Bezier curves   ...   what i learned today , math , cabri
wed 2005-jan-19 20:30:37 pst   ...   permalink

Yesterday I learned the math behind Bezier curves. Ever since I first started playing with a vector graphics program (probably MacDraw or SuperPaint, way back in the day), I always wondered what they were, and how those little handles manipulate the shape of the curve. Turns out the math is pretty cool.

Of course there's a whole lot of algrebra surrounding it -- basically, they're piecewise cubic polynomials. But what's most interesting is the meaning of those little handles in the drawing programs. In particular, there turns out to be this really elegant geometric construction.

( the construction )

the missing Google employee perk   ...   what i learned today
fri 2004-dec-10 14:26:25 pst   ...   permalink

Today I had lunch at the Google cafeteria (something I've done exactly once before, and that was back when the company was about one tenth its current size, and the cafeteria wasn't really much of a cafeteria).

You are probably aware that Google offers its employees some pretty impressive perks, including (among other things) free food in their cafeteria, which -- I am assured by people who are far more expert than me in the field of evaluating eating establishments -- is fabulous.

Today I learned that in spite of all the other great perks, in spite of the apparent fabulosity of other "more important" aspects of a dining facility ...

... the Google cafeteria has no straws.

stupid BIOS default   ...   what i learned today
wed 2004-sep-08 15:13:16 pdt   ...   permalink

Turns out that my work laptop (an IBM ThinkPad T20), is supposed to run at 750 MHz, not the sluggish 200 MHz at which I've been using it for over three years now. Apparently, the default settings in the BIOS were to have it step down the CPU speed after some period of inactivity, and it never then steps it back up again (until reboot). Clever.

To be fair, I would guess that the failure to step back up is a linux kernel issue, not the BIOS' fault (and so presumably, the majority of ThinkPad users don't have this problem since they run Windows), and I haven't all that often maxed out the CPU anyway... But still, over three years, I've executed only about a quarter the number of clock cycles I could have had. That's a lot of missed cycles!

colon-zero   ...   what i learned today , unix
thu 2004-apr-29 20:20:35 pdt   ...   permalink

From the procmailrc man page:

In the old days, the `:0' that marks the beginning of a recipe, had to be changed to `:n', whereby `n' denotes the number of conditions that follow.

Gotta love stuff that's just there for historical reasons.