[r-t] The null change

[John Harrison]

> >> ... must include at least one non-null change.
> >
> > Do you really think that's necessary?

> Personally, no. If someone wants to do something daft it really
> doesn't bother me at all. 

I can go along with that.

> However, I suspect allowing rounds to be a method is just going too far
> for many folks, and it is sufficiently perverse that I'm equally
> unfussed if we choose to prevent it, which seems likely

That seems inconsistent.  If it is so perverse there is not need to
complicate the definition schema by banning it -  no one will invoke it and
anyone who did would be derided.  

I was also surprised to see Don contemplating 'allowing' and  'preventing'
;-)

-- 
John Harrison
Website http://jaharrison.me.uk