lauren@vortex.UUCP (Lauren Weinstein) (12/17/84)
I've gotten a lot of messages from people asking how they can help with the project. I can mention one place right now where detailed technical suggestions can be of use--the design issues of the message buffering board. While the demod/decoder (the pair that will hopefully be available as a set for $500) have to meet all sorts of protocol, video, and cable standards, we are under no such constraints for the add-on buffering board that would sit between the decoder and the "mainframe" computers receiving articles. This board will need to take serial ASCII, at speeds varying (probably) between 1200 and 9600 bps over time, extract and store as many messages as possible (as selected by mainframe commands regarding catagories, sequence numbers, etc.) and make them available to the mainframe in a solid, flow-controlled manner. This board would contain as much memory as reasonably affordable, and the more sophistication that can be built onto the board and its firmware, the better. This board will also be responsible for error correction (through merging of multiple message transmissions), keeping track of received and needed message sequence numbers, and a host of other functions that are best handled outside people's mainframe systems. The practical limit on memory for the board will probably relate to money. As I've mentioned before, I'd sure like to see this board cost no more $150 to $200 (including memory, manufacturing and distribution costs, etc.) if possible. Of course, ideally, it would also be subject to the same sort of buyback and rental arrangements that (I'm hoping) would eventually be available for the demod/decoder. The firmware for this board could get quite complicated quite quickly, and I KNOW that I will NOT have the time (nor the inclination, to be frank) to do that coding. The satellite carrier may be able to contribute to the board's design and manufacturing, since they could apparently get considerable use of their own out of such a device, particularly if it were fully addressible and had some TV display capabilities of its own. Their participation may well serve to help keep the price down since it might mean higher volume demand for the equipment. In any case, I will probably be talking to them about the board issues in early January, and I'd appreciate specific technical suggestions (memory, CPU, etc.) that people out there could make that I could bring to the table for those discussions. Please only tell me about components you've worked with yourself rather than relay second-hand info if possible. Your suggestions regarding memory and related components for the buffering board could be of great use if the project is to proceed beyond the experimental stage. By the way, another place where assistance will be needed down the line will be in the design of the mainframe software to interface the board (and thus the demod/decoder) to the Usenet software running on the mainframes. This may well be a substantial task in itself, since it will involve sending out specialized uucp messages to the stargate for message retransmissions in some cases and involves a number of other unique aspects as well. Please don't volunteer for the mainframe software effort yet--it's still too early for that... All replies regarding the buffering board hardware design should be sent to me as uucp mail, not posted to this group. Thanks again. --Lauren--