[r-t] Proposed definition of a peal

[Richard Smith]

Matthew Frye wrote:

> Perhaps basing it on a lead wasn't the best idea, perhaps 
> a plain course (which dixonoids have?)

The whole concept of a 'course' in dixonoids is fraught 
with confusion.

First, the length of a course (without any calls) is 
variable.  In conventional Dixons, some courses are 64 
changes long; others are 168 changes long.  It would be 
entirely possible to produce a dixonoid with one bell in the 
same place for the whole of one type of course, but for all 
the bells to be involved in a second type of course.  This 
is an issue in practice, because some otherwise-nice 
dixonoids can include a type of course that is very 
short -- perhaps only 4 changes long -- in which case, it 
may very well have several bells fixed throughout.

Second, a course isn't necessarily a linear (well, circular) 
sequence of rows.  If you draw the main part of a course as 
a loop with each row marked along the loop, some dixonoids 
have additional 'tails' (which are potentially forked). 
These are (generally short) sequences of rows that lead to 
the loop, and once you're on the loop, you'll never get back 
on to the tail.  At the ends of the tails, you'll find a 
handful of rows for which no predecesor exists under the 
rules of dixonoid.

These things make composing true extents harder, but that's 
not relevant here.  What is relevant is that using the word 
'course' opens up a whole minefield of problems if a 
dixonoid is involved.

RAS