Sorting floating-point values, and NaN

[Vladimir Panteleev]

On Tuesday, 12 November 2013 at 20:24:11 UTC, Andrei Alexandrescu 
wrote:
> Again, I'll add that "improving the documentation" helps little 
> unless RTFM is at work :o).

The distinction between good and bad documentation in this 
context is whether the respective information is discoverable. 
That is, don't make users ask "How should I have known that?". A 
warning on the documentation of std.algorithm.sort is good, but 
an article on floating-points merely included together with the 
documentation, and not referenced from anywhere, isn't.

> I think your solutions are not good. I also think my solution 
> in http://goo.gl/3dGGkf is good. (I know it sounds awful, but I 
> hope you'll appreciate the brevity :o).)

Brevity appreciated :D