29 Jan
2008
29 Jan
'08
7:40 a.m.
Neil Mitchell wrote:
Orthogonality.
Indeed. If a proposed function's name is "fooBar", because it intends to be a combination of foo and bar, then "foo . bar" is already a perfect name. I think you'd recognize a useful abstraction by the fact that it has an obvious "stand-alone" name. Otherwise, it's not really an abstraction, is it? Best regards, Johannes Walmann.