[cap-talk] Real world distributed capabilities

[Stiegler, Marc D]

 E-Lib was developed for, and first used by, Electric Communities for
Habitats. That project faced many terrible problems, but E-Lib was one
of the things that solved problems rather than causing them. For
example, we had plenty of problems with deadlocks in the small parts of
the system that could not use E-Lib, but no problems with deadlocks in
the bulk of the system that did use E-Lib. The difference was quite
stark. Habitats as a whole had performance problems, but I believe they
were local-computation performance problems, not network performance
problems (though I was gone before that was resolved). 

Dean Tribble later used E-Lib for a major project for a company that
was, during the dotcom boom, one of the Great Powers of the computing
industry. Dean, how much can you say about that project? My own sense,
from a great distance, was that it went well and E-Lib was a significant
success, but that may just demonstrate my ignorance.

Using E-Lib from Java could bear a little more documentation than I
believe exists. Dean, markm, what documentation is there, aside from the
JavaDoc?

--marcs 

-----Original Message-----
From: cap-talk-bounces at mail.eros-os.org
[mailto:cap-talk-bounces at mail.eros-os.org] On Behalf Of Monty Zukowski
Sent: Monday, May 08, 2006 3:18 PM
To: cap-talk at mail.eros-os.org
Subject: [cap-talk] Real world distributed capabilities

We also need to worry about concurrency and distributed programming.  Is
anyone using the E-lib from Java?  I'd love to hear about real world
performance issues, programming complexity, etc.  I realize that Joe-E
intends to use it too, but we need to start programming already.