This is very disconcerting because it looks like it would be faster to
generate mtf code, ship it as a string and parse it at the other end.
Looks like we might have to run the prototyping environment and the model
server in the same process. So, the run-time environment can still use models
coming from remote places, but most development would take place in the
combined model-server/prototyping environment process. The model editing
tools can still run as separate processes, since they don't seem to need to
bind to so many objects.
Anyway, let's wait until we get a better understanding of the numbers. We have
the fallback options mentioned above, and perhaps others I din't think of.
Pedro Szekely
USC/ISI, 4676 Admiralty Way, Marina del Rey, CA 90292
Phone: 310/822-1511, Fax: 310/823-6714
URL: http://www.isi.edu/isd/szekely.html