bazzilion gpios then that's the the perfect amount.

I would eventually like to see a maximite CMM3 with USB kbd and mouse and HDMI. That would make it more contemporary and supportable as PS2 and VGA are approaching their use-by date. How hard can it be? . Having a big "beautiful" powerful core means you don't have to make so many compromises and gives you a lot of future options.  Thats just an opinion and everyone has one.

Anyway, I look forward to discovering the known unknowns and maybe some unknown unknowns.

Ron.
Edited 2026-02-09 19:29 by rrroonn

I think Peter Mather has been considering options for a new CMM version. It's always a case of figuring what components are liable to still be available for some time though.

There aren't a lot of options if you want to buy "off the shelf". Everything about MMBasic is done by enthusiasts and hobbyists, it's not really a commercial setup. There are a couple of people that can provide small numbers of ready built units but no-one produces them in commercial quantities.

rrroonn

I suggest you PM grogster about the CMM2. He is supplying them at close to cost using the new value engineered design files