Comments on: String reversing Part II: Tail Recursion

By: Javier Mena

Javier Mena — Tue, 12 Aug 2008 19:52:13 +0000

I was intensive tail recursion use in java, but I “discovered” that sun jvm (at least for java 1.5) doesn’t optimizes for tail recursion.

Try to implement a recursive fact with a large number, and it will raise a stack overflow exception.

There are many developers that don’t understand tail recursion. :(

By: moose56 blog » Blog Archive » Java string reverse methods analysed

moose56 blog » Blog Archive » Java string reverse methods analysed — Fri, 25 Jan 2008 13:49:42 +0000

[...] A Recursive version from here [...]

By: stephan

stephan — Fri, 16 Nov 2007 06:54:38 +0000

;-)

By: Jonathan Holland

Jonathan Holland — Fri, 16 Nov 2007 06:15:08 +0000

Hi Stephan,

Kudo’s for following up with the tail recursion example. Personally, unless your a LISPer, I find that tail recursive solutions are better expressed iteratively.

But thats just my opinion, and its been considered faulty before :)

By: stephan

stephan — Wed, 14 Nov 2007 18:51:41 +0000

Reading through lambda the ultimate some compilers seem to automatically change recursion to tail recursion, so the beauty in the source is preserved. But I’m no specialist in functional programming.

By: mccoyn

mccoyn — Wed, 14 Nov 2007 18:22:14 +0000

I agree, tail recursion optimizations take the beauty out of recursion. Some things can be solved very beautifully with recursion until you realize the stack frame or speed implications make it an unscaleable solution.

I know a lot of interpreted languages represent the stack as a linked list, or even a linked tree. This gets around the limits on stack space, but makes function calls a little slower.

I haven’t seen any compliers that attempt to make recursion faster other than with the tail-recursion optimization.