[r-t] parity

[Mark Davies]

> I think that, once you apply your system to rows instead of coursing
> orders

It's the same algorithm for both, Michael. I'm sorry I cocked up the example
(I blame talking to Camp at the same time) but you can see it works.

> you are using the same system (or at least conceptually a very similar
> one) that John David is using. You are just able to stop sooner
> because you recognize that 16324578 is a negative row.

Strike a light! YES! That's the whole point - you stop sooner (or
straightaway) by using your brain-resident lookup table. If you have a
brain.

> If you admit this system works for rows, why add the unnecessary step of
> converting to a coursing order?

As I said, it reduces the problem space; and often you know the coursing
order, or can jump quickly from leadhead to CO. Especially on higher
numbers, if you're tenors-together, there's no point trying to apply John's
algorithm to e.g. leadhead 18503T2E4967. It's 64235 man, and that's as
positive as positive can be.

MBD