Tromp's entry is amazing.
Source: http://www.ioccc.org/2012/tromp/tromp.c
Spoiler: http://www.ioccc.org/2012/tromp/hint.html
It's based on a minimal implementation of Binary Lambda Calculus: http://en.wikipedia.org/wiki/Binary_lambda_calculus
Source: http://www.ioccc.org/2012/tromp/tromp.c
Spoiler: http://www.ioccc.org/2012/tromp/hint.html
It's based on a minimal implementation of Binary Lambda Calculus: http://en.wikipedia.org/wiki/Binary_lambda_calculus
Apparently so. Wow.
http://web.archive.org/web/20040602202609/http://homepages.c...
It really is, which leads me to wonder how difficult it would be, to deobfuscate this into code that can be understood by the average programmer?
At least in terms of being able to say what it does, without being told.
While most would probably regard security through obscurity as bad practice, perhaps there are practical uses in certain contexts.
Most of the evolution from that point was finding ways to reduce the code size, then finally reformatting into a lambda. I credit John for not succumbing to gratuitous obfuscation, nor gratuitous use of macros to shave bytes. I suggested to him at one point to consider reducing the entire program to a single loop. He didn't think it would be shorter and pointed out that he had made a similar transformation in an earlier IOCCC entry of his, his infamous maze program, so I believed him and didn't go through the effort to verify for myself.