harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Doug Lea ...@cs.oswego.edu>
Subject Re: Research
Date Tue, 17 May 2005 14:18:22 GMT
Geir Magnusson Jr. wrote:
> The academic community was one of the original groups we reached out  
> to via Doug Lea and others :)

Thanks for the prod. I've been pretty swamped, but here are
a few comments on top of the excellent postings by Steve Blackburn:

It would be a tragedy if Harmony did not at least learn from
work on research JVMs. Just about every aspect of VMs has
been the subject of intense research by academics and industrial
research labs over the past 5-10 years; enough that a whole
conference is now devoted to such work (VEE -- 
http://www.veeconference.org/) in addition to lots of work
reported at various OO and systems conferences.

In particular, people interested in overall architecture
might want to look into OVM (http://www.ovmj.org/)
(Disclaimer -- I was loosely and indirectly involved in some of this).
The goal was to make a highly componentized and configurable JVM; at a
fine-grained  enough level that you could even change object header
layouts and the like. The main demonstration of this was to make
a configurable real-time personality conformant to RTSJ.
While OVM is not a production JVM, it is the only JVM I know of
that has helped guide a pilotless drone airplane without crashing :-)


View raw message