I saw these on mathoverflow:

  • The primary reason Bourbaki stopped writing books was the realization that Lang was one single person.
  • A paper should be like a miniskirt: short enough to be interesting and long enough to cover the subject.