Re: guile-ffi

[Marius Vollmer <mvo@zagadka.ping.de>]

Gordon Matzigkeit <gord@trick.fig.org> writes:

>  MV> That's a weak excuse, IMO.
> 
> I retract that statement.  I shouldn't have used Moore's Law as an
> excuse for ignoring performance issues.

I only said that because I know myself.  I'm using that excuse all the
time, but I'm trying to push me into being more aware of performance
issues, especially at the low levels of a system.  Although a system
might be `fast enough' in the beginning, people will always push it to
its limits.  Look at Emacs.  It's fast enough for most things.  In
fact, it's pretty fast once you realise just how much is actually done
in Elisp.  But Gnus begins to look slow for certain tasks.

There!  I'm lecturing again!  I'm sorry for that, don't take me too
seriously.

> The basis of my argument is the fact that I've played with RScheme.
> RScheme is nice in that it puts Scheme in control over C rather than
> vice versa.

Yeah, I should learn more about other Scheme implementations.

But actually, I don't find it too troublesome to write low-level code
in C.  I can't really imagine writing a largish application completely
in C or C++ any more since Scheme claimed a large amount of my brain.
But for low-level code, where you think about a concept for some days
and then write down a couple dozen lines of code, I think writing C is
ok.

But then again, I haven't really seen an alternative like RScheme yet.

> Let's pay attention to making the dynamic approach work properly, then
> integrate it with the static approach that other people like best.

Yep, you lead, I'll watch. ;-)