[Bro-Dev] Doc generation tests fail

[Jonathan Siwek]

> Ack, except that I'd also like to see that when running tests locally
> already. Can we write a btest that does just the relevant subpart of
> "make doc"? (e.g., doesn't need Sphinx).

Seems reasonable.  I'll work on that.

FYI, I'm thinking about rewriting how `make doc` does things almost entirely.  The scripts I did before don't seem like they're robust enough to deal with some of these new requirements we're working out.
 
> Would it make sense to provide our own fully predictable (and simple)
> RNG implementation (stolen from somewhere) that kicks in only when a
> seed is explicitly set?

Probably worth looking at.

- Jon