RE: Java memory model

["Steven Richman" <address@hidden>]

> From: Vladimir G Ivanovic
> Sent: Friday, December 07, 2001 5:51 PM
>
> Would anyone care to speculate on why it's taking so long to fix the
> Java memory model? The problem has been known since at least June 1999,
> but JSR-133 gives this timetable:
...
> I can understand that (a) the problem is hard and (b) it's important to
> get it right, but I'm still left with the feeling that progress has
> slow. What seems to be the holdup?

I think "the problem is hard" is the crux of it. It's a lot of work to find
out what's wrong with the current model, and then to collect consensus from
compiler/VM implementers, architects, and programmers on what they
want/need, and then to propose a correct solution. Plus, this effort has
been spearheaded mostly by a single man.

You should check out the archives of the JMM mailist list:
http://www.cs.umd.edu/~pugh/java/memoryModel/archive/ The first post is on
June 24, 1999, and the most recent, from November 27, 2001, is titled "The
Memory Model is _done_." There has definitely been a lot of work done over
the last 2.5 years; I don't ascribe the delay to people sitting on their
heels or to JCP red tape.

,steven.