Boost logo

Boost :

From: Vesa Karvonen (vesa.karvonen_at_[hidden])
Date: 2001-07-23 00:08:37


From: "John Max Skaller" <skaller_at_[hidden]>
> Vesa Karvonen wrote:
> > It seems to me that most current metaprogramming libraries for C++ have
been
> > designed and developed without fully considering existing functional
> > programming knowledge. I think that C++ template metaprogramming could
greatly
> > benefit from functional programming knowledge.
>
> The question is, if you knew some FP, why would you bother
> with C++ in the first place?

Well, my FP experience is rather limited. Also, the platforms and products
that I've been involved with have not been particularly well suited to FP. For
example, recently I've been involved in the development of games for the Palm
platform, which is highly resource limited. I'm, however, using more and more
FP style design and programming.

Although FP seems very interesting, I'm not yet convinced that it is The
paradigm. For instance, it is interesting that according to Okasaki: "[...]
theoreticians have established lower bounds suggesting that functional
programming languages may be fundamentally less efficient than imperative
languages." Another issue is that the model of computation offered by
functional languages is decidedly different from how practical (finite
resource) computers work. Whether or not this is purely an advantage is not
obvious to me.


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk