Re: [q-lang-users] "listof" surprises

[Albert G. <Dr....@t-...>]

John Cowan wrote:
> IMNSHO, rebinding of lexical variables in an inner scope is a bug in
> Algol 60 that has leaked into most other programming languages.
> It is almost always the result of a programmer error, and instead
> of quietly repairing the code with alpha conversion, compilers should
> at least warn, and ideally go BZZZT WRONG when they see such a rebinding.

Hmm, do you mean that something like \X.(\X.X) X should be an error? I
don't think that this is a very good idea because it renders alpha
conversion, as it is usually defined, invalid. You might argue that this
is just a theoretical issue and we should give in to the practical
demands of programmers here. But if you write a program that takes apart
and manipulates lambda expressions (as you can do in Q), it's quite
tedious if you can't just rename bound variables without keeping track
of the entire context.

I haven't read the original (1932?) paper on the lambda calculus by
Church, but I'm pretty sure that the possibility to rebind variables in
inner scopes was in the calculus from the very beginning, predating
Algol by some 25 years, so it's certainly not a bug that has "leaked
from Algol 60". Rather, I think that Algol picked up some ideas from the
lambda calculus there. But I'm not a lambda calculus expert, for me it's
just a convenience to quickly write down a "throwaway" function. Maybe
someone more versed in the history of the lambda calculus can shed some
light on this?

Anyway, I agree that shadowed parameters and locals are a big problem in
Algol-like languages, but IMHO that's mainly due to a programming style
which favours deep nesting of scopes with big chunks of code inside.
Alas, lots of Lisp code also looks like this, and to some extent this
also applies to Haskell and ML with their nested scoping. But IMHO it's
much less of a problem in Q (unless you try really hard to program in an
Algol'ish style in Q), since Q pretty much forces you to write lots of
little equations without nested scopes. That style also has its
problems, of course -- sometimes it's hard not to miss the wood for the
trees. ;-)

Cheers,
Albert

-- 
Dr. Albert Gr"af
Dept. of Music-Informatics, University of Mainz, Germany
Email:  Dr....@t-..., ag...@mu...
WWW:    http://www.musikinformatik.uni-mainz.de/ag