[r-t] Re: Decisions

[Richard Smith]

On 5th January, Robin Woolley wrote:

> Richard says "I'm not convinced by the need for this as it is implied by
> the rest of the decisions". Well, this is all very well for those who can
> remember the interpretation of the decisions when you come to create an
> extension, but for 'bears of very little brain' like me, any aide-memoire is
> useful.

I think perhaps we disagree on what, exactly, the Decisions
are for.  My opinion is that the Decisions should aim to
precisely describe how exactly extension (or method
construction, or whatever) should work.  Perhaps
controversially, I do not believe it is the job of the
Decisions to make the subject approachable to everyone.  I
feel this should be left to other documents, whether written
by the MC or by others.

Let me take an analogy from computing.  C++ is a programming
language -- a language you use to instruct a computer how to
do something.  Its formal definition is in a pretty
inpenetrable 748 page document.  No-one in their right mind
would try to learn C++ from this; indeed, no-one but an
expert would benefit from reading it at all.  Its purpose is
to describe in absolute detail how everything should work.
Lots of other books and tutorials exist for actually
teaching C++.  Perhaps they make over-simplifications and
certainly they will omit some of the more esoteric corners
of the language, but, by doing so, they will do much better
job of presenting the remaining 90% of the language.

In my opinion method extension is like this.  The Decisions
should aim to be concise and exact.  They should not aim to
be a tutorial on method extension.


> [...] let's look at my standard example. We know that Roker gives two
> (terminating) extensions at royal 1BC/7HI and 1CD/1FG and an indefinite one
> to stages 4n using the path 4FG/7HI. However, this latter is 'effectively
> finite' since stage 16 gives a short course method as does, I believe, stage
> 28. This leaves us with the situation where, although 4FG/7HI is
> indefinitely extendible, it will never be rung at any higher stage than the
> 'next lowest' - let's call this the infimum - on tower bells and I don't
> suppose it'll be rung to 20 in hand. (There's a challenge for some of you!).

My argument is not that there is any practical use in
extension on 28 (or similarly many) bells.  I think that
much of the time, an indefinite extension series will have a
more "natural" feel to it than a single extension.  Why?
Because I think a pattern that can be extended indefinitely
will often be more "regular" (in some qualitive sense)
or aesthetically pleasing than one that can the extended
once and no further.

> Since Roker is only a ten,

Personally, I don't find this relevant.  If they want to
name a Royal method after their tower, they can always ring
a method of a different class (e.g. a Delight method) or
slightly change the name (e.g. Roker St Andrew's Surprise
Royal).

>  then they might have some claim to ring 'their
> own' method in their own tower. By the way, look at RW88/569. Tony Smith
> never states that Roker-10 cannot be rung, he just gives the maximus
> construction He does NOT explicitly say that the 10 bell versions cannot be
> rung.

I don't have ready access to that RW, nor do I have access
to the Decisions from 1988.  Did the 1988 Decisions include
something saying "any relationship covering an indefinite
number of stages shall have preference over a relationship
covering a limited number of stages"?  If it did, I imagine
that Tony meant his letter to say that no method could be
called "Roker Surprise Royal".

> why is it that this element
> of the decision, originally introduced in 1989 as (D)E4(b) has been altered
> so frequently? You only alter thins when they are not doing what you wanted
> them to. It was change, for example, in 1992 and I suggest that the
> interpretation of the (D)E4(b) is inconsistent with the advice paper.

Not having access to decisions from this era, can you tell
me exactly what (D)E4(b) said when it was introduced in
1989, and how it was modified in 1992?  By my reckoning,
this must have been somewhere in the section on peal
ringing.

> Richard: "I'm not convinced that there couldn't be a situation where one of
> the intermediate modes produced a "better" extension". Isn't this appealing
> to aesthetics?

Yes.  That's exactly what it is.  We have to let aesthetics
guide us at some level, otherwise we could have extensions
that bore no relation to the parent whatsoever.  Before the
CC had Decisions on extension, people extended methods by
taking what they considered to be the essential properties
of a method and trying to duplicate them on a higher stage.

For example, even without thinking about the CC Decisions, I
think it's obvious how London Minor or Bristol Major should
extend; the fact that neither extension is currently
recognised is neither here nor there.  Similarly, I think
its fairly clear what the designer had in mind when Ipswich
Minor was "extended" to Major.  Sure, we wouldn't consider
it to be an extension today, but there's a very obvious idea
being preserved.

> consider the previous (G)A2(b)(i) and ..(d)(i), both of which refer to
> Static extensions and can be conflated as "Wherever the parent has a place
> made immediately above/below the path of the treble, this characteristic
> must be retained in all extensions." This has, in (G)B7, had "above/below"
> replace by "adjacent" and made to apply to all extensions. I argued that,
> giving the original meaning of the two separate decisions elements, this new
> (G)B7 must mean, say, the extension used by Lincolnshire-8 to 10 would no
> longer be allowed. pabs said "that wasn't the intention"

Maybe I'm being naive.  Why might you think that the usual
extension of Lincolnshire is no longer allowed because of
this?

The usual extension is EBC/EDE.  In the parent, all of the
internal places above the treble are adjacent to it.  This
characteristic is preserved by the extension.  Below the
treble, all of the places (except for the thirds) are
adjacent to the treble; this property is also preserved.

I don't see what the problem is.


On 6th January, Robin Woolley wrote:

> As regards the missing out of path 'EF' in 1953, I believe that this is due
> to the same reason as the omission of 'static above'.

I think the example of Cromer Alliance that PABS gave
probably explains this quite well.  Though the absense of
any examples of EF extension is, I'm sure, relevant too.

> Regarding Vancouver, classically, no extension would be allowed.

To Royal, no.  But there's a perfectly good "classical"
extension to Maximus.

 EFG/SHI  &-3-4-5-6-4-5-36.2.7-36.2.7-36.2.7,2

But I agree this isn't just Lincolnshire with a different
frontwork, which is what you might expect.  (Pre-2004
(G)A.2(d)ii only said that "at the half-lead, a penultimate
place ... *may* expand".  It did not require it to.)

> Is 5AB/2DE a valid extension? If it is, this seems to be a departure from
> previous interpretations. It has always been the case that, to take
> expanding extension including 12, that this particular place notation must
> be preserved in the same position. The word used was 'corresponding'. (See
> Roger Bailey's paper.)

Yes.  I believe it is legal.  My 5AB/2DE extension of
Vancouver repeats the DE section below the treble (that is,
a section with no internal places).  As it is mode-2 below
the treble, it is not simply an expanding extension.
Quoting the new (G)C.2(c)i,

| An, Bn, Cn, etc. below the treble are derived in a mode-m
| extension from A, B, C, etc. by expanding all internal
| places above mths place n positions from the lead.

Thus the 12 change has no places above m=2 and so stays as a
12 change.  As to which position it's in, I think PABS has
addressed this before.  The seconds are in section H in the
parent.  Using the formula given in (G)C.2(c)ii

  DE           Royal   A B C D E  D2 E2 F2 G2 H2 I2

the H section "corresponds" to the H2 section in Royal.
Both sections contain a 12 change thus satisfying (G)B.8.

> Finally, a request for info: what are the sporadic extensions of London? I
> know about EDE/EEF and SBC/EFG.

Under the old Decisions, these were the only possibilities.

  SBC/EFG  &5-5.6.5.2.5.36.4-4.3.6-6.5,2  (8)
  EDE/EEF  &3-3.4-2-3.4-4.5.6-6.5,2  (8)

Under the new Decisions, these continue to be valid
(excepting the fact that they don't extend indefinitely).
In addition, we get two new extensions:

  4AB/2EF  &5-5.4.5.2.5.36.4-4.5.6-6.5,2  (8)
  2CD/3FG  &3-3.4-2-36.4-4.3.6-6.3,2  (8)

Richard