[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Getting the Symbolics and the SUN to interact
Date: Fri, 22 Sep 89 13:37:17 -0700
From: email@example.com (Steve Willis)
So, we have developed some C pgms to perform the common image processng functons,
and now wish have the Symbolics invoke these functions on the Sun - typically
click on the sym, process on the Sun, view , interact and cogitate on the Sym, etc.
So how do we get the Sym and the Sun to interact?? There are a couple of
possibilities as we see it, but we are naive, and would appreciate more suggestions.
1) Use some sort of RPC call. Any ideas on how to do this would be helpful.
The recently announced and currently shipping Symbolics UX400S includes
software to do exactly this, which is fully documented and supported.
It's fully compatible with Sun's RPC, and you can use either Sun RPC or
Symbolics RPC to generate the Sun side of your application, whichever
you are more used to. The UX400S Ivory coprocessor can talk RPC to the
Sun it is embedded inside, and can also talk RPC to any other machine on
the network that speaks Sun RPC and TCP/IP. That RPC software will be
available for all Symbolics models in Genera 8.0, to ship in the first
calendar quarter of 1990. So, if you have a UX400S or you can wait a
few months, that's your solution.
We developed this because we knew there were quite a number of customers
in situations similar to yours, wanting an easy-to-use way to divide an
application between a Symbolics machine and a Unix box, with minimal
network and Unix expertise required.
This RPC software will not run in Genera 7.2, so if your Symbolics
machines are 3600s you'll have to wait for Genera 8.0. If you have
ILA-NFS in Genera 7.2, there is a not very well supported RPC
implementation inside that, which you might be able to convince to help
you out. This is not the same RPC as in Genera 8.0, but it is a program
that knows how to talk to Suns.
Additionally, we are interested in a general distributed computing model
that would allow us to easily interface our imaging system with whatever
super duper number cruncher comes along. Any thoughts??
I don't have any solution to offer to that, but I too am interested in
hearing any thoughts anyone might have on what Symbolics could do in
this area to be most useful to you. (This is NOT a promise to develop
anything!) For example, would anyone like us to support Apollo's
Network Computing System model? Is there some other standard that is
better? Or is RPC enough of a solution that you don't need anything
- [no subject]
- From: Steve Willis <bruno%oz.geology.washington.edu@RELAY.CS.NET>