Re: "first form an hypothesis..."

[Leigh Palmer <palmer@SFU.CA>]

> My favorite "method" is to beat my head against a problem far into the
> night with no progress in sight, then to finally doze off.  Suddenly at
> about 3AM the full-blown solution appears, interrupting a dream and
> causing me to wake up in shock and write it down before it vanishes from
> my mind (which it will if I don't write it down.)

Exactly my method, Bill! It's also the best way to find a really
obscure bug in your code.

Leigh