My only question relates to the behavior of an object-filtered request
when the object changes as part of the event trigger --
STEP_INTO, STEP_OUT, MethodEntryRequest and MethodExitRequest
can have a "before" and "after" object. To which does the
filtered object apply. Is it "after" for STEP_INTO and MethodEntryRequest.
How about for out/exit? I am not sure which would be more
useful for MethodExitRequest -- filtering on the "before" object might
be useful if I am trying to debug the interactions between a service
object and its clients, while filtering on the "after" object might
be useful if I am trying to debug the interactions between a client
and all of its servants. Come to think of it, the same is true
of MethodEntryRequest -- "before" gets me client and "after"
gets me servants.

A. break on a call into this object (method entry "after")
B. break on a call out of this object (method entry "before")
C. break on a return into this object (method exit "after")
D. break on a return out of this object (method exit "before")
similarly for step into and step out
...

I guess your plan is to support A and D. You might want to make
the comments a little clearer for numbskulls like me...