http://www.cs.caltech.edu/~adam/jedi/paper/
Abstract:
In this paper, we present the design and implementation of a Java
package, as an alternative to RMI, the Java Environment for
Distributed Invocation (JEDI), which is based on dynamic proxies.
Without statically compiled interfaces, JEDI provides a simple, dynamic,
efficient communication mechanism for Java programs. We describe the
integration of JEDI with the CORBA-compliant ILU, and present an
experiment comparing the performance of JEDI with other communication
facilities such as UDP, TCP, RMI, and the Infospheres info.net.
Speaking of which, we released our beta 3 of Infospheres last week, in
case anyone is interested:
http://www.infospheres.caltech.edu/releases/
Dr. Ernie wanted to know, as punningly as possible:
> So, the only question is: is your environment synchronous or
> asynchronous?
> Or, in other words, do we have to wait for the Return of the JEDI?
Right now the remote method calls are synchronous; we haven't had the
chance to implement callbacks yet. But we've mapped out how callbacks
could be done, we just haven't written the code.
I think Dr. Ernie's just giving us a hard time because we referenced his
Dr. Dobbs paper...
----
adam@cs.caltech.edu
It seemed slightly too trivial for a [FoRKpost]; I guess I forgot about Tim.
-- a FoRKer who wants to remain anonymous