Note to other monks: This sub thread starting with my original reply to L~R is a little bit out of context. The subject of tail optimization came up in the CB and L~R's post was referenced (by L~R) and I replied to follow up on a paoint from the CB. Sorry about that if it appears out of place.
I think the important thing to remember is the the whole hype around "tail recursion" and is really about "optimized tail recursion" is due to a fundamental equivelency between a tail recursive subroutine and its iterative equivelent. Its a very useful equivelency for language designers because it means that you don't need to supply any form of iterative control structures explicitly if you provide for automatic transparent optimization of tail recursive routines. But in languages with rich loop constructs like Perl there isnt much need to exploit it.
In short whenever I hear people talking about how wonderful it is that they have found another weird way to emulate the normal tail recursion optimization in Perl I get a bit disdainful. The language comes with such a tool built in and reinventing it in kludgy ways like with goto is IMO not particularly impressive.
Having said that its a pity Perl doesnt automatically optimize tail recursion, as it can involve a notational elegance that can be quite useful. My understanding is it should be a failry straightforward optimization so I wonder why it is that it hasnt happened. It could be just simply lack of tuits to do an optimization that isnt really needed I guess but it would be interesting to know if there are deeper issues.
First they ignore you, then they laugh at you, then they fight you, then you win.