Sean Young wrote:
>On Sat, Apr 21, 2001 at 11:55:03PM +0200, Patriek Lesparre wrote:
> > MSX is more than the MSX hardware. MSX is more than the MSX software. MSX
> > is also a philosophy.
>
>What do you mean exactly?
In the meeting on Sunday 22nd the 'MSX philosophy' was mentioned quite a
few times.
It's about education, ease of use, flexibility etc.
There are problems though, because what Nishi told us about the FPGA/PLD
design that would cost 10$-20$ was not quite accurate. With CURRENT
technology such a design would cost about $200 per chip, while an ASIC
design costs 10$ but has to be manufactured at least 10 million times.
Who knows what the technology will be like in 2 years. PLD technology has
improved fenominally (both in capabilities and price) in the last 2 years
so if the trend continues, Nishi's words may be right. Also, Nishi is a
Professor at MIT University where he will work on a project that should
make PLD chips far more cheaper than today.
He really has everything planned out, and the timing seems perfect.
Ofcourse people today are laughing at Nishi. But history proves everytime
people laugh at Nishi for a crazy idea, he eventually succeeds!
Because the one-chip MSX will be used in 3rd world countries and such, it
will have to be very cheap, because the governments must buy it. Normal
people in those countries are too poor. So it's possible 2 versions will be
made. One cheap ASIC design and one more expensive PLD design.
Software from the ASIC design can run on the PLD, but if you create
software that reprograms the PLD it ofcourse can't run on the ASIC.
Basically, because the an ASIC is fixed, the ASIC one-chip-MSX is actually
nothing more than a very cheap one-chip-PC with hardware MSX emulation.
But the PLD version is far closer to the MSX philosophy, and in my opinion
worthy to carry the name 'MSX'. The Japanese feel the same way.
The PLD design allows one to program their OWN HARDWARE! Ofcourse not
everybody has to be able to do that. For instance, if one person writes a
MoonSound hardware, you can download it from the MSX Server (for a certain
amount of money) and *BAM* you have a MoonSound in your one-chip-MSX.
It's also really good for emulation, for instance you can create the
Gameboy Advance (soon to be released in Europe/USA) VDP and Sound hardware
in PLD and use the ARM9 core to emulate the GBA CPU (which is a ARM7 by the
way ^_^).
Really the possibilities are infinite. Because of this, the PLD design can
last much longer than an ASIC version. An ASIC version would be obsolete
sooner because it's totally fixed. If you want new hardware functions
(maybe a 3D engine) you have to create it in software, using CPU power. But
with the PLD design you can program the 3D engine in hardware so the CPU
can make full use of its power for other purposes.
Now, most people want a new PC every 3 years. With an ASIC one-chip-MSX
that may also be the case, but a PLD version last longer because of its
flexibility.
> > I'm really convinced this one-chip MSX using an ARM9 core is the best
> thing ever. It's truly limitless in possibilities!
>
>Yes, but since you got an ARM9 and FPGA -- why write software for the
>"old" system? It would be a waste of cycles to begin with. The
>possibilities are endless with a fast CPU and FPGA, so you'll write stuff
>for that.
First of all, don't forger the MSX Player (the ASCII emulator). If a
succes, it will run on millions of cellphones/PDA's/laptops/desktops/etc.
so if you create software for the old architecture you have an instant
market of millions. While if you program the one-chip-MSX only those with a
one-chip-MSX will be able to buy your software.
The MSX Player improves transition between old and new architecture in the
way I described above and also people will start to remember MSX again and
will be more likely to buy a one-chip-MSX when it's out.
Nasu (Yokoi) told about Sharp and their Zaurus PDA. They developed 2 games
for it, 1st was Space Invaders and 2nd Pacman. The games sold for $20 and
they were sold out within 3 days (do I remember this correctly, please
someone who was present Sunday correct me on this). So the market for these
games is huge, even though you can download such games on the internet for
free.
When Nasu told Sharp about the MSX software library, which is between
4000-7000 they totally flipped out! ^_^
Also MSX games are very small. You can send an MSX game in notime at
9600bps which is the current cellphone standard. Also on the new MSX design
was a SD-card slot. It's a very small memory card by Panasonic ranging from
16MB to 1GB. Nasu met with Panasonic people, and when he told them how much
MSX games could fit on the smallest SD card (16MBytes) the Panasonic person
went on his knees and said "HELP ME!" ^_^
There seems to be very good support by the big companies. Sharp will be
happy because they'll have lots of games for their PDA's. So the PDA will
sell more. Then Panasonic is happy because the PDA's use the SD memory
cards so they will sell more. And Toshiba is happy because they make the
memorychips inside the SD cards. And when they are used in new MSX, they
will sell more and more and more. The companies will even PAY you for
including a such a memorycard slot in your design, which helps keep
development costs low.
In the meantime, everybody is using MSX! ^_^
on a less serious note: world domination ;)
>The point is, it's not MSX anymore. It's a completely new system with the
>name MSX.
Sure, you are right it's not the MSX hardware, but it's everything else!!
It's the MSX software, it's the MSX philosophy, it's the MSX feeling!
Does it really matter if there's not a REAL V9938 inside but an FPGA one??
FPGA is hardware too, but it's programmable. The line between software and
hardware is fading and this new MSX will contribute greatly to that.
>There isn't much market for the "old architecture"
You forget the MSX Player. There will be millions of MSX users!
Well this turned out to be quite lengthy and not only a reply to Sean, but
more informative for everybody (I hope ^^;).
Greetz,
Patriek
--
For info, see http://www.stack.nl/~wynke/MSX/listinfo.html