MOSS@cs.umass.edu ("Eliot Moss, GRC A351B, x5-4206 01-May-1989 1904") (05/02/89)
Hi, folks -- I am a professor at Umass/Amherst doing research in object oriented systems (programming languages and databases). I am interested in implementing the new language Modula-3, and the GCC technology seems interesting for that purpose. Modula-3 is an extension and rework of Modula-2, itself the successor to Pascal. Modula-3 provides a module facility (of course!), a simple object oriented type and inheritance facility, exception handling, lightweight processes (threads), and a subspace of garbage collected objects. For low level systems programming you also have a space of objects and pointers that are not garbage collected. The language was defined by some folks at DEC Systems Research Center and the Olivetti Research Center. The language DEFINITION is unrestricted; the language definition document may be freely copied so long as it is not for gain and the copyright notice appears on the copy (details on the document). It is available for DEC SRC or from Olivetti. The Olivetti folks have offered to share their compiler with me at some point, but it will not (to my knowledge) provide the sort of optimization that GCC supports -- their compiler will produce C code actually, which will then be compiled in the conventional way. Anyway, I'd like to consider whether we can get a better result as well as get an interesting language out to a lot of people who might be interested in using it, by building a Modula-3 front end and a run time system. My own goal is to add to support for persistence (accessing and automatically "faulting" objects in from an object store; I naturally happen to have an object store of my own devising available for this), though I intend to add it in a way that would not particularly disturb the language if you don't use the new features. I can provide machine resources (VaxStations mostly), and maybe some compensation from a research grant if someone wants to work with me and my graduate students on this project, though general hacking by qualified interested parties just for the fun of it could be neat, too. If it matters to anyone, Amherst is a little less than 2 hours drive west of Boston; we can be accessed via Telnet from Internet-land. You can reply to me at Moss@cs.umass.edu, or if your response is of more general interest, perhaps to the list. (Maybe someone can suggest the appropriate protocol for this sub-discussion; I am new to info-gcc; note that I am on the Internet, not UUCP.) Looking forward to the response .... Eliot Moss Asst Prof Dept of Comp and Info Sci Univ of Mass Amherst, MA 01003 Moss@cs.umass.edu