[r-t] A plea for help

[Mark Davies]

Philip writes,

> Many thanks Mark - that's really great. Very much appreciated.
 > I've copied below the results of search 2.  I'll try sending the
 > other files also.
>
> I now need to have a play with the results...

Didn't we once have a facility whereby the list would strip out incoming 
attachments and make them available on its web page?

Regarding the results, they need a careful look at to make sure the 
output is sensible, i.e. only includes valid methods and does not miss 
any areas of the search. QA has up to now been very light! I'll try and 
have a more thorough check tonight.

Regarding the general constraints, it has struck me that limiting the 
cycles to (123) and (45678) is not actually imposing any kind of 
restriction. The only other valid leadhead for this type of construction 
is (12345)(678), and all such methods are implicitly generated by the 
search anyway - simply ring the second half-lead first. The reason you 
cannot have any other kind of leadhead is simply that, whatever the 
three-bell group is, the changes XXX..... and .....XXX must be visited, 
and there is only one way of reaching each of them: via XX.X.... and 
....X.XX. Hence, the only way of including these changes is at the 
leadhead/halflead.

So these searches have exhausted more of the search space than I 
originally thought. I am happy to try extending/altering the allowable 
PN sets though.

MBD