ram@shukra.UUCP (09/07/87)
Hi, It's been a looooong time since I posted about starting a newsgroup for parallel architectures. Since then a lot of background discussions had been going on with Steve (comp.hypercube) to use comp.hypercube until traffic builds up. Since ATT is already making sufficent money, we decided to use comp.hypercube until ATT starts loosing money (:-)). OK. So, all parallel architecture enthusiasts - lets do some congestion testing on hypercube :-). As Steve has already pointed out in comp.hypercube - anything with parallelism (architectures, compilers, languages, trivia) will be accepted. Note: There are two other groups that would be of interest -one a transputer based mailing list (originating from cornell) and a dormant supercomputer mailing list exists (I think supercomputer@cmcl2) for the Internet folks. --------------------- Renu Raman ARPA:ram@sun.com Sun Microsystems UUCP:{ucbvax,seismo,hplabs}!sun!ram M/S 5-40, 2500 Garcia Avenue, Mt. View, CA 94043
jclaude@ecrcvax.UUCP (Jean Claude Syre) (09/08/87)
Summary: Expires: Sender: Followup-To: Distribution: WHAT??????????????????????????? I will never post anything in hypercube. until my work becomes related to hypercube. It is as if you asked TOYOTA people to make their ads using the word MITSUBISHI only. Please, the net is already a mess, it's so difficult to read anything interesting. When I look at the hundreds of rec.X, comp.Y, net.Z, (Prolog hackers, instantiate X Y and Z, others just give values), which contain just garbage, I believe that a newsgroup entitled parallel-arch could live proudly among that crowd (we and other people I know are already exchanging data and info through e-mail). I even suggest to create a numpar-arch and a symbpar-arch, to make numerical and symbolic parallel processing distinct (there are many common issues, but even more which are specific; I would subscribe and participate to both). And to make things seriously, I suggest that they be moderated. And if you do not find anybody to moderate, we could do the job for symbolic parallelism. (* private message: Hi Renu, nice to read you *) J.C. Syre CA Group Leader ECRC Arabellastr. 17 D-8000 Munich 81 In ai one thing is certain: West Germany the future is not sure. tel(49) 89 92 699 127
fpst@hubcap.UUCP (Steve Stevenson) (09/08/87)
in article <429@ecrcvax.UUCP>, jclaude@ecrcvax.UUCP (Jean Claude Syre) says:
[....I'll never participate in comp.hypercube....]
I'm the moderator of comp.hypercube.
The combination comes from many considerations. When, and if, it
becomes necessary to propagate a new newsgroup we can do it. Consider
the following, however.
* At this time, the major interconnection available
to a very wide diversity of groups is the hypercube - the Connection
machine all the way down to the 2-cube.
* Problems which arise in parallelism are probably independent
of interconnection methodology - thereby one can profit from the
lessons learned on a particular connection scheme.
* Secondly, nobody seems to know everything. Thus, comp.hypercube
is a meeting place for engineers, (theoretical) computer scientists,
language people, etc.
I'm sorry you feel the necessity to not participate because of a name.
--
Steve Stevenson fpst@hubcap.clemson.edu
(aka D. E. Stevenson), fpst@clemson.csnet
Department of Computer Science, comp.hypercube
Clemson University, Clemson, SC 29634-1906 (803)656-5880.mabell