[moca] Combinators i Have Known And Loved

[Martin Berger <martinb@xxxxxxxxxxxxxx>]

#2 sum

a perennial questions in process theory is which form of summation
to use -- if any. i'd like to know what purposes "really" require sums
and, for each such application, what a minimal dedicated
combinator doing the same job would look like.

for example, sums can used to implement value passing. a
corresponding dedicated combinator would be conventional value passing,
equivalent, seemingly, to input guarded sums where all the
guards wait on the same name. another example is using sums to
model a rudimentary form of timers. here one seems to use mixed
choice. the dedicated combinator would be a proper timer. other
examples i can think of are be expansion theorems, locking,
internal choice and solving consensus problems for symmetric
asynchronous processes.

are you aware of more examples where you think some form of sums is
necessary?

martin

 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The "models for mobility" mailing list     mailto:moca@xxxxxxxxxxxxxxx
http://www-sop.inria.fr/mimosa/personnel/Davide.Sangiorgi/moca.html