Social problems of Python doc [was Re: Python docs disappointing]

[Steven D'Aprano]

On Mon, 17 Aug 2009 13:56:13 -0700, Paul Rubin wrote:

> Python 3.0 went overboard by actually removing the cmp argument and
> requiring use of the key argument.  That requires various kludges if the
> key is, say, a tree structure that has to be recursively compared with
> another such structure.  Maybe then can bring back cmp someday.

I'm having trouble understanding this, which may be my weakness rather 
than yours, but could you give an actual (simplified?) example I can run, 
where cmp would work but key wouldn't?



-- 
Steven