Compiler theory operates at mathematical time-frames, not Web 2.0's.

Best compiler texts were written before 1985 :-)

It's just a custom here as I'm sure you know.

So is "posting the obligatory xkcd comic". So is posting "[citation needed]" or reminding people that "correlation does not imply causation" like a goddamned robot.

That doesn't mean you have to do it.

But with pg making comments that are just a link to the readability bookmarklet, I think this battle might be lost.

Right, but Yegge's 2007 audience is likely a big pile of programmers coding in Java, C++, and Perl who have never even heard of Backus Naur Form.