Apps Messaging - Semantics of a Message

[Doug Tody]

On Tue, 17 Apr 2007, John Taylor wrote:

> On 16 Apr 2007, at 16:31, Rob Seaman wrote:
>
> > I'd also argue that the astronomical software community itself
> > contains many of the "better qualified people".  We're a small, but
> > feisty, group that attracts interest from many larger players due
> > not only to the niftiness of the science of astronomy, but also to
> > the interesting bit of computing phase space that astronomical use
> > cases transect.
>
> I guess the proof will come when we see companies dropping their SOAP
> APIs in favour of VOOOM.

I realize this is probably a joke, but no one here is trying to produce
the ultimate messaging system for use outside astronomy (although I notice
PLASTIC was done on Sourceforge :-).  Rather, we are trying to meet the
needs of typical distributed/heterogeneous applications within astronomy;
not just stand-alone "tool" apps (although this is an important new use
case), but a typical range of astronomical applications.  Ideally we can
build upon what has been done with general purpose messaging capabilities
and related research outside astronomy, rather than build this up from
mere sockets, with little benefit from work done elsewhere.

Actually there are a couple of very healthy things about this discussion
(compared to some other projects I have been involved with recently):
1) the emphasis on ultimately producing something which is lightweight
and easy to use, yet suitable for our applications; 2) the willingness
to do something new, rather than merely adopt some blessed "standard"
from outside, which is a poor match to our applications.  The art is in
doing these things, while still making it general enough to last a while,
while building upon related work which IS relevant.

	- Doug