HN Explorer

5 comments

[ 2.1 ms ] story [ 25.8 ms ] thread
JadeNB
Needs more combinators! It's already got Y, so why not SKI (http://en.wikipedia.org/wiki/SKI_combinator_calculus) or BCKW (http://en.wikipedia.org/wiki/B,C,K,W_system)?

(I'm not sure if it's kosher to link to it—if not, then please let me know and I'll remove it—but I implemented a SKI combinator reducer in Perl regexes a while back: http://perlmonks.org/?node_id=809842 .)

cstrahan
Nice! I'll have to read that in full later tonight - thanks for sharing!

This was mostly just me trying to wrap my head around Church encodings and such, but I could surely do with a couple more combinators in there :).

JadeNB
Just a warning: at least for me, once I got started, it became addictive to try, Oulipo-like (http://en.wikipedia.org/wiki/Oulipo), to write simple programmes in these ever-more-restrictive languages. I was helped along immensely by Hindley–Seldin (http://www.cambridge.org/us/academic/subjects/computer-scien...) and, for, as one would expect, a more friendly but still thoroughly mathematical introduction, Smullyan (http://en.wikipedia.org/wiki/To_Mock_a_Mockingbird). I am not alone in my admiration for the latter; see, for example, http://www.angelfire.com/tx4/cus/combinator/birds.html.

EDIT to add: Speaking of 'Mockingbird' articles, you may be interested in Appendix A of http://dkeenan.com/Lambda/index.htm, whose author discusses another interpretation, attributed there to Barendregt, of Booleans and numerals. For someone who regards the Church encoding as 'intuitive', they were very surprising!