There is no single evidence saying that V9978 was intended to have V9958 compatibility, only rumours.
Instead the decision of using V9990 for this new VDP is a clear indication that yamaha want to tell us "This is a vdp, but not a retro compatible thing, totally a new product, with break anything with V99x8 series".
Otherwise they should have named it V9978 .....
Nah, i do not agree. MSX is made of "standard" on the market components. They choosed what the market had to offer. Maybe the best if one would not afford a custom chip.
Ironically,they would have been forced to custom chips as many other vendors with later generations...
Interestingly, it's only the MSX1 that can be made of off-the-shelf components. The MSX2, released only two years later, already needed at least one custom part: the VDP. This made me think MSX wasn't actually meant to be always based on off-the-shelf components, but MSX1 eventually ended up that way by force of necessity. Instead, it would be the other way around: MSX chips would be available off the shelf for anyone to use in their designs (and Ascii would happily collect their royalties), MSX or otherwise. The large number of MSX-like unofficial hardware people like @PiterPunk helped unearth seems to support this (and of course we know Yamaha sold their MSX chips to anyone).
If this is true, and it seems to be in line with Nishi's ambition of an MSX in every corner, then I believe he was telling the truth when he said he intended MSX1 to have an updated VDP from the outset. But even if this new VDP existed at the time, in 1983 RAM was still too expensive, so it wouldn't have mattered much.
The situation in '80 was extremely clear: while there was a relatively large kind of general purpose chips for any usage (serial, parallel I/O, mmu, disk controllers, even math coprocessors and sound chips) there is nothing available for video interface. only the 6845, rather limited and not a true vdp, or vic-I maybe. then the TMS.
If TMS improved the 9918 quickly, most probably the msx could have adopted new TMS chips.
But what, a manufacturer can do if it want to improve the existing video hw but there is no improved hw from its first choice? (TMS). The only thing one can do is to develop a custom chip.
Sega did the same as msx. First gen: TMS Standard chip. Next Gen: Custom chip based on TMS VDP
The same happened to Amstrad CPC, that used a standard component. Later with CPC+ they were forced to custom chips.
There is no single evidence saying that V9978 was intended to have V9958 compatibility, only rumours.
Instead the decision of using V9990 for this new VDP is a clear indication that yamaha want to tell us "This is a vdp, but not a retro compatible thing, totally a new product, with break anything with V99x8 series".
Otherwise they should have named it V9978 .....
If it wasn't then ASCII would have quite some troubles manufacturing it, two VDPs, two VRAM banks, composing image from two different sources... Or, not backward compatible.... Doesn't make any sense at all, it would be a very bad project choice as it would increase already high costs and make it not competitive price wise with other computers of the same price range (with 2 VDPs, that is).
There is no single evidence saying that V9978 was intended to have V9958 compatibility, only rumours.
Instead the decision of using V9990 for this new VDP is a clear indication that yamaha want to tell us "This is a vdp, but not a retro compatible thing, totally a new product, with break anything with V99x8 series".
Otherwise they should have named it V9978 .....
What you say there does not make sense. And you forget that it was Panasonic who first said that the MSX3 could not be released because the VDP was not ready. It's not only Nishi. It's not just a rumour. And then, when Nishi lied, as for example for the significance of MSX, it was to generate buzz. This does not call into question the reliability of what is said.
If this is true, and it seems to be in line with Nishi's ambition of an MSX in every corner, then I believe he was telling the truth when he said he intended MSX1 to have an updated VDP from the outset.
No, the MSX1 were made with common components because Nishi wanted a simple to use machine so components that are known and available pretty much everywhere.
The same happened to Amstrad CPC, that used a standard component. Later with CPC+ they were forced to custom chips.
Custom chips were very expensive back then. When the LSI, VLSI and other circuits arrived, it made it possible to make them for a reasonable price.
There is no single evidence saying that V9978 was intended to have V9958 compatibility, only rumours.
Instead the decision of using V9990 for this new VDP is a clear indication that yamaha want to tell us "This is a vdp, but not a retro compatible thing, totally a new product, with break anything with V99x8 series".
Otherwise they should have named it V9978 .....
What you say there does not make sense. And you forget that it was Panasonic who first said that the MSX3 could not be released because the VDP was not ready. It's not only Nishi. It's not just a rumour.
Obviusly panasonic said this. But who was in charge of following the development of V9990 ? Surely not panasonic.
Then you say it's not a rumour about v9978. I've never seen a true evidence that a similar v9978 chip with retro compatibility was AT LEAST took in account from yamaha .
We can say that is nishi, panasonic or santa claus that wanted the V9978 but those are only words and speculations. There is no real evidence.
If this is true, and it seems to be in line with Nishi's ambition of an MSX in every corner, then I believe he was telling the truth when he said he intended MSX1 to have an updated VDP from the outset.
No, the MSX1 were made with common components because Nishi wanted a simple to use machine so components that are known and available pretty much everywhere.
No, this is a weak justification, nishi wanted only to cut down costs needed to develop custom hw. In the '80 there where a lot of different and relatively 'not simple to use machines' but this did not stop at all the availability of good software for such machines.
The same happened to Amstrad CPC, that used a standard component. Later with CPC+ they were forced to custom chips.
Custom chips were very expensive back then. When the LSI, VLSI and other circuits arrived, it made it possible to make them for a reasonable price.
VLSI was already on place when MSX1 was developed.
Others customers afforded the costs of making custom chips without too much trouble.
Then you say it's not a rumour about v9978. I've never seen a true evidence that a similar v9978 chip with retro compatibility was AT LEAST took in account from yamaha .
We can say that is nishi, panasonic or santa claus that wanted the V9978 but those are only words and speculations. There is no real evidence.
There is a one page preliminary release datasheet for the V9978. Very similar to the V9990 one that came later. So Yamaha started to work on something. It may not be more than an intention. Often those one pagers come before the design.
The rumors that the compatibility was removed after the project could not make it into an MSX is harder to verify. On one side it could explain the way V9990 is. What makes the rumor compelling. On the other side, once the features are in, removing them is a lot of hard work. What makes me think that the rumor side that Yamaha launched the chip anyway after missing the boat is very plausible. But the part that they would sink more money into removing the compatibility features ir harder to believe.
Hard to say. But maybe part of the plan was to have both V9958 and V9990 side by side.
Then you say it's not a rumour about v9978. I've never seen a true evidence that a similar v9978 chip with retro compatibility was AT LEAST took in account from yamaha .
We can say that is nishi, panasonic or santa claus that wanted the V9978 but those are only words and speculations. There is no real evidence.
There is a one page preliminary release datasheet for the V9978. Very similar to the V9990 one that came later. So Yamaha started to work on something. It may not be more than an intention. Often those one pagers come before the design.
Surely!. Yamaha started to work on V9978/90. No doubt about this. What i say they intented the v9978 as a chip to be side by side with 9958. This is proved by the docs on v9978. they do not mention at all the legacy compatibility.
This is because they not even planned it.
Hard to say. But maybe part of the plan was to have both V9958 and V9990 side by side.
this is probably the most logic solution they had in mind. But i suspect that nishi wanted an all-in-one solution maybe for cutting cost. Obviously this yamaha didn't liked it. And boom.... the agreement broke. then yamaha changed the 9978 into 9990.
But the part that they would sink more money into removing the compatibility features ir harder to believe.
What i say they intented the v9978 as a chip to be side by side with 9958. This is proved by the docs on v9978. they do not mention at all the legacy compatibility.
Maybe the compatibility was just at the pinout level. Maybe there was the possibility to link them together easily. Also, v9990 needs more I/O ports so for the system, if v9978 was to come close, they had to be separated from those of v9958 (which is easier) or find a way to separate them into two parts.
I don't understand the idea why one theory should be in need of evidence while the other one isn't. Since there is no proof that the V9978 is not backwards-compatible with the V9958 either, neither theory is anything but speculation.
Do we even know that the V9990 isn't hiding unused/disabled V9958 circuitry? It hasn't been decapped yet IIRC.
