In response to Dave Winer, yes, this is generally true. The longer coffee beans are roasted, the less caffeine they contain by volume, because the beans expand as they roast. Dark beans tend to contain less than lighter roasts. This also varies by bean variety; arabica beans have more caffeine than robusta beans, etc...

You can of course, make up for this in the brewing. Coffee made in a French press tends to have more caffeine than drip coffee.

While doing some refactoring, IntelliJ IDEA thought it would offer a helpful suggestion:

I really like IDEA, but I don't think I appreciate it calling my code "silly" (even if it is).

It is absolutely beautiful today. I'm glad I didn't go to work.

POTD

click for high-res

I'm not sure why, but I feel like posting wide images.

UnixReview: Dept. 1127: going, Going, GONE!

I just might be biased, but the creation of Unix was a crucial, ground-shaking event in computer technology. The corporate-funded R&D organization seems altogether extinct. This one's passing is notable.

This past Sunday, it rained like I can't remember it raining before.

click for high-res

I was lucky enough to be watching some soccer at the time. So several thousand of my friends and I sat beneath the second level of Giants Stadium and watched the rain fall.

click for high-res

It came down in sheets. And in torrents. And we watched.

click for high-res

Three inches fell. Fourty minutes later, it was gone.

Business 2.0: What if Google wanted to give Wi-Fi access to everyone in America?

Here is one for the geeks.

If someone walked up to you and said, "I need you to write a parser ASAP in Java that can parse basic RTF files," what would you do?

What libraries / frameworks / tools would you use? Assume you just had to separate header (stylesheet) from body, and parse out basic tags like bold and italic, as well as some "special" RTF characters (just numeric values for things like curly quotes).

I just did it by hand, but is seems like javacc or something could make this easier.