Commander X16

On 10/9/2021 at 8:30 PM, Tatwi said:

...



On a personal note, having now read quite a lot about how the VERA is programmed and the implementation of BASIC 2.0, the plain truth is that I am too stupid to program this machine. C and even wacky C++, fine, but my brain just refuses to juggle hex/binary /assembly/register mumbo jumbo. And it's not like it's new to me, I've failed to grasp it for at least 25 years. I'm too stupid to program for PET and VIC20 too. As such, I'd rather continue doing my own thing in languages/platforms that already work for me; i won't be programming for the Commander x16. ...

Bear in mind that Assembly and Basic won't be the only options for programming hosted on the X16. There is going to be others. The fact that they will mostly be loaded from the SD rather than being in the ROM doesn't mean much at all given the speed that it reads files from the SD card.

There's already a Forth, and the beginnings of a Small-C written in Forth, and the list will surely grow.

On 10/9/2021 at 5:30 PM, Tatwi said:

the plain truth is that I am too stupid to program this machine. C and even wacky C++, fine, but my brain just refuses to juggle hex/binary /assembly/register mumbo jumbo. And it's not like it's new to me, I've failed to grasp it for at least 25 years. I'm too stupid to program for PET and VIC20 too.

Ha, I can relate. I love my 1981 VIC20 and like my 2020 TheVIC20 (despite the TheVIC20's hacky memory expansion issue). I can program the VIC in Basic (with a book of PEEKs and POKEs flopped open nearby) but once hex and assembly come into view my eyes glaze over. Basic 2.0 programs running at the 8-12Mhz speeds of this X-project might make assembly unnecessary in a lot of cases.

I program the old-fashioned way: tapping away at the petscii keyboard and frequent checks of ?FRE(0). No PC or modern PC tools involved. I find it to be almost as engrossing as it was 40 years ago. I only wish there was some way to output TheVIC20 listings to a tractor-feed dot-matrix printer so I could reproduce the full 1980s programming ambiance.

The C128's Basic 7.0 was a thing of beauty. That was a Basic that really opened up access to the computer's capabilities. Note to the 8BG: I demand you implement the X16 project with my pet preference of Basic 7.0 or you will have failed the project and proven in front of the entire world and for all time that you are a terrible leader! Wah-wah-wah!

On 10/7/2021 at 5:42 AM, Ju+Te said:

@sundownAt least for the VERA an FPGA seems to be necessary, unfortunately. So having an FPGA-free X16 based solely on standard "factory new parts" seems not to be an option.

That may be the case for the X16, however, it is not he only system out there in development.

I hear the whole "cannot build a (dream) computer without...... (compromises like FPGA)" all the time everywhere from everyone. I reject it for what it is, nonsense. Rubbish. FALSE (goto start).

There is no barrier to producing a $50 SVGA (or better) hobbyist computer. That is the fact. The only thing lacking is firepower (in the design department). It is not impossible, it is in fact inevitable.

On 10/7/2021 at 8:47 PM, smartroad said:

I was wondering if you were going to go down the FPGA route with no expansion, why not just have a small ARM board and run an emulator? I understand that FPGA is hardware but from the end users perspective it is just a chip on a board doing a thing, it is almost irrelevant if it is fpga hardware or software emulation. Have a custom Linux install for say a Raspberry Pi which can boot straight to the emulator and thats it, a Commander X16/X8 without custom hardware.



Not trying to be controversial but the question is what is more important, the hardware, the software or both? If hardware go Phase 1, if software Phase 3/Arm emulation, if both then Phase 2.

[Quoting the deceased designer in I, Robot]  " That [] is the right question. "

On 10/8/2021 at 6:43 AM, maktos said:

Considering that it doesn't cost anything to post an update.

Have you met the internet ?

The less time you spend on the internet the more productive you are. There have been shows made about the problem with the internet.

Just adding my thoughts to the discussion about why the Commander X16 is worth the wait.

I've spent all year so far working on a DirectX 12 project, with the goal of designing a 3D game for Windows desktop that's neither an FPS nor a flight sim. It's all being developed in Visual Studio/C++/HLSL, ie. not in Unity or any other game studio. My game framework so far successfully implements many features expected of a modern 3D game environment, including bump mapping, dynamic reflections, real-time shadows, a first-person camera, ability to pick geometry by mouse, and collision detection. It also supports post-processing effects such as HDR tonemapping, bloom and depth of field.

However, after a recent month-long battle to get skeletal mesh animation working (which was finally solved), the burnout finally set in. What got me through each day of DirectX pain was turning to my favorite YouTube channels, namely 8-Bit Guy, LGR, RMC Cave, and Retro Recipes to name a few (see the pattern here?)

Yes, despite being up to my neck fussing with modern PC graphics, I can't wait to wind down on YT with a classic DOS, C64 or Amiga game review, or a show-and-tell with a classic piece of computing hardware from days gone by. So naturally I did see all of David's episodes about the announcement and subsequent progress of the X16, but after sighing wistfully, it was mentally filed away and I thought if only I wasn't committed to modern-day 64-bit game programming for Windows...

So there was a lot of internal conflict going on between what I really wanted to do deep down, and what I 'ought' to be doing. My justification for continuing with the current project was along the lines of, well, imagine if kids and developers had access to today's graphics hardware in the 80s? Wouldn't it be a dream come true? It's a reality today, so why not revel in it and take advantage of the latest technology instead of limiting ourselves to 8-bit era computing power?

Well take it from me... I've found out the hard way that the latest and greatest GPUs and 3D APIs may be a boon for gamers and AAA game studios, but for the indie developer or hobby programmer it can be an absolute nightmare. Firstly, if you aren't able to cope with matrix math or pixel shader lighting equations then 3D programming isn't for you. Assuming you do grasp the math and physics of lighting, then there are the 3D APIs to wrangle. You absolutely MUST have experience with previous versions of Direct3D or OpenGL if you want to program with the latest generation APIs (DX12/Vulkan). Ok, assuming you have previous 3D experience.  Sure, with a few tutorials you'll be able to render a few spinning cubes and even maybe a textured 3D model imported from a tool like Blender. But to populate a reasonably decent game world will take hundreds of hours at the very minimum. Building an animated 3D character and its associated textures and armature rigging alone may take weeks... and of course you'll want more than one game character. Then what about the game world itself, and physics, and AI?

The crunch for me came when I realized my 'game' would never progress beyond the engine phase. There are just too many literal moving parts for an individual game designer/programmer to create and manage on one's own. And of course, when developing for Windows there will be change. We're on the cusp of Windows 11 right now, and that likely means an update to DirectX isn't far behind. In other words, be prepared to push that boulder all the way up to the top of the hill again.

And then I remembered the X16 project and looked at Matt Heffernan's 65C02 assembly tutorials. Luckily I've got some x86 assembly behind me, so I dived right in and I haven't looked back! It's helped me break out of an unending loop and I'm certain I'll have something to show for it at the end. Programming for the X16/VERA graphics chip emulator has brought back the fun and anticipation that had been drained by trying to go it alone with a Windows game. That's why waiting a little longer for the final product (even in kit form) doesn't bother me.

So that's why I'm a convert and I'm sure I'm not alone. As far as I'm concerned the X16 has already achieved its goal to make coding fun again. A bit long but... thanks for reading!

 

On 10/10/2021 at 2:45 AM, Oldrooster said:

That may be the case for the X16, however, it is not he only system out there in development.



I hear the whole "cannot build a (dream) computer without...... (compromises like FPGA)" all the time everywhere from everyone. I reject it for what it is, nonsense. Rubbish. FALSE (goto start).



There is no barrier to producing a $50 SVGA (or better) hobbyist computer. That is the fact. The only thing lacking is firepower (in the design department). It is not impossible, it is in fact inevitable.

I think it is important not to over-generalize here. There is an argument that it is not possible to build Dave's dream computer without a custom fabricated video chip, and the only feasible way to create a custom fabricated video chip at this scale is using an FPGA with the build in hardware that can be wired together to perform the process ...

... which is an argument about Dave's 8bit dream computer.

It is not an argument about an 8bit dream computer in general. If someone has a dream that can be satisfied with a pure tile display, or with a screenbuffer chip, there seem to be options that will work. It's the tile and hardware sprite implemented with a scanline by scanline rowbuffer display that requires either an FPGA to wire the circuit together or an MCU to emulate a notional custom fabricated chip.

So people shouldn't make unwarranted generalizations. It is certainly possible  to dream different specific dreams and avoid the choice between an FPGA implementation and a software emulation of a notional chip.

As for that specific choice, for a system that has a design goal of being built to the extent possible with in production ASIC chips, I have no hesitation in arguing that a netlist wiring of hardware is closer to an ASIC chip than CPU emulating the behavior of a notional chip. But if somebody's dream was a "neo retro 6502 GEOS system", there are plenty of framebuffer chips to choose from.

On 10/10/2021 at 1:26 PM, BruceMcF said:

So people shouldn't make unwarranted generalizations. It is certainly possible  to dream different specific dreams and avoid the choice between an FPGA implementation and a software emulation of a notional chip.

This. So much this. All the this.

I'm 53. I missed certain eras of computing. I wonder if there was as much pushback from earlier generations against ASICs instead of all discrete components? Or if there was pushback against the transistor when it replaced the vacuum tube?

It brings to mind the quote from (I think) Arthur C. Clarke that any sufficiently advanced technology is indistinguishable from magic. My corollary would be "to some people all magic is black magic and to be avoided."

On 10/10/2021 at 4:04 PM, Scott Robison said:

This. So much this. All the this.



I'm 53. I missed certain eras of computing. I wonder if there was as much pushback from earlier generations against ASICs instead of all discrete components? Or if there was pushback against the transistor when it replaced the vacuum tube?



It brings to mind the quote from (I think) Arthur C. Clarke that any sufficiently advanced technology is indistinguishable from magic. My corollary would be "to some people all magic is black magic and to be avoided."

I think that the early ASICs were such an incremental change from discrete components that there might not have been any pushback, but I could easily imagine mini-computer designers arguing "CPU's will never be able to replicate the capabilities and extensibility of a central processing board build from bitslice chips" ...

... and while the computer industry seems like they were enthusiastic in switching from vacuum tubes, the transition was much slower in radio ... remembering how the "cheap transistor radios" of the 60s were "cheap Japanese imports" (which was still a thing in the 60s), while "high quality" radios were still bigger vacuum tube based designs. 

I suggest you release the X8, and call it X16.

Because you guys have been working for years on this and proven how serious you are about this.

Now, I'd like to see, how serious this community really is about this project.

Is the community able to leverage on the full capabilities of it, and provide a number of titles that makes the Commander a thriving 8-bit platform?

Who needs more RAM, when loading from flash is fast?

Who needs more VRAM? I don't want to play Wing Commander on the Commander X16, I can do that on an Amiga just as well.

Heck, even a 12 MHz CPU already feels like it's missing the point of this project.

I think the X8 is the perfect way to find out, what this community really is ready to contribute.

The X8 won't fail because of too less RAM.

It will only fail because people love to dream and talk about things much more than actually contributing to it, and developing for it.

Call the X8 X16, because it's fine as it is, and it allows this 8-bit platform to come to live.

And only if the X8 proves to be successful, go back to the original plan, because only then you (and we all) know, that the additional investment is worth more of both, your time and money.

Thanks for working on this! (you made me look into 6502 assembler again ? )

(on another note: I'm not on Facebook, so I'd highly appreciate a different channel as primary source of news @The 8-Bit Guy)

While I don't share the grave interpretation @maktos put forward, I do have to agree it seems like a mistake to not post any kind of follow-up yet. I don't share the idea that saying nothing is better in this instance, nor that saying something might require unreasonable commitments. So I think its pros would outweigh any cons. IMO.

The audience here does not seem like an unreasonable bunch, so there is that. It also seems David does post on Facebook, so there is that as well. A quick summary to similar effect or even just a quick word every now in the spirit of "planning still going on!" would seem like a generally good idea given the significance of this thread.

In fact it is something that I readily expected when reading through this thread a few weeks back and was surprised to find out did not happen already.

On 10/10/2021 at 2:46 PM, Janne Sirén said:

While I don't share the grave interpretation @maktos put forward, I do have to agree it seems like a mistake to not post any kind of follow-up yet.

I can agree with this just because of perception. The grave interpretation was I think what was being argued against, not that updates wouldn't be welcome. Just that I can understand why updates haven't been forthcoming.