Stamp out singularities: DataStamps and derived PartialOrgls

[Eric Dean Tribble <tribble>]

The may semantics only has Stamps.  What is the rationale for adding
DataStamps as a distinction?

I think what we are currently calling a DataStamp is simply an
informable Stamp with the added convention that there is only one
informable Stamp for any given DataObject.  I don't see why this
should be the case in general.  Imagine that I make a partialOrgl A
and then make derived PartialOrgls B and C.  There is no reason to
prevent me from informaing B or C rather than just A.  Inform
permission on a Stamp lets a user make new informable Stamps onto the
same Partial material.  The only problem is combining two informable
Stamps with different permission sets.  Each user might have
permission to inform different parts of the underlying PartialOrgl!

Note that in any case, issues about inform permission can be
completely outside the Orgls and Berts semantics.  At the basic level,
any Stamp can be informed.  The permission system at the higher level
handles all the restrictions on informability.  For the DataStamp
model, informability is simply non-transferable.

dean