[blub]

Yes, "A Philosophy of Software Design" is brilliant, precisely because the author supports their statements through the many experiments they performed with students and

The fact that they're the author of a well-known piece of software helps, although we should not use this as the main criteria to evaluate a text.

Offering at least evidence based on experimentation is essential if we want to move past personality cults and having entire teams work based on what someone wrote on a blog somewhere.

[sriku]

I'm wondering whether another factor that might be at play is that as consultants they may not be at liberty to discuss that kind of detail. Barring that, it would be interesting to have Fowler sit in Ousterhout's class, take on an advanced problem and see what they come up with.