New Model structure

[Mark S. Miller <mark>]

Date: Sun, 29 Oct 89 17:52:58 PST
   From: tribble (Eric Dean Tribble)

   ... I
   don't want to pay for a duplication check on every add or remove of an
   element.

That's ok.  I wouldn't think saying it's a set would obligate a
duplication check.  It would simply allow it.  
Remember, {a, b} === {a, a, b}.

   Consider the Set version of this:  If I add a dependent twice, then
   remove it once, should the dependent still receive updates?  I think
   the least surprising answer is yes.  

Hmm.  Interesting.  Note that this is also a possible reason for MuBag
(in addition to MuSet).

   Besides, I just thought of a style of usage that might cause this situation.

I curious.  What?

   Well since we need a separate class anyway (subclass of Dependent), we
   don't need to invent any Bag semantics (unless I come up with an
   interesting use).

Huh?  What about "model->dependents()"?  Even if we don't have this
message, and don't seperately reify the collection of dependents, we
should be able to explain it in terms of our other collection
abstractions.

   OK.  Thanks.  End of thread (unless you really feel like arguing :-)
   dean

I suppose I'm just in an argumentative mood.  Perhaps hypertext
shrinks of the future will coin a "last word addiction" neurosis for
people like me.  And the first lady will tell people to 
"Just Say Nothing".