Ralph's comments on JRWG's "Parallel OGRe" paper [Sections 1-3] ...


2.1 - 2.3
At this point I felt it might be helpful to be given at the outset an
enumeration of the classes of symbol - data constructors, process state
constuctors, primitive value constructors, process names, data variables
(have I got it right?).  As I understand it, what you call process names
are dummy variables - in the sense that their interpretation is always
relative to the context of use (unlike constructors - which are drawn from
a static "constant" set).  And these process names are presumably disjoint
from the variable names used to denote values in rules?

In 2.2, by "Messages" do you mean message values/contents - i.e. the
component that comes after the "!"?  Also, I think it's worth emphasising
that in a ! blah, a is the _target_: a reader familiar with Occam/CSP would
tend to think of a as something at least partially owned by, and thus
intimately associated with, the _sender_ of the message.

2.4
in Figure 1, the left input to m is IntVal(3), where in Figure 2 you have
IntVal(15).


3.2
Final rule of Figure 4: shouldn't "DQ" be "VQ"?

3.4
Penultimate rule (first for Free) of Figure 8: shouldn't "Data(v)" "Data(d)"?