Why is Python popular, while Lisp and Scheme aren't?

[Vlad S.]

anton at vredegoor.doge.nl (Anton Vredegoor) wrote in message news:<aqnv3e$ejs$1 at news.hccnet.nl>...
> On 10 Nov 2002 13:57:37 -0800, voodoo1man at hotmail.com (Vlad S.) wrote:
> 
> <snip some descriptions of the devastating effects of compromising
> Lisp>
> 
> >the language, and get a YAPFL (yet another pseudo functional
> >language). Last time I checked, Python fit into that category, so you
> >already have what you say you want.
> 
> Without confirming that I want that, I would like to mention that
> functional languages can't have side effects. So in order to generate
> any output at all some concessions have to be made. Strictly speaking
> Lisp is not a functional language. 

My definition of a pseudo-functional language is one that implements
the idea of first-class and anonymous functions (although the former
is implied by the latter) and supports recursion (which, well, implies
that it has functions =]). Continuations also seem to be becoming
popular in these languages. I think Python pretty much satisfies the
idea of a pseudo-functional language (admittedly from the little
textbook-exercise experience I have with it, and the one look I took
at stackless python), so Pythoners already have some of the major
language construct benefits that Lispers enjoy, without the
parentheses. Enjoy your syntax while you can, because judging where
the "industry" is going (Java, XML and friends), in 10 years
everything will look like Intercal.

> To get at a strictly functionanl
> language - now switching into science fiction mode - a possible
> strategy could use an idea found in one of A.E. van Vogt's novel's.
> Here a description of a part of the universe is made inside a human
> brain which is indistinguishable from the real thing up until the
> twentieth's decimal. This enables the main character to effect changes
> in the universe because at that level of equality the description of
> the universe and the universe itself cannot occupy separate locations.

No need for all that =]. You can just use Church numerals (
http://www.wikipedia.org/wiki/Church_integer ), and that satisfies my
definition, although I'm not very demanding. By that virtue Unlambda (
http://www.eleves.ens.fr:8080/home/madore/programs/unlambda/ ) claims
to be Turing complete.