SMS2 in FPGA

[XorA]

"What kind of nonsense is this"

I mean that was my response to some of TTs writings! Not that I am an expert in OSes but Ive written a couple for various products!

[M68008]

A while back I had a personal connection to ARM's board of directors. I used this connection to convince ARM to look at TTs work. The were given a demonstration and in a nutshell their response was "What kind of nonsense is this"

Funny anecdote I had not heard before.
Even assuming SMS2 was state of the art when written, you should not expect that computing has remained still for the past 30+ years.

[Tinyfpga]

Post by RalfR » Sun Jun 15, 2025 2:37 pm - Link?

Found it as follows:- viewtopic.php?t=4679&start=40
8th post, 2nd image.

or download/file.php?id=7331&mode=view for image

[Peter]

Presumably one or two of the small ICs to the left of the FPGA.

Yes, the lower one.

I assume code has to be loaded into flash memory from an external source

Yes, initially with a JTAG adaptor. Later on there is an SMSQ/E program which can also do the job.

that the code has to reloaded into the the FPGAs on power on?

Yes that happens automatically.

[Peter]

I remember TT saying that C is not a suitable language to express his OS ideas. In particular the work involved in hand checking the object code for errors would be greater than simply rewriting the OS in assembler for another processor.

Just that rewriting his OS in ARM assembler is the opposite of simple. In reality, Stella is bound to the 68K architecture.

[Dave]

In reality, Stella is bound to the 68K architecture.

^^^ This

It also has other failings:

* It's wasn't finished
* There is NO Stella software
* Documentation doesn't say much about hardware expectations
* No toolchains for developers
* You'd be a user-base of 2-3 people


You should see the huge list of everything it's got going for it though!

[RalfR]

Post by RalfR » Sun Jun 15, 2025 2:37 pm - Link?

Found it as follows:- viewtopic.php?t=4679&start=40
8th post, 2nd image.

Well, this is a list of files, but not a piece of source code.

[Tinyfpga]

I thank Peter for answering my FPGA startup questions. I had asked the same questions elsewhere but received no answer.
Is system software bound to 68K really a problem? What's wrong with the 68K?

I had another thought on SMS2 in FPGA. There is a version of SMSQE that runs on Atari STs using the same drivers as SMS2. SMS2 runs on Atari STs in the same way. Thus it would seem that, likewise, SMS2 should be able to run on his FPGA hardware.

I wrote that I had posted a copy of the files not the source code contained within those files. It would be for the author to decide to distribute the code itself.
With a potential user base of one, sorry I meant two or maybe even three, I am not sure it would serve any purpose.

Dave wrote:- "You should see the huge list of everything it's (TT's stuff) got going for it though!"
I have seen that list and I find it impressive, but do I detect a smidgen of sarcasm in his sentence.

Is UNIX and its ilk, dating from the 1960s and still reflecting the hardware available at that time, really "cutting edge" technology, with its need, on a personal and single user computer, for an administrator and an MMU.
What kind of nonsense is that!

Is QDOS, SMS2 and SMSQE, just a load of old nonsense? Do the members of a forum that use these OSs really think that.

[tofro]

I had another thought on SMS2 in FPGA. There is a version of SMSQE that runs on Atari STs using the same drivers as SMS2. SMS2 runs on Atari STs in the same way. Thus it would seem that, likewise, SMS2 should be able to run on his FPGA hardware.

I don't quite understand why you are so fixated on SMS2. SMS2 is basically a stripped-down version of SMSQ/E and has zero features beyond SMSQ/E. SMSQ/E is a maintained operating system (admittedly, with very few active developers, but still more than the zero for SMS2) that is constantly being updated and ported to new platforms. SMS2 is not. I don't get why you don't simply use SMSQ/E - There's nothing you cannot do with that that you could do with SMS2.

Is UNIX and its ilk, dating from the 1960s and still reflecting the hardware available at that time, really "cutting edge" technology, with its need, on a personal and single user computer, for an administrator and an MMU.
What kind of nonsense is that!

Well, there's reality and there's wishful thinking. Reality is that Unix/Linux concepts have been adopted by the industry and thousands of developers are working pushing it forward. There's no point in shouting "we've got something better" with a handful of developers behind it (and, as sad as it may sound, I doubt we "have something better").

Why do I say that:
The QDOS family (that includes SMS2, SMSQ/E and likely Stella) is tightly bound to the M68k family of CPUs. There's absolutely no point in "porting" an OS written in m68k (or any other) assembly to another CPU. That would end up in a complete re-write. That is why modern operating systems are written in higher-level-languages to be able to keep the continuity, even if that may have a performance penalty. An ARM version of QDOS, for example, would be a pain to write (as anyone who has ever done ARM assembly will likely agree to) and a huge amount of work. It is impossible.

Is QDOS, SMS2 and SMSQE, just a load of old nonsense? Do the members of a forum that use these OSs really think that.

Nope. Otherwise this forum and the enthusiasts around it wouldn't exist. But these OSs have set their bets on a boat that has sailed. They were really, really good OSs at their time, but are stuck with an architecture that is gone forever. Of course you can proceed with FPGA-based hardware which could maybe achieve x-100MHz performance if you invested a lot of effort and money, but still, modern CPUs that work at multiple-GHz ranges would run circles around it.

Not everything is negative. Working with QDOS and SMSQ/E is still a lot of fun - I think the members of this forum agree with me that we should simply enjoy what we have and keep on tinkering. Tinkering, it is, because for "productive, modern" work, our OS is simply missing way too many features that are taken for granted today (I won't even start listing those).

[RalfR]

We're actually talking about "hardware" here, but what the heck.

Great, now we've seen a list of individual files from the Stella source code. And we know (by hearsay) that there's no user interface for it. So why bother with it? There are certainly still a few things that SMSQ/E is missing.

For example, an alternative hard disk driver to remove this, in my opinion, unspeakable association of "subdirectory" and filenames. Never mind whether that might cause problems with different programs; we have DEV, PTH, and SUB for that. That would be something, but the question is (as always), who would develop something like that? I don't know how previous hard drive implementations (e.g., the drivers for the "Steinkopf" hard drive) handled this. Was it the same there?