> Hope this helps.
Well, what might help more is if real type theorists (rather than
amateurs like me) gave a shot at replaying the GJ scenario: Define a
type system for an idealization (for example based on Featherweight
Java), and then compromise more than you thought you ever would in
order to apply to RTJ, and then, despite serious misgivings about its
resulting limitations, argue strenuously for its adoption.
:-)
-Doug
-------------------------------
JavaMemoryModel mailing list - http://www.cs.umd.edu/~pugh/java/memoryModel
This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:00:26 EDT