snout:

Why is it taking so long to put all the information about the contents of the package, casing, code license, etc ? If all information is not available, why not release what they have now ? There are only 30 days left to deadline, if this information can help increase the sales, why they don't release them ?

I agree...

Probably because all information is released, they will know that is a lie. I will not buy still without knowing ALL about it...

That, of course, is your personal choice, but it seems to me that you are only looking at what you don't like about the One Chip MSX in its out-of-the-box config, completely overlooking the potential of the device. Just to give a few examples (some of which were given by Javi Lavandeira on the HispaMSX Mailinglist)

• The FPGA chip used is fast enough to allow an implementation of a Z80 running at 84MHz or even 105MHz. Just imagine your MSX running on this speed. How can you think of that as not cool? ^_^
• The 32MB of memory on board can be used for a wide variety of things. How about having an internal mapper of 4MB, 4MB VRAM (so how many SCREEN 5 pages is that? and a couple of MB for samples for your own VHDL audio device (or an implementation of the Moonsound, for that matter).
• After optimization of the MSX2 config, an estimated 42% of the amount of Logical Elements available is still all yours. Now what you have is an MSX2 with 4MB RAM, 128kB VRAM, MegaSCSI, MSX-MUSIC, SCC, MSX-DOS2 and a FAT16 compatible hot-swappable SD card slot. If all that takes 58% of the amount of LE's available, imagine what you can do with the rest. Some very interesting optimizations (increasing the Z80 speed, implementing ADVRAM, adding more VRAM, implementing V9958 compatibility, implementing Ymixer, reducing sprite color and amount-per-line limitations, adding higher resolutions/more colors) don't have to take so many gates at all. An optimized Z80 core that uses less than 50% of the amount of LE's it's currently using already is close to completion

To wrap it up with another Javi example: if you see an MSX1 with F1 Spirit plugged in, turn it on and see the game. Do you consider the entire machine to be 'just a racing game', or are you going to look beyond the capabilities of the device?

That is all nice, exiting and very ^_^, but we're not exactly there yet, are we? Also, I fear that for every future 1cm game/demo/whatever we need a different VDHL core. Not much of an MSX standard then, is it?

I can surely see the potential of the 1cm, but I cannot see it all happening. Not -yet- anyway.

I'd say, if every game/demo/whatever doesn't use the standard MSXA-certified stuff then it prolly needs to come with its own VHDL .. just like Konami came-up with the additional SCC for their games. Since the FPGA can be updated with an MSX-DOS tool, I guess one can perform this 'update' aswell without that tool, meaning: in your own software. So I see the following option:

- save FPGA-content to a temp-file on SD-card
- upload/adjust FPGA with own VHDL
- run game/demo
- quit game/demo
- restore original FPGA content from SD-card

Naturally, if one's VHDL-code contains bugs, it might ruin your 1cm. So I bet you need that cable just for security-matters anyway.. let's hope ASCII is nice and adds the cable to the 189euro-package. ^_^

Well, that was the point of my previous post - it's not making things any easier.

Nishi already spoke about standardizing certain OCM extenstions by acknowledging them as 'certified by MSX Assocation' (or something like that) during his lecture at MSX World 2005. The necessary VHDL to upgrade the OCM with the user-made VHDL extensions will become downloadable as well. However, as such VHDL can be written only after the OCM has been produced, details can only be thought about when the first VHDL appears.

Are you as a developer free to release/spread/propagandise extensions yourself, or do you *need* an MSXA certification? Because if this thing gets popular I could forsee that the 'scene' creates a better core than MSXA with better and way different extentions.. I dunno if MSXA likes that (for sake of conserving the MSX-'standard') ^_^

Of course you don't -need- MSXA Certification. You don't need certification for releasing MSX software or hardware either, do you?

Quote:

Of course you don't -need- MSXA Certification. You don't need certification for releasing MSX software or hardware either, do you?

That's right, but what about the fact that the extensions made by people need to be applied to MSXA VHDL code ? Do we have the right to modify and distribute the full VHDL code made by MSXA with our modifications or just a patch to the original code? That's why is so important to know which conditions apply to the VHDL code that comes with OCM...



btw, nice FAQ page. :-)

forgive me if I'm asking a stupid question, but I read somewhere that the FPGA runs at 21 mhz. You're saying that is possible to code a 100 mhz z80 with it. Can you explain that to me, I believe I need to learn more about processors in order to be able to see the advantages of OCM. :-)

Well, the R800 runs on 7,14mhz, but it can be compared to a 28mhz z80....Probably, similar mathematics can be done to create a 100mhz z80 with a 21mhz fpga?

I doubt you can create a 100MHz Z80 on a 21MHz FPGA. I don't think we need a 100MHz Z80 though, we need a 100MHz VDP!

On original Z80 few instructions executes in one cycle , actually a lot of them executes in 4 and more.
So if you make a Z80 with one instruction per cycle , its runs 4x faster at the same frequency.
R800 processors does this for a lot of Z80 instructions.