[clean-list] Clean and the future of computing

[Richard A. O'Keefe]

Siegfried Gonzi <siegfried.gonzi@kfunigraz.ac.at> wrote:
	I often read now in magazines about quantum-computers and
	DNA-computers (travellling -salesman problem is now possible to
	solve for great numbers).
	
"DNA computers" are not general purpose computers.
They are special purpose systems which use enormous parallelism to
solve combinatorial problems.  They do not solve these problems
more *efficiently* than general purpose computers.  On the contrary,
in terms of computational steps they are vastly *less* efficient.
The huge amount of parallelism available makes them produce solutions
*faster*, is all.

What this means is that for large combinatorial problems of great
economic value, DNA computing will be extremely useful.  But it is
not likely to replace conventional computers for problems that have
solutions of low-order polynomial complexity.

Quantum computers also work by massive parallelism, of a rather
weird sort.  Some have actually been built, but as yet, they can
manipulate only a few *bits* of information.  They have great
potential, but they are a very long way from practicality.

	What the case such computers will become realized lets say in 10
	years,

No, let's *not* say 10 years.  Feynman wrote about quantum computing,
and he has been dead longer than that.  Fusion power has been 50 years
away for the last 50 years.  I expect practical quantum computers to be
20 years away for the next 20 years.

	would Clean (or functional programming in general) be prepared
	for that?  Are we than also programming such computers with C/C++?

Pure functional programming languages such as Clean and Haskell will be
much better suited to quantum computers than imperative languages.

	I mean will be a functional-programming language more natural to
	such a quantum-computer?  Or have we to throw away all our
	programming languages in the future?
	
I think we might want to recall some designs from the past.
There was a small book from CWI Amsterdam "Associons and the Closure
Statement" which described a programming language that would suit
massively parallel programming very well.