I have just returned, in one piece that is, from a 6-hour fucking marathon of an evaluation on Multiprocess Distributed System Architectures. Given 13 back-and-forth pages of regurgitated nonsense in said period, I hereby declare that as of June 11, 2011, I have no right hand! Thus, until further notice this is how we’re going to proceed:

Let K (“Khun be pa mikonam!”) be the dynamic nonserial-polyadic function based on Floyd’s 2D block-partitioned all-pairs-shortest path algorithm on a PRAM with N processors whose scattered values are the sum of infinite A values, such that A is noted by “Asabet khorde haa”. To parallelize, distribute matrix rows to N/P processors using cyclic mapping to alleviate load imbalance. This will result in an isoefficiency function of “Fuck off” with a recursive decomposition of “hymph, ba to ye kalam harf nemishe zadaaa”.

You get what I am saying. It's the only way I can remember it!