[r-t] 23-spliced specs

[Richard Smith]

Alex Byrne wrote:

> Richard wrote: ... but every method has a different initial four changes.
>
> I don't think there is any merit in this feature here.

OK.  That's good to know as it's really quite difficult to
get every method with a different start.  After about a
week's computer time, I had only found four such
compositions.

> This is brilliant fun
> if there's a change of method every lead, but the excitement on this plan is
> only once per course. I'd like a 7-part of these methods please.

I assume someone has already produced 7-part 23-spliced
composition with this feature?

> Anyone had any bright ideas about music optimisation?

I think your suggestion of including the number of CRUs as a
term in the merit function is probably a good start.
(Sorry MBD -- suggest something else if you want!)

I'll try to do this in the next day or two.  (There had been
a technical problem stopping me from implementing this.
Hopefully that is resolved now.)

> I think this is what
> will determine which compositions are rung. I think the plan has excellent
> scope for bands wanting to ring 23-spliced without necessarily having the
> experience to ring (well) all the split-tenors faffing that usually comes
> with it.

I hadn't thought of it like that.  (To be honest, I'm still
a bit surprised that there's any non-theoretical interest in
it at all.)

I guess if you're looking at it as an easier way of ringing
23-spliced, having every lead different might be a bit
tricky.  Also, the methods that I have found so far leave
quite a bit to be desired from ease-of-ringing point of
view.  Compare it to Smith's 23, for example.  At least half
of the methods in Smith's are likely to be familiar already.

Richard