Thank you - I appreciate hearing that :)

Thank you! Will do! I am working on one on Crystal-Lang right now - its a lot of fun.

I got them from Amazon - search for 'YUNGUI PCB Protoboard'

Agree! Being able to program directly on the device - no matter how painful - seemed pretty core to the concept of being a computer. My next step is to make it a little easier to program on the device.

I've just started playing with Raku a little and can see where it has a lot of uses - particularly in making your own domain-specific-language. I have never paid attention to Crystal until your comment, but it looks amazing and can't wait to play with it a bit. I like the macro system and the built-in ability to call native libraries. Exciting stuff - thanks for the pointer!

Interestingly, I just watched a video on a found homebrew based on the RCA 1802 where the maker had built their own backplane (not S100) - it might be interesting to you: kzbin.info/www/bejne/aqHKgGp5q9iXnKs - thanks for watching - cheers!

I am unclear on the context? Please expand :)

If I am 100% honest, I got emotional making it. Thanks!

You're right about the keyboard module - it definitely has plenty of clock-cycles left over that can (and will) be put to use. There are lots of optimizations and honestly this could been done on one or two RP2040s. This was mostly an artistic design and it does an over-abundance of processing, but that has the advantage that I can develop out services on the nodes over time to make it into a capable little computer. Cheers!

I really appreciate you saying that - thanks. Cheers!

Thank you! I'll make it a better computer over time, but I'm super happy to have gotten it working in the first place 😊

Please do! Its time well spent. Thanks!

Absolutely - Perl has set the gold standard for regular expressions.

This genuinely made me laugh - and gave me a target to shoot for. Check back :)

Yes! I recently subscribed to them and am working my way through the series. Thanks for the tip!

Thanks! The time spent diving deep has been a great diversion - and having the creative freedom of how to connect the pieces together makes it fun. Time well spent.

Would like to hear more about that - any overview information?

@wrongdogreckons-cp2sy that's way to much info for KZbin to handle. But... It is similar to an FPGA but completely different. The highest level overview... I got what I call a cluster, truly thow that's the entire architecture... Each cluster connects to eight other clusters in a three-dimensional stack. There is no multiplexing. All clusters are flawlessly identical so no way to address a particular X,Y,Z if its not in use. but the eight ports on each cluster are numbered so you can "snake around" through an XYZ matrix of clusters that are "blank/formated" but doesn't matter if blank or not... The full address of a cluster is broken down into three parts. The application ID, function ID, and sequence ID. , That should be self-explanatory enough, that part with the ID's. So when you are writing your code aka populating clusters, you are also giving them their address but you don't have to really think much about it if you use it in the logic that I mentioned. It is a process in memory style type of architecture. There is no ALU, GPU or CPU but it is All and None at the same time. There are three types of buses that run through the entire architecture the way the clusters connect to each other through ports but their identical in function but their used differently, I don't have anything in particular I call them just high medium low buses. If you are trying to get ultra fancy with your coding, on the low level bus, you would not really need more than two duplicates of the low level bus. But just one is perfect. For the middle level bus, six duplicates is probably pushing it to be overkill, buh that's trying to seek the most absolute godlike performance. For the high level bus, however many duplicates of it you have, essentially you could 'almost' consider that your core count, but not truly. The high level bus is used for application to application communication, system to system, application to system etc, Also including external hardware, mouse, keyboard, speakers, printer, monitor, network, USB, etc. The middle level bus is used for function to function communication inside of an application. And the low-level bus is just following sequence, the order of operations. Your assembly code one instruction after the next. The clusters physically defragment to each other. So it does not matter if it looks like a perfect cube or if it's branched out like lightning. As long as the appropriate clusters are neighboring each other at least by one port. Let me give you a visual example. You have a bag with a thousand marbles aka clusters. Each cluster holds an assembly instruction. (following the sequence), So on the low level bus, the particular function of a particular application will zone off all of its clusters, making a pool on the low-level bus. So however many assembly instructions are used to make up that function, they can follow the order of operations, the sequence, all day long and not interfere with any other functions of that application or any other application, because the ports are cut off on the low-level bus that do not neighbor the same function of the same application. And to have functions communicate with each other inside of the same application., they will use the middle layer bus. With the middle layer bus, the way those ports are cut off, is representative of the application as a whole. And when one application needs to communicate with another application or hardware, they will use the high level bus. Physically externally you are 'kind of' in a way limited to the duplicates of the high-level bus. But internally you are not physically limited to high level buses because of a feature I developed called the rope bridge, but using it wastes precious cluster space, It forms a physical bridge connecting one or more applications or even particular functions between applications without using the high level bus at all.

@wrongdogreckons-cp2sy Part 2, here is the continuation because KZbin on my phone could not handle editing the above text as I was typing. I need to bring up hard coding addresses on any other architecture creates security issues but not on mine. Here's why. Since there is no multiplexing, you have one wire per bit. One cluster that wants to activate another has to put its identity/ address on the bus, the cluster that it wants to call out to, its instruction, its data, etc. And the receiving cluster sees its identity being called. The receiving cluster checks its security ID, a section within itself, kind of like a registered but permanent storage. that aka register has enough bits to store an entire address application ID, function ID, sequence ID. There is a way to store an unlimited number of security IDs starting with one cluster, but it's too much to go into here. Not important at the moment. You can be mediocre about the security by just filling in the application ID and leaving the others blank or you can have stronger security by filling in the application ID and the function ID or extreme overkill security by setting all three in the security ID. If the security ID filled in matches the identity of the calling cluster then execution continues. otherwise the calling cluster gets terminated. There is no way to soft code or hard code around this. And obviously if the security ID is left blank then any function of any application can interact with that particular cluster. I haven't mentioned this part yet, but it should have been obvious. The callings are a broadcast message. So if you've got a thousand clusters pooled up on the low-level bus and the first sequence calls out to the second sequence all thousand or 999, get that message. And no address is ever used with all zeros for all three together or individually application function sequence. This saves on electricity usage.

@wrongdogreckons-cp2sy Part 3, In regards to external hardware using the high level buses, so physically you would be kind of limited. But the reason I said not exactly is because of, ... For one, how many of the clusters are you willing to expose outside of the periphery of the chip? And for a piece of hardware like speakers, for example, you could hard wire to the middle level bus or even the low level bus. So your sound processing app would have isolated access to the speakers and of the hundreds of applications that need to output sound. Well, they have to go through the sound processing app anyway.

@wrongdogreckons-cp2sy part 4.... ....... For one, if you are willing, of course, and if you think hard enough, you will know what the following is for... Nehine oh fur, for stix hero, ...... ..... ..... .....

Thank you very much! Those are some very kind words - I appreciate them and I appreciate you.

Agree - they open a whole new field of possibilities - far beyond another NES emu :)

Thanks for the sub! I've had a great time, so expect more :) Cheers!

@@wrongdogreckons-cp2sy make more that kind of thing and you will gain a bunch of followers 😜

Thanks so much! I've had a great time making it and so I appreciate the feedback!

Its all an incremental journey. Thanks for watching - cheers!

Well said - I appreciate that statement. Thank you.

Thank you so much for watching and letting me know - I've had a great time with this project so I'm glad people are enjoying it - cheers!

Good thinking, but some side-projects would feel too much like our day jobs :)

That's really cool - I like the idea of passing the retro tech concepts to new generations. Good luck on your journey - cheers!

Thank you kindly! It was a blast to make, so glad you enjoyed :)

oh boy, are you gonna hate the next rev... thanks for watching and cheers!

I really like how you said that - 'Programming is physical. But its not obvious.' - I think I've missed a lot of 'obvious' things because I didn't take the time to reflect and introspect. Its worth doing. On the students that you saw excel, where would you put curiosity as a factor?

of course - but optimization was never my point. Cheers!

THAT is cool AF. I especially appreciate the online simulator, playing with that now. Thanks for the pointer - I subscribed to your hackaday project and yt channel. Cheers!

@@wrongdogreckons-cp2sy thanks means a lot 🙂

Thank you! The original Raspberry Pi brought excitement back into computing for me - I even made magic mirrors for Christmas one year :) Discovering the RP2040 this past year has done it again - I appreciate what Rasberry Pi has given to computing.

Thank you! Its been a blast to make.

Thank you! Making the video has been one of the best parts of this project.

I appreciate that, this was all about trying something new. On your second point - I agree! There are so many optimizations to be had here - but that would make it feel too much like my day job to be fun :) In all seriousness though, recent updates have made it super snappy and my next steps on this are to consider how I can make use of the extra horsepower - its a fun experiment.

Thank you! Cheers!

Those are incredibly kind words and made my day - thanks!

Thank you! I appreciate that you get what I was going for here. Cheers!

Thank you so much! I absolutely will keep posting, have two in the works now :) Cheers!

You are correct - the latency is largely about communication timing and there isn't any scheduler. In the context of 'running user commands' it is not mult-tasking, whoever gets to the processor first gets full reign. A lot of my decisions were about the look of the piece and it looked better with more chips :) I do like that I have plenty of processing power left over to add functionality. Trying to read, but I am super confused about GreenArrays though - trying to Grok what they are doing there. Its great that they have 144 computers on that board, but am unclear on why...

@@wrongdogreckons-cp2sy I reckon most who hear of them are as bemused. Consider it an experiment in architecture; its closest relatives are TIS-100, FPGAs and transputers, each of different granularity. It's supposedly quite efficient but not clearly integrated either socially or industrially. I think a part of it is some old hand Forth enthusiasts saw the extensibility of their platform and wanted to build the new world. They just never stopped, while other things changed around them. A bit like GNU Hurd. They're not wrong, their systems are cool, but they could be improved with basic routing and recruiting people who grew up watching youtube on 64 bit pocket computers to work on 18 bit systems that don't even have peripherals may be a challenge.

Couldn't agree more - its fun to get into the basics :) Thanks so much for your kind words, and share your project out so we can see. Good luck!

Thank you - its a fun project to work on. Cheers!

That's amazing! Please share what you build!