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

[holger krekel]

John Roth wrote:
> 
> "Kenny Tilton" <ktilton at nyc.rr.com> wrote in message
> news:3DE2B626.6000009 at nyc.rr.com...
> >
> >
> > Martti Halminen wrote:
> > > Alexander Schmolck wrote:
> >
> > >>Well, if you prefer then elt is inconsistent :)
> > >
> > > Sure it is. We are talking about a language that had as one of its
> > > primary design criteria at least some compatibility with existing
> code
> > > bases in its predecessors, so many design warts had to be retained.
> > > Obviously not a problem for people getting to define their language
> on a
> > > clean slate.
> >
> > And this will set an upper bound on how far Python can go, with an
> > obligation not to break all the stuff piling up in the Vaults. (Not
> that
> > I am aware of any actual desire to take the language in new
> directions;
> > I get the feeling "less is more" was the Prime Directive.)
> 
> I believe he was talking about Lisp in that paragraph, not Python.
> However,
> your comment re. Python is IMO correct; there's a desire to avoid
> gratuitously breaking older code in newer releases. However, that hasn't
> stopped things like the integer division change, or the new yield
> keyword for generators. Both of those *will* break code.

How does the introduction of 'yield' break old code?  

cheers,

    holger