Wednesday, January 15, 2003 :::
Yesterday I posted a math question; last night 1/(log(z)+i*pi) hit me upside the head. For |z| between .5 and 1.5 with the negative real axis removed and the principal branch of log, not only do the limits from different sides of the boundary differ greatly, approaching z=-1 diverges from one side but not the other.
It's interesting how this sort of thing works; I had been mulling over this for a week, but it was only after sharing it that I quickly came upon an answer. I have a book on the computer language C that refers to "confessional programming" at one point, where one solves a problem by explaining it to someone else, who often doesn't have to say a thing.
::: posted by dWj at 9:55 AM