At the risk of carrying this thread too far (and thus missing the best stopping point):
There is always a best solution to any problem.
Ahh, that is the notion I'd like to disabuse you of because holding such a notion can be a big impediment to getting any work done, or even providing a suitable solution to a problem.
There are several problems with seeking an ideal "best solution".
- There may be many "almost best" solutions that are entirely suitable for some particular task. Seeking the "best" solution may waste a great deal of time where accepting a slightly less ideal solution provides the same effective performance and gets the job done in a timely fashion.
- There may be many essentially equivalent "best" solutions so choosing the "one best" solution may not be possible.
- There may be no "best" (or even any) solution at all so seeking one is futile.
- The problem specification may be so poor that it is not possible to determine where a best solution may lie.
The last is often encountered on PerlMonks, especially in nodes that ask "What is the best way to ...".
TIMTOWTDI is a common Perl mantra, often with the follow up "and most of them are wrong". A better codicil may be: "and none of them are best". Neither codicil need be correct, but they both contain a large grain of truth.
True laziness is hard work
|