wm@ogccse.ogc.edu (Wm Leler) (05/24/89)
There are several groups doing some sort of work with Linda on transputers. The Trolius people (in particular Moshe Branner at Cornell) have a version of Linda called Brenda, but it is very limited. At one time, Topologix had a version of Linda that ran on their hardware, but I'm not sure of the current status. And of course, here at Cogent Research we have an implementation of Linda called Kernel Linda that we have used as the interprocess communication mechanism for a distributed, UNIX-like operating system. Kernel Linda is a subset of Linda in that it only supports tuples with a single key (this was necessary in order to avoid the preprocessing step required of Yale Linda, which could not be done in an operating system). Kernel Linda also supports many advanced features, like multiple tuple spaces, language-independent communication, shared data, and so on. We plan on adapting the Yale proprocessor to use Kernel Linda as the Linda run-time system. If anyone would like more information, we would be glad to send it. Wm Leler Cogent Research, Inc. 1100 NW Compton Drive Beaverton, OR 97006-6998 503/690-1450
wm@ogccse.ogc.edu (Wm Leler) (06/02/89)
I'd like to add a couple of things to my earlier posting about Linda on transputers. Human Devices makes a system called Chorus that is transputer- based, and runs Linda. I should also point out that the "Yale Preprocessor" that I mentioned in my earlier posting was developed jointly by Yale and SCA, and is a product of SCA. In fact, the people at SCA are a good source of information about Linda, since they have versions that run on a number of systems. Finally, for people who are currently doing work with Linda, there is a mailing list (run out of Yale) that has discussions of various Linda implementations (among other things!). Long Live Linda, wm