[plug] non-procedural programming

[The Thought Assassin]

On Thu, 5 Oct 2000, Mike Holland wrote:
> On Wed, 4 Oct 2000, The Thought Assassin wrote:
> > On Wed, 4 Oct 2000, Mike Holland wrote:
> > > C++ is still very procedural.
> > > Very little non-trivial programming isnt.
> > I would dispute that, but I feel I am probably just disputing your meaning
> > of "non-trivial". Any program which must do extensive interactive
> I should say something more like: very little non-academic non-trivial
> programming doesnt have lots of procedural content.
Well that is certainly not the case. Anything that is sufficiently
calculation-based should not be done procedurally. As I have said earlier
in this thread, procedurality is for interactive, temporal IO. If you are
not doing that, and your program is non-trivial, you probably shouldn't be
programming procedurally. Of course, much is done procedurally that
oughtn't be, since procedural programming is the de facto standard for the
most visible situations. (see Windows. :)

>   I have played with logical programmimg with prolog, but dont see much
> use of that in the real world. I guess that expert systems would
> count. Anybody familiar with those?
Not much real-world use for logic programming, no. I can't think of a
single example where one would want to make intensive calculations of a
purely logical nature. Expert systems probably come the closest, but these
days people have learnt that they cannot be purely logical. To my meagre
knowledge, almost all classical expert systems are functional these days.

> And it depends on your definition of "programming". Setting up a simple
> spreadsheet or database may be non-procedural, but is it "real
> programming"? :)
Of course it's real programming. Is it non-trivial? Depends on the scale
and complexity, I suppose.

-Greg