Tail call elimination

[Walter Bright]

bearophile wrote:
> If D wants to become more fit for functional programming, then D
> specs may talk about this, and the DMD may learn to perform part of
> such optimization, and GCC does.
> 
> One of the several possible alternative solutions: 
> http://www.score.is.tsukuba.ac.jp/~minamide/papers/sas03.pdf
> 
> A bit of related discussion: http://lambda-the-ultimate.org/node/1331



I know how to do tail call optimization (it's a very old optimization),
but there are some technical problems with it in the back end.

In any case, tail call optimization is not necessary to do D-style
functional programming, because loops with mutable indices are practical
with pure functions.