You've hit upon the two main means for content distribution that we've thought of. A hub-and-spoke model for distributing data from a VSAT terminal works as long as the intervening terrain is reasonably flat (and it can give a 2-way link to the 'net)
The terrestrial radio option is a good one. We've considered using subcarriers on existing shortwave or FM broadcast systems to provide a continuous stream of (for example) news and weather information. But the backchannel to the 'net is the big problem...
Mike, this is exactly what we had in mind- except that modern surface-mount parts require a lot more soldering skills (and a stereomicroscope) than the DIP packages used in the "old days".
The Pengachu team grew up on the Vic20 and the C64 and we miss the days when the hardware was as hackable as the software. One of the main uses we envision for such a system is as a low-cost tool for computer science education- because one person can completely understand the whole system, both hardware and software.
It is in fact possible to do this for right around the $50 OEM cost target. The typical markup is 3x-4x rather than the 10x you specify. At the moment RAM and Flash are both available at about $1/MByte on the commodity market.
In the endgame we want to develop our own ASIC that will integrate many of the peripheral functions on to the same die as the processor.
A couple of comments and clarifications from a member of the Pengachu team:
1. The $50 price tag is OEM cost in bulk. Our goal for this project is to build these devices and give them away to people in the developing world as information access devices. We're therefore not including the 300-400% markup that a commercial product would experience. If you consider the bill of materials cost for a Handspring Visor you will find that something around $30-$40 OEM cost in volume translates to the $110-$120 retail price.
2. The LCD is the single most costly component of the device (about $10 of the $50 cost target). While it would be nice to have a bigger display, you pay dearly for it. So it makes more sense to figure out how to build a decent UI model for limited screen real estate than throw all your budget into the display, which will cost more, break more easily, and eat up more power. This is especially important if you charge your batteries from a solar or wind-up power source!
3. This made it to Slashdot before we finished a proper documentation set. We wish to acknowledge the uClinux/Lineo team and TomW (http://www.openhardware.net) for doing the groundwork that made it possible to build these devices. TomW's commitment to open source hardware is especially laudable. Our device is not based on their netlists or board layouts, but their work with other uClinux hardware made it much easier for us.
-- Matt Reynolds, matt@media.mit.edu (hardware engineer for Pengachu)
Slashdot Mirror
The terrestrial radio option is a good one. We've considered using subcarriers on existing shortwave or FM broadcast systems to provide a continuous stream of (for example) news and weather information. But the backchannel to the 'net is the big problem...
-- Matt Reynolds (Pengachu hardware engineer)
The Pengachu team grew up on the Vic20 and the C64 and we miss the days when the hardware was as hackable as the software. One of the main uses we envision for such a system is as a low-cost tool for computer science education- because one person can completely understand the whole system, both hardware and software.
-- Matt Reynolds
In the endgame we want to develop our own ASIC that will integrate many of the peripheral functions on to the same die as the processor.
-- Matt Reynolds (Pengachu hardware engineer)
1. The $50 price tag is OEM cost in bulk. Our goal for this project is to build these devices and give them away to people in the developing world as information access devices. We're therefore not including the 300-400% markup that a commercial product would experience. If you consider the bill of materials cost for a Handspring Visor you will find that something around $30-$40 OEM cost in volume translates to the $110-$120 retail price.
2. The LCD is the single most costly component of the device (about $10 of the $50 cost target). While it would be nice to have a bigger display, you pay dearly for it. So it makes more sense to figure out how to build a decent UI model for limited screen real estate than throw all your budget into the display, which will cost more, break more easily, and eat up more power. This is especially important if you charge your batteries from a solar or wind-up power source!
3. This made it to Slashdot before we finished a proper documentation set. We wish to acknowledge the uClinux/Lineo team and TomW (http://www.openhardware.net) for doing the groundwork that made it possible to build these devices. TomW's commitment to open source hardware is especially laudable. Our device is not based on their netlists or board layouts, but their work with other uClinux hardware made it much easier for us.
-- Matt Reynolds, matt@media.mit.edu (hardware engineer for Pengachu)