Wow! That's what I have to say after see some movies on Youtube of a MSX2 running SymbOS. This is truly amazing.

I not a die hard MSX fan (though I enjoyed it 20 years ago), but this system was brought to my memory after read about recent difficulties of OLPT (One Laptop Per Child) project, specially due high costs to run Linux (memory, solid state disk etc) and the new (?) idea of a couple MIT researchers to recreate OLPT project based on 8bits computers after a trip to India where they saw a NES-like home computer been sold for just US$12.00 (twelve dollars!) in some city streets.

I know that would be extremely hard to sell a laptop for US$12.00. Even the Indian machine does not have a screen, just a standard TV output connection, but I wonder what would be the final cost to have SymbOS running on some MSX2 with a LCD screen and a wifi connection (this is what I call MSX3). Final specs would be something like:

1MB RAM
4 GB (SD card)
Wi-Fi connection
SymbOS software
LCD screen (MSX2 high resolution mode) (maybe some kind of e-Paper)
Zilog eZ80 CPU 20MHZ with TCP/IP support
PS2-style Keyboard
1 USB port
2 AA battery powered!

We are talking about several millions of computers to be sell to third-world governemnts. Such scale could make parts costs very competitive, huh?

Anyways, I think MSX2 standard is a much better candidate than NES-like computers to create an educational computer. MSX software library is still awesome (compilers, educational soft, etc) and over the years amazing software has been developed as well (browsers, MP3, MPeg players, FTP etc).

Links:
http://www.symbos.de/
http://www.insidetech.com/news/2791-mit-students-build-12-pc-based-on-nes
http://revolv.in/2008/03/15-computer-in-india.html
http://www.zilog.com/products/family.asp?fam=218

Come people, lets see this as an opportunity to relive a legend! Post your comments and ideas. Maybe someone (Nishi, where are you man? This is the chance that you may waiting for) could create a prototype to show to some government...

Regards, Emerson

It's probably not quite as simple. Cost is not so much determined by the age or low complexity of a system, but by simple production volume. The OLPC is quite a lot more powerful than an MSX2. But since it should be able to access the internet, it really needed that power. With the current state of websites, you're just not going to get away with a 20 MHz z80 and a TV resolution display.

Also, SymbOS is a very nice system, but it's not complete yet for daily usage. It's going to need a lot more applications. Which brings you to the main issue, what should a machine be able to do to get potential market of many millions. After you figure that out, you can determine whether an MSX2 (or NES for that matter) based PC has any chance of succeeding. Personally, I don't see anything like that happening without proper internet access. Which pretty much discounts either option.

elopes, Nishi has been involved in the OLPC project. He probably has brought up the MSX concept as some kind of basis. Maybe it's not even a coincidence you connected the OLPC with your MSX memories. Still, I don't see it happening anymore the MSX concept will somehow lead to a new OLPC-computer. We've passed that phase and it didn't happen.

Hello,

You are right about connectivity, that's why I've put eZ80 in the specs. Take a look in its specs. It has built-in TCP/IP capabilities (including web server with cgi processing).

Regards, Emerson

PS: I mean OLPC, not OLPT (it was a kind of joke with brazilian Workers Party (PT in portuguese), the Brazil's president party who is interessed in OLPC project).

BTW, MSX disappeared because its market disappeared in first place. We are not talking about to sell computers to people that can spend US$200 or US$ 2000 for a computer. We are talking about to sell millions of machines directly to governments so they could give it for free to students whom does not have enough money to buy a notebook. A window-like computer for US$25 would make a difference (forget about US$100 or US$200, it will not happen).

OLPC so far is a massive failure. besides that, it does not seem inspired by projects like MSX (which stablished a formal standard, so anyone could produce a compatible machine). It is a one-stop shopper approuch (they make the specs, they produce the machine and they sell it, no room for innovation - again an outstanding point to MSX). I think the idea here would be to create the specs. Hopefully, it will attract people who want to make money by making a machine compliant with defined standard.

BTW: I'm not advocating the idea to create a computer to connect to a TV. I think it HAS to be mobile (from classroom to room at home).

MIT is moving in this direction. The point is that the machine they selected as a model does not have anything like SymbOS. MSX does!

Just because MIT is experimenting with it, doesn't mean there is a real market. Especially with internet connectivity it's probably not going to work. It's not because of the connectivity itself, it's about rendering web pages. An average webpage already contains more than 1MB of data. After decompressing images that will increase to several MB. Ploughing through that requires a bit more than an 8 bit cpu with 64k of address space.

Sorry, I cannot agree with you simply because such machine would not be used by people like you and me. This is to be used by third world students, to allow them to access email, forums, FTP servers, email and tailor made web content such as education oriented websites (there are thousands of lightweight websites like that). Maybe wikipedia, but not youtube (though I believe SymbOS could handle such content).

Another point is that eZ80 can address up to 16MB of RAM, so machine specs could be easily updated to something like:

8MB RAM
4 GB (SD card)
Wi-Fi connection
SymbOS software
LCD screen (MSX2 high resolution mode) (maybe some kind of e-Paper)
Zilog eZ80 CPU 20MHZ with TCP/IP support
1 USB port
2 AA battery powered!

Cheers