K312240@AEARN.BITNET (Klaus Kusche) (05/31/88)
This is the second (third, fourth, ...) try to send this message. I received no response or acknowledge up to now. I heard that there may be serious problems with the European BITNET backbone interconnections and BITNET routing tables. Please don't be angry if you receive this message for the n-th time! Dear Mailing List! This question is not very transputer-related, but I don't know a better mailing list to post it (if you know, please tell me): I'm looking for any kind of information about concepts and implementations of parallel Lisp (again, not only transputer-based!). Especially, I would be interested in some good papers on Multilisp. Moreover, I heard that there is a Common Lisp on the intel Hypercube -- sounds also very interesting, any informations welcome!!! Klaus Kusche (K312240@AEARN.BITNET).
ran@mit-vax.LCS.MIT.EDU (Randy Osborne) (05/31/88)
In article <8805301428.AA17144@tcgould.TN.CORNELL.EDU> K312240@AEARN.BITNET (Klaus Kusche) writes: >I'm looking for any kind of information about concepts >and implementations of parallel Lisp (again, not only >transputer-based!). > >Especially, I would be interested in some good papers on Multilisp. Papers on Multilisp: The original paper. Gives detailed description of Multilisp and its implementation. Halstead, R., "Multilisp: A Language for Concurrent Symbolic Computation", ACM Trans. on Prog. Languages and Systems, October 1985, p. 501-538 The following two papers are more tutorial in nature. Halstead, R., "Parallel Symbolic Computing," IEEE Computer 19:8, August 1986, p. 35-43 Halstead, R., "Parallel Computing Using Multilisp," J. Kowalik, ed., Parallel Computation and Computers for Artifical Intelligence, Kluwer Academic Publishers, 1987 Describes exception handling implementation: exception values, catch/throw, unwind-protect. Halstead, R., J. Loaiza, "Exception Handling in Multilisp", 1985 Int'l. Conf. on Parallel Processing, St. Charles, Ill., August 1985, p. 822-830 Discusses the three major ways to use futures and describes other "lessons from experience". Halstead, R., "An Assessment of Multilisp: Lessons From Experience", International Journal of Parallel Programming 15:6, Dec. 1986, Plenum Press, New York Multilisp is available from the Parallel Processing Group at M.I.T. The present implementation is a byte-code interpreter which runs on uniprocessor machines supporting UNIX (BSD 4.2) and the multiprocessor Encore Multimax (and our Concert Multiprocessor). Little work should be necessary to port the implementation to other shared memory multiprocessors, which like the Multimax, support UNIX. I'll post details on how to obtain this implementation of Multilisp if people express interest. -------------------------------------------------------------------------- Randy Osborne ran@vx.lcs.mit.edu
UHAC010@vaxa.rhbnc.ac.UK (05/31/88)
The connection machine Lisp is described (briefly) in pp 412-418 of the second edition of Parallel Computers by Hockney and Jesshope (Adam Hilger 1988). I've just received an inspection copy and haven't seen it in the shops yet. Adrian Johnstone, Computer Science, RHBNC, University of London A.JOHNSTONE@UK.AC.RHBNC.VAXA 0784-39025
ran@mit-vax.LCS.MIT.EDU (Randy Osborne) (06/02/88)
In article <4285@mit-vax.LCS.MIT.EDU> I wrote: >Multilisp is available from the Parallel Processing Group at >M.I.T. The present implementation is a byte-code interpreter >which runs on uniprocessor machines supporting UNIX (BSD 4.2) >and the multiprocessor Encore Multimax (and our Concert >Multiprocessor). Little work should be necessary to port the >implementation to other shared memory multiprocessors, which >like the Multimax, support UNIX. > >I'll post details on how to obtain this implementation of >Multilisp if people express interest. Here's the details: To obtain the Multilisp implementation described above, obtain a licence agreement from the following address: Multilisp Request Room 205 M.I.T. Laboratory for Computer Science 545 Technology Square Cambridge, MA. 02139 Complete and return the agreement (no fee required) and we'll mail you a tape (UNIX tar format, 1600bpi). (We may have to charge for the tape if we receive too many requests.) You can also send me your postal address via e-mail. (No need to do this again for those who have already done so ...) ------------------------------------------------------------- Randy Osborne ran@vx.lcs.mit.edu