I'm not keen on Raspberry Pi idea. If it is a FPGA + CPU (e.g. ARM) that we want, I prefer a SoC FPGA, like Altera Cyclone V SoC, I have a Terasic DE1 Soc and DE0 nano Soc, but haven't done anything with them... at least you get a better FPGA-CPU integration.
@PAC I think also original VDP and FM sound chip ICs are necessary, the FPGA emulations are not good enough and require a lot of time investment to get them up to the desired level.
For most MSX chips, FPGA implementation is better (cheaper / smaller / faster / upgradeable / ...) in almost every way. V9958/V9990 and some sound chips would be the exception. Note that for a sound chip, an FPGA board's analog section may also be to blame, vs. an MSX with the real IC + better sounding analog circuit.
Personally I think that as the original IC's get older, [improving FPGA implementation] may be an easier problem to solve than [obtain working specimens of the old IC's]. This is already happening with other systems, for example try finding a spare / New Old Stock ZX Spectrum ULA... FPGA route has the added advantage that you can fix shortcomings to IC's (like those rare logic bugs/glitches that original IC's may show). And -relatively- easy to port the design to new FPGA boards or experiment with new hardware. Much like with emulators.
What may be useful, is reserve some FPGA I/O pins for much-used functions in cartridges, and then leave it to (internal) plug-in modules or external modules to implement a specific flavor of that. One example: serial I/O. An FPGA could include an implementation of often-used 8251/8253 combo, and route the serial signals to a pin header or small external plug. Then you could plug in a cheap/simple/DIY-able module that uses those signals to provide either proper RS232, MIDI, infrared, or some other serial interface. And maybe even (ab)use those pins for other interfaces like I2C.
That FPGA's 8251/8253 combo would become an instant de-facto standard with the widest possible software support, the plug-in modules would be very cheap & simple.
Similar thing could be done for sound: say reserve a few FPGA pins to provide a serial, digital sound input. And then leave it to logic in the FPGA if/how those signals are mixed into output sound. That could do away with a lot of the volume / ratio issues people are having with SCC/FM vs. PSG etc.
Also nice: expanded cartridge slot(s). One that still allows plugging in regular MSX cartridges, but has some more pins for "advanced" cartridges. Kind of like how 32-bit vs. 64-bit PCI, or PCIe x1 vs. PCIe x16 is implemented. Maybe for a 16 or 32 bit data bus, DMA signals, 3.3V power, video-related or additional clock signal(s), etc, etc.
Though for the VDP, I understand it is difficult to provide HDMI output with the original IC. (How hard is it to make an XRGB-mini Framemeister in FPGA? )
First you'd have to capture the analog output of original IC with a 3-input video ADC (to allow processing by the FPGA). Then you'd have to figure out a good way to 'map' each screen mode produced by the original IC, onto a standard/supported HDMI mode. And then write the Verilog or VHDL code to implement it (possibly borrowing building blocks elsewhere). And have a big enough FPGA. Oh and hook up a HDMI connector... Each of those things is doable in theory. But combined, very much NOT easy I think...
Oh btw: Altera DE2 I have, has a composite video input + video ADC, and a sample application that converts the captured video data to VGA output. The IC's involved are not very expensive. So it's not like this is impossible. Just takes a lot of work.
Now a HDMI output is not something you simply wire up to an FPGA board. HDMI signals are very high-frequency and too timing-sensitive for that. So in general HDMI means you need an FPGA board that already includes HDMI output. Which... is easy to find.
I think that to implement anything and everything into FPGA, should been done on 1chipMSX.
I don't interest to make a copycat of failed project
What I'm aiming is between GR8BIT and 1chipMSX.
BTW, I’m wondering that nobody considers personnel expenses but only material cost.
I'm not keen on Raspberry Pi idea. If it is a FPGA + CPU (e.g. ARM) that we want, I prefer a SoC FPGA, like Altera Cyclone V SoC, I have a Terasic DE1 Soc and DE0 nano Soc, but haven't done anything with them... at least you get a better FPGA-CPU integration.
One idea behind FPGA + RasPI approach is that RasPI is usable as is, with very good software and hardware support to almost any use. You could use the same computer for web browsing, media player and everything, beside all the MSX stuff.
Other is modularity. The base model could be just the FPGA board. It would be up to the user if he wants more features and power.
There's only one person that could realistically make a NEW MSX machine and that is Eugeny Brychkov, as he's already made the GR8bit MSX kit. As the GR8bit may be too expensive for some hobbyist a Single Board , cut down version for about 250 Euros would be a good idea ? Perhaps someone should discuss the possibility of this with him as they could potentially be a sizeable market for it !
There's only one person that could realistically make a NEW MSX machine and that is Eugeny Brychkov, as he's already made the GR8bit MSX kit.
Rubbish. Sure, he could do it. But, certainly he is not the only one on the planet who could do it.
A little off-topic maybe, but I think important nonetheless.
Apart from the technicalities of the internals, I also believe a true MSX successor should at least look like it. So it should be a full product, not just some barebones development board.
Every MSX model and brand looked different, but every MSX also had a very specific and recognizable look. To really "feel" MSX-like, a new one should at least have a lot of that in it.
I think the most common look was the NMS8245 "keyboard+floppy in one" look. The sonys, panasonics, sanyo models all had that look. I think some sort of derivative should be created for a successor. Or maybe even one of the earlier MSX models without floppydrives - they were the most colorful as well.
Also: this discussion pops up every now and then on the forum Isn't it about time we put our heads together seriously and start thinking about a serious plan for a Kickstarter campaign?
If this is the case, why hasn't somebody at least started making a NEW hardware MSX. And I'm not talking about software emulation like running a MSX emulator on a Raspberry-pi. The ZX Vega uses this approach, which is dreadful !! Or using Hardware emulation (FPGA) like a Altera DE0 board programmed to act like an MSX ! I mean using Real chips like Z80,V9958 YM2149 etc !!
And I'm not talking about software emulation like running a MSX emulator on a Raspberry-pi. The ZX Vega uses this approach, which is dreadful !!
ZX Vega is dreadful mainly for other reasons than using emulation...
Or using Hardware emulation (FPGA) like a Altera DE0 board programmed to act like an MSX ! I mean using Real chips like Z80,V9958 YM2149 etc !!
FPGA implementation is NOT emulation, it's a real circuit! Take a discontinued chip, like the MSX-Engine S3527. You could re-create it using discrete logics. Or you could do it in VHDL. If you had a lot of money, you could have that VHDL manufactured in ASIC, a real chip! But burning that design to an FPGA works as well. For a user, there is absolutely no difference!