If I had more time, money, and a supply of patience, I'd create a website which placed events and commentary in their historical perspectives. How interesting would it be to see many people who complained about the rhetoric in last year's US Presidential election using the same sorts of arguments and innuendos in the previous election?
Another fun hobby is comparing predictions to actual results.
Take an April 2008 interview with TIOBE director Paul Jansen. (Yes, this is the same TIOBE with results no one can replicate, which apparently suggests that LOGO is making a big comeback, and which believed for several months that Google used Delphi to write their Chrome browser.) Less than a year ago, he made this statement:
C and C++ are definitely losing ground.
Fortunately, TIOBE's index provides a year-over-year delta. It's not April yet, so his prediction isn't a year old, but March 2008 to March 2009 shows that C's rating has improved by a percentage point and C++'s has improved by 1.08%.
(Delphi hasn't left the top ten yet, so I predict that their index uses the scores of previous months, compounding flaws.)
How's that prediction business working out?