Hi all, My two pence while we are on the subject.
The professed goal of the Simputer is a low cost simple to use device.
Costs of a product depend on (briefly) a) design costs, b) dies, moulds and tools, c) raw material d) marketing. If volumes are large a and b reduce to ignorable values. Which leaves c and d. Lets ignore d as a matter of individual idosyncracies and company policy. The I386 arch scores over all others on rm costs. An 810 board with a Cryix cpu and flash instead of drives will give you much better performance at the same costs as a simputer. Morover an 810 can be easily upgradable is available from multiple vendors and has all the software you would ever require. If you want to cut it fine there are several implementations of Ix86 cpu with VGA and peripherals and some combo of i/o (PCI, ISA ) buses available as single chip devices. They are also power effecient (compared to Cryix). As Mayuresh pointed out the simputer is to be plugged into a phone line, so portability & power effeciency is a non issue. Finding a phone line in the interiors is anyway not a computing issue and best left to the DOT.
Handhelds were developed with mobility as the primary goal. It is a tool for presenting very very brief synopsis of very important information and presumes that you are in a position to use and act on this info. In India where even the most rudimentary infrastructure is lacking, I think Simputer pong would be a good application while waiting for the ST bus or the tehesildar.
A corolary of mobility is very low power consumption. Hence simple architecture (read as strip_off_everything_ useful). Which resulted in "simple to use" software (useless). The arm development is progressing mainly because there are requirements where low power un attended operations are mandatory - industrial electronics- and developers are fed up of porting applications every 2nd year when their microcontrollers go obsolete. And ofcourse all this makes sense only under the opensource development framework.
Ease of use, fancy Gui's, phone list etc are farthest away from the development requirements for this arch.
Development effort is required at the software level for the Ix86 arch cause gadgets like the simputer can never hope to achieve the maturity and spread of the PC.
Which brings up the question how will any Company successfully commercalise such a product? Government subsidy? As somone pointed out our village dwellers are a very intelligent lot. They will quickly see thru the charade being enacted in the garb of bringing IT to the masses.