| [Return to Home Page]
Feedback on 'Freeform Futurology (2) - The $20 Computer'
I trust your $20 computer will be easier to use than a TV remote control because, as TV remote controls improve they are becoming more and more difficult to use and the buttons on them are proliferating at an amazing rate.
I like your $20 computer (A$20 or US$20, and what will the difference be then?), but see it more as a pocket calendar/planner/calculator/communication device than a computer in the sense I use one now (which is primarily as a word processor). And since I use neither a cell phone nor electronic pocket planner at this time, I rather doubt that I'll use one of your $20 computers either (even if it does turn out to be only US$10).
I reply (August 2000):
I suspect the A$20 computer will (in time) end up looking like those PADD things in "ST:TNG." Those of us who create with PCs will want an ergonomic alternative, others who use PCs mainly to surf the Net, look up reference materials & only occasionally make notes will be happy with a handheld device.
Regarding web and email on pocket devices, the Psion 5 organiser and Ericsson SH888 phone have been able to do that wirelessly since sometime in 1998 (and it may have been 1997 or 1996 for the Psion 3c). I'm sure other companies had similar facilities around the same dates. The WAP and WML techniques now promoted seem to me mostly a way for phone companies to make out like bandits charging for custom portals that don't access the internet, but do have some fast changing or "locality" based information. I don't see widespread "permanent connections" via wireless broadband. The bandwidth isn't there (except for really short range stuff in the home or office as a local area network). Wireless comms (except for analog phones) is heavily into MUX as a result. Optical fibre beats wireless every way on bandwidth (by many orders of magnitude). That is why fibre still gets installed.
Not sure how much speed or memory battery powered devices can deliver. batteries improve at most 5% a year, power consumption for CMOS goes up as the square of the clock speed. Power needs also drop as voltage goes down, but several chips seem near the likely noise levels now, so you eventually run out of tricks at that end. (There is a reason fast AMD chips wear a humongous heat sink ... and could kill a set of batteries in a few minutes). Approaches like the Crusoe chip may help, since they move a lot of hardware into software (not really emulation, if I've grasped it correctly).
I reply (October 2000):
Thank-you Lindsay - your technological briefing is fascinating.
I feel confident that the technology wizards (I refuse to insult them by calling them boffins!) will solve the wireless bandwidth and PC energy problems. (Possibly with some ingenious hybrid data-transfer technology & a new development in fuel cells.)
If people want something badly enough - they can usually work out a solution.
(The latest "Scientific American" has articles on wireless Internet technology - but I haven't had time to read the articles yet, and cannot comment on any of the issues raised.)
There are computers that are that small. Ben has one, it is called a Palm Pilot and can do some amazing things, especially considering its size. It is a lot more than $20 at the moment but like all new technology the price will come down in time.
The Palm Pilot can be used by writing on the screen or else you can attach a keyboard to it and type as you would for the normal PC. He would be lost without it now. In fact when the first one he had crashed he went through withdrawal symptoms for a while until he managed to get it fixed. Now he has two and if one does crash at least he has the other as a bit of a back-up.
I'm sure we'll keep in mind your predictions for "ubiqitous distributed computing."
Copyright © 2003 by Michael F. Green and others. All rights reserved.
Last Updated: 26 April 2004