> Another example is prematurely choosing a complex data structure for theoretical efficiency (say, a custom tree for log(N) lookups) when the simpler approach (like a linear search) would have been acceptable for the data sizes involved.
replyTo be fair, a linear search through an array is, most of the time, faster than a hash table for sufficiently small data sizes.
I'd say that this is premature optimization.
replyIt doesn't take long for hash or tree lookups to start outperforming linear search and, for small datasets, it's not frequently the case that the search itself is a performance bottleneck.
Yeah, for every Knuth there are 10000 copies of schlemiel the painter
reply