AG2 arch, api, protocol walkthrough

[Michael E. Papka]

Rick,

This is a great idea, but this will take time, first Terry and I would
outline the requirements, Ivan, Bob, and Tom then would take those
requirements, as well as the requirements collected from 4 years of AG use
and propose a solution. Then we would have a critique after Ti, Justin,
Susanne, Mark, Ed, and Eric (if we are being inclusive, plus more eyes the
better) had a chance to review. 

In the timeframe of 2.0 release dates (and I don't want to miss those dates)
this might be a bit hard. I also understand we don't want to ship/build
something that is not fully baked. What can we do in the short term? Can we
do some group brainstorming / real-time design / code review at the retreat
next week?

Mike

-----Original Message-----
From: owner-ag-dev at mcs.anl.gov [mailto:owner-ag-dev at mcs.anl.gov] On Behalf
Of Rick Stevens
Sent: Tuesday, March 25, 2003 8:24 PM
To: ag-dev at mcs.anl.gov
Subject: AG2 arch, api, protocol walkthrough



I've discussed with several of you that I would like to schedule a
walkthrough of the current design requirements, specs and plans for AG2
interfaces and architecture.  This is a critical step since we now have over
a dozen groups that planing on building to our services interfaces, registry
interfaces, network services etc.  Recently we made major inroads with the
high-energy physics community for them to based the distributed data
analysis infrastructure for the LHC on AG2 mechanisms.. this will ulitmately
affect hundreds of sites and thousands of users.  I'd like to get a
discussion thread going on scheduling this and organizing who is going to
talk about what aspects of the design and the current implementation.  I'll
toss out a strawman of something like Terry and Mike to talk to requirements
as they see them currently from the bio, hep, medical and atmospheric
modeling perspective to name few.  I'd like the team of Ivan, Tom and Bob to
present a joint set of solutions to these requirements.  I'd like Ti, Eric
and Justin to critique the proposed solutions. (note that in the future we
will change these roles so dont get hung up on the assignment of the
moment).

comments.

--Rick