[cap-talk] "Composite", was "Same" key

[David Hopwood]

James A. Donald wrote:
> David Hopwood wrote:
>> Nothing else in section 6.2 contradicts a single
>> object being a special case of a composite.
> 
> That is silly.

Do you have a technical argument? If not, then there is nothing for me to
respond to here.

>> More to the point, it is useful to view it as such,
>> because clients of a composite cannot necessarily tell
>> whether it is made up of more than one object.
> 
> No.  It is often useful to view a composite as an atomic
> object.  It is never useful to view an atomic object as
> a composite.

Again, do you have a technical argument, rather than just an unsupported
assertion about what is "never useful"?

At risk of repeating myself, analysing an object as a composite/abstraction
is useful because, for a large proportion of uses, the client neither knows
nor cares whether the object depends on additional hidden objects. To require
that there not be hidden objects would be an overspecification. We should
therefore analyse most software abstractions as composites (in MarkM's sense),
even if they *may* be implemented as a single object.

-- 
David Hopwood <david.nospam.hopwood at blueyonder.co.uk>