lauren@vortex.UUCP (Lauren Weinstein) (12/07/84)
I am pleased to announce that I successfully installed site
"stargate" on Monday, December 3rd, 1984. The stargate is a
Fortune 32:16 system, with 30 megabyte disk, which Fortune Systems
has graciously made available for the purposes of satellite
netnews delivery experimentation. Primary expenses for the experiment
are being covered by Usenix, with special thanks due for Lou Katz,
who has strongly supported my efforts in this area through
thick and thin. Thanks are also in line for Bell Communications
Research, which has provided both moral and tangible support
for the experiemnt. BCR's Brian Redman, Mike Lesk, and Stu Feldman
deserve special thanks.
The stargate is located at the satellite uplink facilities of
Southern Satellite Systems (SSS) in a rural area about 20 miles
outside of Atlanta. The stargate data transmissions are appearing
as a portion of the vertical interval on "Superstation" WTBS, which
is uplinked to the satellites via SSS facilities. SSS has
graciously made a continuous 1200 bps channel available for the
experiment.
WTBS is available to over 31 million cable TV subscribers across
the country via over 8000 cable companies. With the appropriate
decoders, the netnews transmissions will be available at any
point where satellite-delivered WTBS is received, either by cable TV or
direct satellite pickup (through inexpensive earth stations) when
direct pickup is not available.
We are currently in a demo/test mode--the exact shape of a production
system is still subject to change. Usenix's support of the experiment
does not in any way obligate them to supporting a production system,
though there are all sorts of favorable possibilities. A production
system would require a backup computer and would most likely operate
at a data rate ranging from 1200 through 9600 bps depending on
hour to hour load factors (that is, we'd probably be running at higher
speeds during the evening/night). Since all points receive WTBS
simultaneously, this is a true broadcast medium which makes much
greater use of even 1200 bps transmissions. 1200 bps reaching
the entire country at once is a considerable improvement over
point-to-point phone calls or data links!
Please don't ask me how you can sign up for this service now.
Decoder availability, costs, and various other factors are still
being worked out. As a rule of thumb, the basic decoder package
will probably cost about $500, with another $150 or so for our
special processing/buffering board. The possibility may exist
for "buy back" and rental of decoders, though this is not a certainty
at this time. A small monthly fee will most likely be assessed
for netnews delivery in a production environment, but note that
any such fee would be miniscule compared with the phone bills that
most sites now pay, or are likely to soon pay, for large portions
of netnews. Please remember, however, that for now this is
an experiment.
Transmission of some netnews materials will commence shortly,
via the "netnews screeners" who have volunteered after a previous
message I sent to this group. It is not contemplated that ALL
netnews would EVER be sent via satellite, due to technical and
broadcast restrictions, but a large (and growing) chunk of netnews
is suitable right now, and this is the portion of netnews that can
be expected to see the highest growth in the near future. Submission
of materials to the stargate will be via UUCP dialup (using a special
encryption technique to avoid "unauthorized" submissions) and control
is via a separate dialup line from vortex (here in Los Angeles).
The current demo system, at 1200 bps, will be relatively simple
since we do not yet have the specialized hardware to do local
buffering and flow control/error correction for high speed data.
Flow control will be handled through uplink data delays of fixed
length for now. It is important to understand that this is
only for the current experiment and would be handled differently,
via a special board attached to the decoder, in a production
environment.
Testing of netnews transmission reception is taking at my location
and will also take place (via a decoder that I'll take to various
sites) at (most likely) Lucasfilm, NASA-Ames, and various other
locations. If all goes well, a demo of the system should occur
at the Dallas Usenix Conference in January, via a 9-foot dish
portable satellite receiver to be brought in for the occasion.
I'll also be giving a progress report on the experiment, and
should hopefully be able to show a nice collection of slides
that I took while I was out at the satellite facility earlier
this week.
The first message I sent over the satellite link (sent for about
24 hours, in fact, as a test run) was, appropriately enough:
hello, universe
Within the next few days, more elaborate testing will commence
and distribution of actual netnews materials will begin
as soon as possible.
Once again, I'd like to thank everyone who has supported my efforts
to make this idea a reality. Usenix, Fortune, BCR, and SSS have
enabled us to potentially take a giant step forward toward improved
communications, via one of the most widely-distributed basic
cable broadcast services in existence. About six months ago, when
I first proposed broadcasting netnews via satellite, I never
imagined that so much would happen so fast.
I'll be happy to try answer questions from any of you about the
project, but please note that I cannot be too specific about
aspects of any possible production system at this time. Thanks again!
--Lauren--
{ihnp4, decvax, seismo, clyde, allegra}!vortex!laurendavid@bragvax.UUCP (David DiGiacomo) (12/11/84)
In article <463@vortex.UUCP> lauren@vortex.UUCP (Lauren Weinstein) writes: >As a rule of thumb, the basic decoder package >will probably cost about $500, with another $150 or so for our >special processing/buffering board. This is totally bogus! There is no reason why the decoder should cost more than $100 (video input) or $150 (RF input). If Lauren is turning a profit on this, why is Usenix funding the project? >The current demo system, at 1200 bps, will be relatively simple >since we do not yet have the specialized hardware to do local >buffering and flow control/error correction for high speed data. Yeah, specialized hardware -- an SBC with some RAM on it. This project seems to be about as well organized as the "Usenet project". -- David DiGiacomo, BRAG Systems Inc., San Mateo CA (415) 342-3963 (...decvax!ucbvax!hplabs!bragvax!david)
jack@vu44.UUCP (Jack Jansen) (12/12/84)
David Di Giacomo (sorry if I misspelled the name) writes about the $500 Lauren wants for a netnews decoder : > This is totally bogus! There is no reason why the decoder should cost > more than $100 (video input) or $150 (RF input). If Lauren is turning a > profit on this, why is Usenix funding the project? > > ... > > Yeah, specialized hardware -- an SBC with some RAM on it. *FLAME ON* I think this is totally unfair! Without knowing *any* details, you start accusing Lauren of trying to make a profit out of this project. I have *no idea whatsoever* how the scheme is implemented, so I have no idea what it will cost, but I know that your "SBC and some RAM" approach is doomed to fail. Ever tried to grab a few bits that race past at 5Mhz with a 8080? I think that it would be nice if you first asked for some details before you start shouting. By the way, the data can be found on your own TV set, so why don't you make your own decoder, and sell it at $150,- if you think that $500,- is too expensive????? -- Jack Jansen, {seismo|philabs|decvax}!mcvax!vu44!jack or ...!vu44!htsa!jack
henry@utzoo.UUCP (Henry Spencer) (12/12/84)
In Lauren's defense... > This is totally bogus! There is no reason why the decoder should cost > more than $100 (video input) or $150 (RF input). Lauren is a pro in this area; I'm afraid I put rather more faith in his cost estimates than yours. > ... If Lauren is turning a > profit on this, why is Usenix funding the project? Maybe because they think it's desirable, even if some awful capitalist swine does happen to be making a profit on it. Or even Lauren :-). Seriously, David, we're talking about a substantial production volume for this widget -- even 10 copies counts as "production" when it comes to the practical end of building them, and there'll be more than that -- and such things work out much better if people who work hard to make them go are allowed some compensation. > Yeah, specialized hardware -- an SBC with some RAM on it. Plus a wee bit of software imbedded in that hardware... "Software should be called hardware, and hardware should be called easyware." > This project seems to be about as well organized as the "Usenet > project". And what have you done for the network lately? -- Henry Spencer @ U of Toronto Zoology {allegra,ihnp4,linus,decvax}!utzoo!henry
david@bragvax.UUCP (David DiGiacomo) (12/15/84)
In article <528@vu44.UUCP> jack@vu44.UUCP (Jack Jansen) writes: >David Di Giacomo (sorry if I misspelled the name) writes about >the $500 Lauren wants for a netnews decoder : >> Yeah, specialized hardware -- an SBC with some RAM on it. >... how the scheme is implemented, so I have no idea what it will cost, >but I know that your "SBC and some RAM" approach is doomed to fail. >Ever tried to grab a few bits that race past at 5Mhz with a 8080? Lauren stated that the specialized hardware was necessary for upgrading the system from 1200 baud to 9600 baud on the decoder output. This has nothing to do with grabbing bits at 5 MHz (1 MHz is more likely anyway). >By the way, the data can be found on your own TV set, so why don't >you make your own decoder, and sell it at $150,- if you think >that $500,- is too expensive????? That's exactly what I'll do if Lauren will provide the encoding format. Will he sell me the decryption algorithm/key if I don't buy his decoder? -- David DiGiacomo, BRAG Systems Inc., San Mateo CA (415) 342-3963 (...decvax!ucbvax!hplabs!bragvax!david)