Re: JavaMemoryModel: JLS3 contains glitch concerning volatiles?

From: Victor.Luchangco@sun.com
Date: Thu Aug 18 2005 - 07:09:33 EDT


From: Bill Pugh <pugh@cs.umd.edu>
> I agree it is a fix that needs to be made. And we should start
> a document listing fixes and clarifications to the JLS, since it will
> probably be a while before even small changes to the JLS
> make it into print.

Would you be willing to maintain such a list at your website?

I'm not sure what you mean by "a while", but certainly doesn't
have to wait for another edition, or even another release of Java.
It could be just part of an errata sheet, and corrected in the
next printing, right? And couldn't that be done relatively done
relatively soon for the online version? We don't have to make
a big deal about it: just treat it as if it were a typo.

> But the fix is only to the explanation of the semantics, not
> the semantics itself. Thus, there is no question of whether we
> need to go through the JCP to get a change to the semantics
> approved.

Yes.

- Victor

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



This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:01:10 EDT