Could it make sense to reuse existing hardware for the X16_fpga?

[Ju+Te]

E.g. the SiDi FPGA or ZX-Uno+? That way people could get hands on some real hardware instead of just an emulator. Those people who want the DIL throughhole version of the real X16 will buy it anyway if it comes out after the chip shortage.

The only problem I see is, that there might not be an FPGA description of the whole X16 yet.

[Scott Robison]

I don't know if the rules have changed recently but I got a warning months ago for even talking about such topics.

The biggest problem is one of licensing and paying Cloanto for the ROMs used as the base of the X16 firmware. It would be possible to sell a core that came with licensed ROMs depending on the license terms which we don't know. It would also be easier to pirate.

But technologically speaking, it is a course the team could consider if they were so inclined. Just a lot of moving parts to orchestrate.

[TomXP411]

Yeah, we talked about building a MiSTer core, for example, and was politely asked to not discuss ports to other hardware. 

Which is a reasonable request, since the team had to license code from Cloanto.

On 10/31/2021 at 8:50 AM, Ju+Te said:

there might not be an FPGA description of the whole X16 yet.

There isn't a single FPGA X16 yet, but all of the components in the X16 already exist as parts of other FPGA projects. So the hard work would really be in combining those into one system.

Still... since we've been asked not to port the Commander to other hardware platforms, we should probably end the conversation before it gets locked. 

 

[Scott Robison]

On 10/31/2021 at 3:19 PM, TomXP411 said:

There isn't a single FPGA X16 yet, but all of the components in the X16 already exist as parts of other FPGA projects. So the hard work would really be in combining those into one system.



Still... since we've been asked not to port the Commander to other hardware platforms, we should probably end the conversation before it gets locked. 

Agreed, though I do think there is a difference to this question. Namely: I infer "has the X16 team considered using an existing hardware platform as a target for their FPGA X16 version?" as being the question. It may be that Cloanto only licenses the code to go with bespoke hardware, since they sell C64 Forever themselves. I posed a similar question to MEGA65 and the impression is that they are only authorized to include Commodore derived ROM with their physical hardware. More to the point, I wondered if one could sell a core that would allow money to be collected to be sent to Cloanto for a licensed ROM, and the answers I got indicated that wouldn't be possible. If that is true, they could create a cost reduced MEGA65 that didn't include the keyboard or FDD or other non-essential pieces of the system and include firmware with that, but they they probably wouldn't be able to sell a binary core that could be installed on MiSTer.

Given that MEGA65 has to get the license from the same source as Commander X16, it is probably safe to say similar if not identical terms would apply.

[Ju+Te]

There is the X16 hardware and the software. Each is IMHO independent of the other. Maybe licensing the early versions but planning to switch to an Open Source OS later would be a reasonable plan. Using other FPGA hardware as starting point could be a boost to quicker develop the Open Source OS while the X16_dil needs to wait for available chips. Sometimes, when the parameters change, it makes sense to change the plan, too, or reorder the plan's subtasks.

[TomXP411]

On 11/1/2021 at 12:51 AM, Ju+Te said:

There is the X16 hardware and the software. Each is IMHO independent of the other. Maybe licensing the early versions but planning to switch to an Open Source OS later would be a reasonable plan. Using other FPGA hardware as starting point could be a boost to quicker develop the Open Source OS while the X16_dil needs to wait for available chips. Sometimes, when the parameters change, it makes sense to change the plan, too, or reorder the plan's subtasks.

Not that I disagree, but this is also a huge potential can of worms.

First, VERA is not open sourced. So even though the 6502 and all of the other vital chips are available as open source FPGA code, the critical part - the video interface - is not.

But let’s set that aside for just a moment and examine the other issue: Let’s say someone created a working VERA reproduction without using any of Frank’s code. What happens then? The obvious answer is: people will borrow the ROMs from the emulator and load it into the FPGA computer. With or without permission. And since David and the crew will have no way to stop this, they will have to close down the emulator and lock us out of future updates until they have hardware to sell - or face issues with their Cloanto license.

The X8 was a potential way around this: by selling a computer that was “not quite but entirely unlike the Commander X16”, they would get a VERA computer out there for us to use, learn from, and funnel money into the project. More importantly, it would have been a good way to develop the open ROM, since it’s close enough to the X16 to be useful as a development platform, but not identical enough to allow the Cloanto ROM to run without modification. 

However, 58% of the people here said “No, we don’t want that.” So there will be no commercial release of the X8.

If someone was to build an FPGA core for MiSTer that’s similar to the X16 in concept, I’d certainly use it. I have used and enjoyed the MSX core, for example, which is very similar in concept to the CX16. In fact, the MSX is nearly exactly what David outlined in his original Dream Computer post, way back when.

However, with Frank’s code still being private, there’s no chance of actually building a proper Commander X16 FPGA core. Having said that - I would happily spend $50 on a CX16 core - even if it came with only the KERNAL and Supermon code. In fact, I’d kind of like to get away from Commodore BASIC, due to it’s super limited nature, and explore one of the other 6502 BASIC interpreters - ones that include some of the many commands missing from Commodore BASIC.

This would also have some interesting side effects - for example, what about releasing a 65816 version? What about a 20Mhz or 48MHz Z80 version? Since at this point, it would be just firmware, get away from the 6502 entirely. Making it impossible to use the Cloanto code certainly solves the licensing issue.

 

[Tatwi]

Excellent points Tom!

Personally, I actually prefer using QBasic (1.1 or QuickBasic 4.5) over other versions of BASIC, especially version 2 for Commodore machines. Code wise, it's just a better experience all around while still being nostalgic. For BASIC, I don't think it's necessary to go full PEEK/POKE everything just to be "true to the spirit of old stuff" or whatever. Maybe that's just me.

The machine's capabilities and the tools provided to make content for it are truly what's most important. Who cares if it's compatible with existing machines that anyone can emulate and have been done to death? Like... just go use that stuff, right? Otherwise one could sensibly ask, why use Commander stuff? Especially when there isn't even a Commander machine to own anyway.

VERA, cool. The rest of it? Meh... At this point I don't know why I should care; There are already so many other options for BASIC and Assembly programming.

[BruceMcF]

On 11/1/2021 at 3:04 PM, Tatwi said:

Excellent points Tom!



Personally, I actually prefer using QBasic (1.1 or QuickBasic 4.5) over other versions of BASIC, especially version 2 for Commodore machines. Code wise, it's just a better experience all around while still being nostalgic. For BASIC, I don't think it's necessary to go full PEEK/POKE everything just to be "true to the spirit of old stuff" or whatever. Maybe that's just me.



The machine's capabilities and the tools provided to make content for it are truly what's most important. Who cares if it's compatible with existing machines that anyone can emulate and have been done to death? Like... just go use that stuff, right? Otherwise one could sensibly ask, why use Commander stuff? Especially when there isn't even a Commander machine to own anyway.



VERA, cool. The rest of it? Meh... At this point I don't know why I should care; There are already so many other options for BASIC and Assembly programming.

Yes, I would hope that someone would develop a QBasic for the X16.

[kelli217]

On 11/1/2021 at 2:42 PM, BruceMcF said:

Yes, I would hope that someone would develop a QBasic for the X16.

I would actually like something similar to Visual BASIC for DOS, but I am probably in the minority on that.