JavaMemoryModel: Choice of formalism for the JSR-133 memory model

From: Bill Pugh (pugh@cs.umd.edu)
Date: Fri Apr 02 2004 - 16:53:18 EST


OK, the expert group has discussed the options for
a formalism for the JSR-133 memory model. We have decided
that, unless unforeseen problems arise, the JSR-133 memory
model will be based on the unified proposal by Jeremy,
Sarita and myself.

Other proposals, such as Vijay's CCMM, have some promise
and provide important insights. However, in the opinion
of the expert group, they are not sufficiently mature
at this point to be usable within the time requirements
of JSR-133, which is part of the Tiger 1.5.0 release
of Java.

Additional work on alternative memory models is encouraged
in that they may continue to provide additional insights
into the Java thread semantics, or useful ways of reasoning
about multithreaded Java programs.

Victor Luchangco is working with Sarita, Jeremy and myself
to develop a new presentation of the unified model. Sarita
just got back from some travel, so that isn't quite done yet, but
should be ready by Monday. If anyone is really eager to see
drafts of the current version, let me know and I can give
you today's version, but Monday's version should be better.

I'd like to thank everyone on the list for their contributions.
We aren't done yet, and I'd like to encourage everyone to
continue to ask hard questions and try to poke holes in
the unified proposal, or other aspects of the JMM.

        Bill Pugh

-------------------------------
JavaMemoryModel mailing list - http://www.cs.umd.edu/~pugh/java/memoryModel



This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:01:02 EDT