Change of product direction, good and bad news!

[BruceMcF]

37 minutes ago, Scott Robison said:

The interesting thing is that I'm engaged with this exact topic in a FB group right now. I don't know what the credentials of the other person are, but my instinct is that it is someone who's never worked with FPGA.

Part of the confusion is that we call it "programming" an FPGA, but rather than creating a binary or being interpreted, an "FPGA program" is used to generate a netlist which is then translated to a concrete hardware configuration for the FPGA.

[paulscottrobson]

1 hour ago, BruceMcF said:

Yes, using dictionary definitions rather than the technical definitions is part of it ... the technical usage of simulation for, eg, a soft 6502 core and emulation for, eg, VICE is not subject to the same confusion, while in casual English usage, "emulation" and "simulation" are not so sharply differentiated.

You could argue many CPUs emulate CPUs, with microcode ?

[Tatwi]

I think another important consideration is that we live in what is essentially the science fiction future of the 1970s and 1980s, where people have computerized high resolution graphical tablets that they carry around with them everywhere. At the bear minimum this means that it's possible to take the amazing VIC20 printed manual and turn it into a fully functional, interactive 2D/3D representation (in the form of an HTML5 website) of the VIC20 hardware and software that works not only on one of those nifty graphical tablets, but on literal millions of other devices too, from cheap Chromebooks to million dollar workstations.

In short, the X8 and X16 do not exist in a vacuum and I think it is kind of silly to pretend that they do.

For those who do not wish to assemble a full sized X16 kit, the only difference between an entirely FPGA version and one made with discreet components is the physical size of the circuit boards. Otherwise the machines are functionally identical. In today's day and age, where a manual can so detailed as to be a complete virtualized replica, if you're not interested in the kit aspect (and you lack the electronics lab to maintain and troubleshoot said kit), then it doesn't really matter what's on the physical circuit board.

Even good old block diagrams and circuits are plenty good enough to teach people what's going on inside the computer, how their code gets translated into physical actions. Very few people are going to sit there with an oscilloscope and diagnose a circuit just to learn how their assembly code makes a sprite function, etc. But if that were truly important, the community could absolutely make an interactive website that allows a person to explore those very concepts without having to own a lab or even know what a multi-meter is. If the X16 was really about teaching people how an 8 bit computer works, such a website is the most accessible, affordable, and contributor-friendly way to go about it, because it would simply be available to everyone all the time, just like YouTube videos.

I admit that this was a bit of a rant, but I find the pedantic nature of some of these discussions to be very frustrating. And that's just reading them; I really don't participate in the "internet" anymore.

[Ed Minchau]

3 hours ago, Starsickle said:

 Software is not developed to completion and subsequently supported by a single guru with unlimited time in a garage, anymore. It is the exception now, not the rule. Waiting for manna from heaven is not how you get a killer app. You need a development environment and ecosystem with high level ease-of-use. The cuttthroat University CS department "Make-everything-as-arcane-as-possible" way is not the way of real life.

It's probably up to the members of this group to write the code for development tools.  I wrote an assembly language editor that's really a fancy Monitor program with labels, and I've been using that a lot, but most of my development for the X16 has been on spreadsheets.

Other people have worked on things like sprite editors and text editors and palette editors and music sequencers, and I believe David Murray has a version of PETDRAW out for the X16. There's also a lot of c libraries being written by people here.

[Wonderdog]

I still reckon the X8 could be a really cool devices if the video memory access mechanism could be modified to work just like the X16's will for code consistency, then functionally its just a cut down Memory + sound hardware version of the X16, but otherwise identical in operation. I think that's something I'd buy for <$100.

 

[Carl Gundel]

25 minutes ago, Wonderdog said:

I still reckon the X8 could be a really cool devices if the video memory access mechanism could be modified to work just like the X16's will for code consistency, then functionally its just a cut down Memory + sound hardware version of the X16, but otherwise identical in operation. I think that's something I'd buy for <$100.



 

I agree.  As much as the X16 can be a compatible upgrade for X8 users, the more compelling it is, I think.

[BruceMcF]

3 hours ago, paulscottrobson said:

You could argue many CPUs emulate CPUs, with microcode ?

Quite so ... any abstract distinction is prone to running into border cases like that.

[BruceMcF]

1 hour ago, Wonderdog said:

I still reckon the X8 could be a really cool devices if the video memory access mechanism could be modified to work just like the X16's will for code consistency, then functionally its just a cut down Memory + sound hardware version of the X16, but otherwise identical in operation. I think that's something I'd buy for <$100.

We'd have to ask @Frank van den Hoef whether that is even feasible ... that is, whether the X8 has the spare logic resources to do the auto-incrementing memory access.

If it's feasible, then a 32byte space added to the I/O device control space (which is I think are part of the $0600 page?) would allow CX16 binaries to be easily patched to run on the X8 ... as long as it fits in the RAM and Video RAM requirements.

[Scott Robison]

1 hour ago, BruceMcF said:

Quite so ... any abstract distinction is prone to running into border cases like that.

Exactly. Someone else brought up that very point regarding microcode. The other side in the debate seemed unwilling to accept that anything other than a 100% transistor for transistor, mask identical copy, is "real" and everything else is "emulation".



Now, to be fair to them, their point was that emulation is often used in a pejorative sense and that it isn't a dirty word and that calling FPGA emulation isn't meant as a bad thing. My point was "you can have good and bad software emulation, and you can have good and bad FPGA implementations; the difference is not the tools, it is the craftsman using the tools; regardless, words and more than the sum of their dictionary definitions, and quality software emulation is sufficiently different than quality FPGA 'emulation' that they deserve different words to describe them".

Edit: Fixed a typo and "the difference is the tools" should have been "the difference is not the tools, it is the craftsman using the tools". Too little time spent proofreading.

[BruceMcF]

Yes, the reason that there is a technical distinction is not because one is good and the other is bad, but because they are different ways to replicate the functioning of existing hardware with different strengths and weaknesses.

Of course, just as emulation is not the only thing software can do, simulation is not the only thing that SPLDs / CPLDs / FPGAs can do ... in the case of Vera, it is implementing an all-new linebuffer VGA display.