[r-t] Blue line (/grid) generators / Jump Changes

[Don Morrison]

I find it mildly amusing that GMail, in its infinite wisdom, suspects
this message may be spam. Perhaps it is in cahoots with the more reactionary
elements of The Exercise.

On Sat, Mar 21, 2015 at 12:52 PM, Philip Earis <pje24 at cantab.net> wrote:
> However, I can't find anywhere the functionality to generate and
> view methods with jump changes.

The place notation parser underlying the stuff on ringing.org/methods
(as opposed to methods.ringing.org, which points to Martin's and
Richard's stuff)
does actually understand jump changes. At any point where a new change could
begin an "extended change" of the appropriate stage may be inserted within
square brackets, where an extended change is an arbitrary permutation written
just as if it were a row. So, for example, at minor

x3x[134265]x2x

applied to rounds generates the rows

123456
------
214365
124635
216453
264135
621453
624135
261453

I have found this useful not just for methods with jump changes, but for
combining blocks from fragments of compositions for various purposes,
eliding other bits that I have not settled on yet; or even for eliding
the interior bits of methods so that I just have, say, the sections
when the treble is in 1-2.

Unfortunately, while the parser understands "extended place notation,"
the method stuff in ringing.org/methods explicitly rejects it. I presume
I made this choice for several reasons:

- the blue line drawing stuff would undoubtedly get catastrophically
confused

- the false course extraction stuff examines a method for palindromic
symmetry, and that becomes somewhat harder with jump changes since,
unlike normal changes, they are not necessarily their own inverses

- the method classification engine would also get catastrophically
confused, not least because the CCCBR classification it implements
would be, if not confused in terms of broken bits, certainly confused
in spirit.

However, there seems no reason it couldn't still spit out the
appropriate rows, and just not display a blue line, falseness or
classification. I will look into what I need to do to make that
happen. That said, please don't hold your breath; I can't even
count the changes I am "planning" to make to the site that I
haven't started on yet.

[On the other hand, the current stuff on the site is completely happy
if you include the identity change, and will draw a perfectly sensible
blue line. If the method has Plain Bob lead ends/heads it will also
describe it as having false course head group 'A' :-)]

So, what *would* a blueline (or grid-like thingie) look like for
methods with jump changes?

The method library underlying ringing.org/methods is just the one
from Martin's and Richard's stuff*, which I believe is derived from
the CCCBR stuff, so there aren't going to be any treble jump
methods in there. On the other hand, once I'm able to do something
sensible with them, I certainly have no objection to using a more
comprehensive library if one were available. I'd even volunteer
to host one, but I'm so far behind in some many other things I
want to do, that I'm not sure I'd get to it in my daughter's lifetime,
let alone my own.

> It would also be helpful to document rung methods that have jump
> changes. I've certainly rung (in peals or extents) at least:
>
> Jump Stedman Doubles (4 different sixes)
> Mersey Ferry Minor
> Cambridge Jump Surprise Minor
> London Jump Surprise Minor
> Double Oxford Jump Minor
> Stedman Jump Triples (2 different sixes)

It would be nice to include other things disavowed by the Council,
too, things like the false-in-the-plain-course asymmetric plain minor
method rung in the Cambridge seven illegalities peal, and the three
lead course royal methods rung in a peal of spliced long before the
Council started accepting "differential hunters" (to my way of
thinking a horrible name that makes me think of horses with an
uncomfortable gait, or of a bellman in pursuit of infinitesimal
snarks).



* More accurately it tries to use Martin and Richard's library, but
because of a bug in my code it fails to correctly reflect things
when the CCCBR makes changes to the definition of a method (which
then *are* correctly reflected in Martin and Richard's version, just
not in mine). Fixing this is another of those uncountable changes
I'm "planning" to make.




-- 
Don Morrison <dfm at ringing.org>
"His crew are simply Tom, Dick and Harry, with the Baker as Everyman.
We are all there, all in the same boat, all heading in the wrong
direction, going the wrong way."
        -- Alexander Taylor, _The White Knight_