Agreed proteins as "words" encoded in a 2-bit DNA alphabet surely do get much bigger than even human languages like German. :-)
That said, "galactic algorithms" suggests impracticality/inefficiency "in the small" which is not really true of the Myers bit-vector approach. So, I don't think "only faster for big problems" is a great excuse for the wider world. Myers bit-vector is faster all the time (compared to many other things in far more frequent use in the wild). It's also not that involved. Here is a link to an implementation in about 34 lines of Nim code: https://github.com/c-blake/suggest/blob/master/suggest.nim#L...
Fair enough. IMO, Hyyro's description is easier to follow than Myers' own which may be the real explanation for its relative obscurity.
Re: concision - an incomplete summary of why I like Nim is that it is more concise than Python, as fast as C, and safe, both in its metaprogramming and otherwise.
That said, "galactic algorithms" suggests impracticality/inefficiency "in the small" which is not really true of the Myers bit-vector approach. So, I don't think "only faster for big problems" is a great excuse for the wider world. Myers bit-vector is faster all the time (compared to many other things in far more frequent use in the wild). It's also not that involved. Here is a link to an implementation in about 34 lines of Nim code: https://github.com/c-blake/suggest/blob/master/suggest.nim#L...