On Wed, Apr 21, 2010 at 9:46 AM, Matthias Felleisen <matth...@ccs.neu.edu> wrote: > > On Apr 21, 2010, at 10:37 AM, Sam Tobin-Hochstadt wrote: > >> It seems like what you want is for multiple Racket programs to be able >> to share the same VM process, thus sharing memory. I think this would >> be useful in some cases, and I think people have investigated it in >> the context of Java. But it also has some drawbacks. For example, a >> crash in one program brings down everything. And SirMail and DrRacket >> can't run on separate cores. And both programs pause at the same time >> for GC. > > Our whole claim is that mred eh mrracket is an OS. We should > push it hard so that one error doesn't take down the whole thing. > And I think we can. > > (I could even imagine a fake shell :-)
FWIW, the thing I had was called "mzos". But I think a better course is to try to attempt some kind of integration / reuse for the parts where we don't see a direct research interest (but maybe you see one here?) Robby _________________________________________________ For list-related administrative tasks: http://list.cs.brown.edu/mailman/listinfo/plt-dev
