What I think an MSX 3 Should be ? | MSX Resource Center (Pagina 7/24)

Resident (45)

01-08-2015, 10:50

Perhaps a group of hardware guys should get together to design an MSX 3 or some sort of Crowd-funding initiative could be made to do a serious attempt at an MSX 3 to keep the system alive well into the future.

Van anonymous

incognito ergo sum (116)

01-08-2015, 13:23

Follow this recent thread but there are no reasons to be optimistic on this matter. Like you, I also think that joining forces this work can be done 'easily' but both people and developers do not show much interest...

Van Grauw

Ascended (10821)

01-08-2015, 12:41

I hope those hardware guys do it in cooperation with the software guys .

Van maxis

Champion (512)

01-08-2015, 13:07

Putting my 2 pennies worth, IMHO, before embarking on the MSX3 HW definition, we would better finding the answers to the following questions:

1. What is the purpose of MSX3 and who is the potential user of the new platform: educational, entertainment, hobby, etc.... ?
2. What kind of applications we would like to run beyond MSX1-MSX2+ legacy, which require the radical increase in the CPU/GFX performance (multi plane platformers/SHmUP/3D)?
3. What SW/FW tools we have to develop to support the new HW standard? C/C++/C#/Java compilers, OpenGL libraries, graphic constructors etc (you name it)?
4. What is so terribly wrong in the MSX2+/TR architecture, which doesn't allow to use the existing HW concept?
5. How the backwards compatibility will be achieved?

Benchmarking the past attempts, let's see how many SW titles were developed after the commercial demise of MSX TR? Also how many new SW titles were developed on the commercially successful HW like MoonSound and GFX9000 for example?

One more thing: how many TR R800 native applications including the famous game titles were created?

Van Pac

Scribe (7116)

01-08-2015, 14:15

maxis wrote:

Putting my 2 pennies worth, IMHO, before embarking on the MSX3 HW definition, we would better finding the answers to the following questions:

1. What is the purpose of MSX3 and who is the potential user of the new platform: educational, entertainment, hobby, etc.... ?

To keep enjoying MSX experience on a real hardware. Our machines are more than 30 years old and probably will fail sooner or later. Yes we already have 1ChipMSX (and clones) or emulators but in my opinion they are not the same and the current FPGA design is limited. I'm not criticizing the FPGA based platforms nor emulators, they are really useful but I obviously would prefer a platform as close as a real MSX. Besides with current knowledge and technology we could improve some aspects. Potential use? Hobby, of course...

Quote:

2. What kind of applications we would like to run beyond MSX1-MSX2+ legacy, which require the radical increase in the CPU/GFX performance (multi plane platformers/SHmUP/3D)?

I'm not asking for a super computer with a powerful CPU, 3D graphics and so on, just a little more powerful and modern hardware based MSX. We shouldn't change the spirit of MSX as Tsujikawa (the 1ChipMSX designer) told me by e-mail time ago, do impressive things with limited hardware.

Quote:

3. What SW/FW tools we have to develop to support the new HW standard? C/C++/C#/Java compilers, OpenGL libraries, graphic constructors etc (you name it)?

Not sure about this...

Quote:

4. What is so terribly wrong in the MSX2+/TR architecture, which doesn't allow to use the existing HW concept?

What do you mean with the existing HW concept?

Quote:

5. How the backwards compatibility will be achieved?

We know that the MSX has pros and cons but that is its distinctive feature. Keeping the current architecture is a must if not that supposed computer wouldn't be an MSX, that's clear. I mean a Z80 core, MSX VDP, etc

Van maxis

Champion (512)

01-08-2015, 15:27

PAC wrote:

To keep enjoying MSX experience on a real hardware. Our machines are more than 30 years old and probably will fail sooner or later. Yes we already have 1ChipMSX (and clones) or emulators but in my opinion they are not the same and the current FPGA design is limited. I'm not criticizing the FPGA based platforms nor emulators, they are really useful but I obviously would prefer a platform as close as a real MSX. Besides with current knowledge and technology we could improve some aspects. Potential use? Hobby, of course...

PAC wrote:

I'm not asking for a super computer with a powerful CPU, 3D graphics and so on, just a little more powerful and modern hardware based MSX. We shouldn't change the spirit of MSX as Tsujikawa (the 1ChipMSX designer) told me by e-mail time ago, do impressive things with limited hardware.

PAC wrote:

Keeping the current architecture is a must if not that supposed computer wouldn't be an MSX, that's clear. I mean a Z80 core, MSX VDP, etc

And I do agree with you, that in order to keep the MSX in a name, it must be based on the STANDARD CPU and VDP. So, MSX2+/TR are based on ASCII V9958 VDP. Unfortunately we don't have MSX3 VDP specification in hand to create it (we could add 9958 features into 9990 specification).

IMHO, unless we change the memory architecture, the CPU instruction set or create a new VDP, it would be difficult to name the new creation as "MSX3".

So, in other words, we are not looking for MSX3, but for MSX2+ or TR, which is faster. Let's say 10 times faster, than the standard 3.57MHz MSX2, right? Not only the CPU, but the VDP data transfer too.

The difference between the SW emulation and the FPGA is that FPGA can be cycle accurate or EXACT when comparing to the real MSX. It is only the matter of the time investment into the design.
You can read C64 Direct To TV story and some comments about the design being a cycle accurate emulation of C64.

Van Grauw

Ascended (10821)

01-08-2015, 15:42

maxis wrote:

1. What is the purpose of MSX3 and who is the potential user of the new platform: educational, entertainment, hobby, etc.... ?

For me I would like to have a MSX turboR+, the machine which could have been. I’m talking as a hobbyist and msx fan, I think any effort for a broader appeal is doomed (MSX Association already tried it a decade ago ).

Unfortunately all of the new MSX hardware (OCM, GR8BIT) is built on MSX2/2+ specs, and since turboR computers are expensive and relatively hard to get, I hesitate to fully utilise them, never mind require one.

maxis wrote:

2. What kind of applications we would like to run beyond MSX1-MSX2+ legacy, which require the radical increase in the CPU/GFX performance (multi plane platformers/SHmUP/3D)?

4. What is so terribly wrong in the MSX2+/TR architecture, which doesn't allow to use the existing HW concept?

One more thing: how many TR R800 native applications including the famous game titles were created?

For me, developing turboR software is appealing because the CPU speed is so high. Both VGMPlay and Synthesix benefit from it, though often subtly (reduced latency).

For a lot of Basic(-kun) and C programmers, the increased CPU speed is also really useful. You can see it with a lot of turboR software actually (e.g. Illusion City), it has nothing that could not be done in highly optimised Z80 assembly code as well, but writing such code increases the cost of software development and reduces the pool of available developers. Because of this higher CPU power directly increases the amount of software available even with other specs remaining the same.

Graphics wise, personally I would like to see a V9990 built-in. I know it’s hard to get and it doesn’t exactly have a stellar track record in terms of how much software is developed for it, all good points, but the new pattern modes, expanded palette and improved blitter speed are just too delicious to ignore, and it was originally developed for the MSX3, after all . Some pretty amazing things can be made with it, while the hardware does stay within “the spirit of MSX”.

That said, I’d also be happy with just a V9958, it’s still got quite some unused power to unlock.

Sound-wise I’m fine with PSG + OPLL, plenty of OPL4 cartridges around and having it built-in would just further increase the cost.

Van sd_snatcher

Prophet (3675)

01-08-2015, 15:46

Adding to what PAC mentioned, I would also say:

1a. We have ancient hardware that soon will be so classic that no one will want to modify or maybe even get it out of the glass case. Just like it happens with old/classic cars

1b. The newer MSX generations integrated hardware that were clumsy external extensions in the previous generations. It was the case of the slot expander, disk drive, Lightpen, MSX-Music, etc etc etc. It's now the case of V9990, OPL4 and IDE (or Flashcard), MegaROM emulation. The V9990 is the clumsiest of them all, because of its separate video output. An internal V9990 superimposed over the V9958 would solve this perfectly.

2. All existing OPL4 software, and the very few V9990 software, i.e. IMHO, there's no need to get wild in the specs. We all know more or less what the MSX3 specs were supposed to be: R800, V9978, maybe OPL4. And we add some handy modern extensions like IDE/Flashcard and a better megarom emulator than the one Panasonic had built-in on the Turbo-R. (Where in the world are we going to get R800s? Well, that's a too big question for this moment, when people are still not convinced it it's worth designing the machine itself)

3. Probably the same we already use for the existing MSXs.

4. Nothing terrible. The main points were explained on (1) and (2).

5. Keeping inside the lines detailed in (1) and (2) wouldn't break the compatibility in almost any way. One or another program would maybe have trouble, but probably those programs already have trouble in some of the existing MSX models.

But more important than everything else: this mustn't turn into a sacred war. It's something for grown up men discussing like good friends drinking beer in a bar.

If it's something to bring up anger, frustration and ferocious discussions, I rather leave the MSX3 as a failed chapter of the past and keep my good friendships here. No MSX3 is worth more than the 25+ years old friendships I have here.

Van maxis

Champion (512)

01-08-2015, 15:42

Grauw wrote:

For me I would like to have a MSX turboR+, the machine which could have been. I’m talking as a hobbyist and msx fan, I think any effort for a broader appeal is doomed (MSX Association already tried it a decade ago ).

Unfortunately all of the new MSX hardware (OCM, GR8BIT) is built on MSX2/2+ specs, and since turboR computers are expensive and relatively hard to get, I hesitate to fully utilise them, never mind require one.

Exactly to the point!

Van Grauw

Ascended (10821)

01-08-2015, 15:55

sd_snatcher wrote:

(Where in the world are we going to get R800s? Well, that's a too big question for this moment, when people are still not convinced it it's worth designing one)

It can be solved with new FPGA MSX-ENGINE. You probably already want it anyway to reduce the amount of logic and to keep the board size small. Not only include a Z80, PSG, PPI and similar logic, but also an R800 CPU implementation, copying the architecture from the turboR with two CPU modes and the MULUB / MULUW instructions and a 3.58 MHz bus. Timing wouldn’t even need to be identical to the R800 (though it would be nice), just as long as it’s in the same ballpark.