[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Is OO a deliberate fraud?
Andrew Delin writes:
Well, I don't think OO is a deliberate fraud...
Object orientation is an essential means of controlling complexity in
the design of complex systems - we can describe each change in the
context of the data involved; we can specify constraints and
relationships between data; we can classify and re-use.
When it comes to implementation, things are not so clear. OO is fine
for the design of sequential components, but the fact that the UML
lacks a behavioural semantics with a sound treatment of concurrency
reflects unsolved problems in programming language design.
The description of the problem as an impedance mismatch is a nice one
(and reminds me of the Michael Jackson reference in Tony's book).
Many of the books on OO are quite unhelpful. But that's down to the
generally poor state of the publishing industry - I don't think that
it indicates any increase in the amount of bad science; in fact, I
think that the science is in rude health. Around here, certainly.
Some of those books may be so unhelpful - wrong, even - as to
constitute fraud, but there is nothing deliberate there. And the
fraud has nothing to do with the substance of OO; it is the product of
bandwagon-jumping and self-promotion.
And I shouldn't pay it any mind, if I were you. The concurrency
revolution is here, and whatever solution emerges will need to extend
existing object technology, even if only at the design level.