JavaMemoryModel: Updated JMM schedule and status report

From: Bill Pugh (pugh@cs.umd.edu)
Date: Thu Jul 24 2003 - 17:12:48 EDT


OK, here is our updated status.

Things are getting pushed pretty tight, and I'm sorry about that.

We need to clear public review by September 8th, or else we don't go
into Tiger.

That means we should plan to go into public review August 1st, to
allow for 4 weeks of public comment, and then final votes by the
expert group.

Other than the formalization of causality, we haven't received any
recent feedback and have no current plans to change other parts of
the JMM unless we get feedback very soon.

Here is the plans Jeremy and I have:

* We've been cleaning up the new formalization of causality in the memory model
   that we made available on Monday. Once again, this doesn't change any of the
   intended semantics, it just tries to define causality in a more formal and
   clear way.

* Jeremy and I will release a beta of causality definition tomorrow. We've
   been updating the alpha releases we made on last Monday, but if you haven't
   jumped in yet, wait until tomorrow.

* We still have two slightly different but equivalent ways of defining
   causality (versions A and B). We would very much appreciate feedback
   on which version people prefer. We can work with either, and Jeremy and
   I are so deep into it that we can't get a good feel for what other
   people would find more understandable. If you can get us this feedback
   by Monday, it would be much appreciated.

* We will integrate the new definitions of causality into the spec and
   supporting documents, and have those documents available by late Wednesday
   of next week.

* I'd like to aim for agreement to go into public review by Friday,
   August 1st.

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



This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:00:46 EDT