I bring this up because it lets me tell you a nice little piece of CS culture.
The running time of a sequence of m JOINs (unions) and
CONNECTEDs (finds) implemented this way on n objects is
.
Here, is the pseudoinverse of Ackermann's function.
It's basically the world's slowest growing function. It's almost
impossible to imagine an x large enough so that
.
Big improvement over the data structure I gave
earlier! Linear, for all intents and purposes.