Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!henry From: henry@utzoo.UUCP (Henry Spencer) Newsgroups: can.general Subject: Re: A Canadian Unix Network? Message-ID: <8380@utzoo.UUCP> Date: Thu, 6-Aug-87 14:00:16 EDT Article-I.D.: utzoo.8380 Posted: Thu Aug 6 14:00:16 1987 Date-Received: Thu, 6-Aug-87 14:00:16 EDT References: <1987Aug4.235121.22896@gpu.utcs.toronto.edu> Organization: U of Toronto Zoology Lines: 54 > If it is possible to reduce the telecommunications component of > usenet, there would be no real reason to institute a special Canadian > edition of uunet... Actually, there is some reason for it, although for a slightly different reason. The key observation is that unless I'm missing something (this is not an area I'm expert in), reduced night rates for Datapac etc. benefit only those connections which have a Datapac hookup on at least one end. This is more common than it used to be, but remains far from universal. My impression is that it's not possible to use Datapac as a long-haul carrier between two phone-only sites -- and many, many Usenet sites do remain phone-only, since phones and modems are much easier to justify for mundane purposes than X.25 is. (There is actually a more global problem here in that Uunet services may not be easy to justify for mundane purposes either, but services and hardware often come out of separate budgets.) > ... The premise that Stargate was developed upon was that > there was this bandwidth available at relatively no cost that could be > piggybacked onto a national cable channel. The reality of the situa- > tion is that other institutions want that same bandwidth and are wil- > ling to pay much more money for it... Yup, quite true, as the Stargate people found when they started looking for a cable channel. A Canadian version of Stargate would be viable only if we could find what Stargate found: a channel (or at least vertical- interval) owner who was interested enough in the idea to donate the bandwidth at minimal cost. This possibility shouldn't be ruled out until/unless somebody actually looks around for such a benefactor, but I agree that pessimism is appropriate. > ... or Telecom Canada with their Megastream > offering that is a lot more expensive but also a lot faster than Anik. > For shorter portions of this network should terrestrial microwave be > used? Which protocols would be supported and how? As I've said before, clearly what we ought to have is coast-to-coast optical fiber, and never mind the silly microwave stuff. However, this would be rather a large initial step... Protocols should be left entirely to the customers except insofar as access to the medium demands adherence to standards. The last thing we need is a glorious wonderful network that is accessible only to those who have implemented a foot-thick ISO standard, because the network planners have decided to legislate compliance to the One True Way. The question should be not "Which protocols would be supported?" but "How can we give the customers the freedom to make this decision themselves?". Actually, right now the choice is obvious: almost everybody will want to talk either TCP/IP or uucico 'g' protocol, since they are an order of magnitude more widely available than any of the alternatives. -- Support sustained spaceflight: fight | Henry Spencer @ U of Toronto Zoology the soi-disant "Planetary Society"! | {allegra,ihnp4,decvax,utai}!utzoo!henry