OSCON 2004 - Tutorial Day 2
Getting a full night of sleep makes one feel much better, but both tutorials today are somewhat disappointing. The first, Taming Legacy Perl Code by Peter Scott, essentially indicated that the way to survive maintenance of someone else’s code is to 1) talk to the author if possible, 2) assess the author’s state of Perl knowledge before hacking away at his/her design choices (or indentation style), and 3) test, test test! The last covered most of the talk, and covered mostly the same ground as a (1-hour!) talk on—imagine—testing of Perl code that I heard at last year’s OSCON. The second, MySQL Performance Workshop by Jeremy Zawodny, pretty much said that you need to figure out what a database will actually be doing to determine how best to optimize its performance; like the first talk, this information is true but not necessarily new or enlightening. The handouts for both talks contain points that will probably be good for future reference, but I expected more pithy ideas and fewer generic topics from each. To some degree the same could be said for Damian Conway’s talk yesterday, but he’s a much more dynamic speaker than either of today’s tutors.
The evening talks were much better. Larry Wall’s State of the Onion address was entertaining as usual, who else could base a whole coherent talk on screen savers and stomach surgery? Paul Graham had some interesting insights into on hackers, nothing that would astonish anyone who’s actually worked on such projects but that might surprise their managers. And then Damian Conway stepped in to end the night with a bang—and lots of laughs. Only he could put together a coherent presentation that includes Conway’s (no relation) Game of Life, Turing machines, Perl 6, Latin, Klingon, James Clerk Maxwell, the Second Law of Thermodynamics, and a little demonology.