Two of the areas highlighted as needing more work are:
1. Considering whether to broaden the definition of a peal (and other lengths) so that many of the 'standard performance' requirements currently in Section K would effectively move to Section P (Performance Reporting) where they would be disclosed if not followed – the 'norms' approach.
2. Incorporating layered method ringing, such as a 12-bell performance rung using two Minor methods.
There are no doubt other areas that would also benefit from being revisited. As a start, I will send out an updated version of the document in the next few days that tries to incorporate 1 and 2 above.
Hi Tim,
I am happy to support MBD as a member of the Methods Committee. I would probably be best seconding his nomination as a newby member of the Council. I would actually like to get on the Committee myself and push for reform, but I think MBD would be a better bet. Let's talk before the meeting about a strategy.
I thought we had got rid of the 'heard by the public' requirement. Re-introducing it is a retrograde step. Why not just I.41 Performance The ringing of a Block: 1. for which details are intended to be published to the ringing community; ?
Could I also raise the possibility of allowing what I would call 'method extents' to be used in peals and quarters and for method naming. I am thinking particularly of alliance and little minor methods that do not give true extents and therefore have to be spliced under the current rules. It seems an un-necessary imposition for these methods.
I can find no definition of cross-section for the surprise etc definitions. This is a term which I can remember puzzling over about 50 years ago when I first discovered the Central Council definitions (and thought even then that many of the points were either over prescriptive or just illogical!)
As I think I have said before, if place is defined by referring to position in a row, why not use position, and thus simplify by using Place (or place-making) to refer only to repeated blows in the same position (which I think is how most practical ringers think of them) (I appreciate that there would still be confusion among those who talk about place bells when referring to starts).
I would like to (a) get onto the methods committee, preferably as chairman, (b) implement in software most of the tasks Tony Smith currently does, and move this and the method collections to a CC-owned site, and (c) push through Tim's new "definitions" at the earliest opportunity.
I know how to do step (b), which I think is the key to it all, although I will probably need people to help me (e.g. RAS). How do I do (a)? Will people support me in this?
>I hesitate on this primarily because of the complexity it might add to the document. Expanded definitions of both 'extent' and 'true' would be required.
Method Extent The complete set of distinct Rows possible at a particular Stage using a particular method or methods.
Whilst I am not happy with the wording in v7 I understand Tim's desire to
find a definition that differentiates between a 'performance' and other
ringing that is not a 'performance'.
... however the functionality is already in place to make this process even smoother.
On Wed, Apr 13, 2016 at 3:32 PM, Don Morrison <d...@ringing.org> wrote:
> I think a rough summary is that true and complete extent ceases to be meaningful in a vacuum, and instead needs to be understood as true and complete extent with respect to some property or properties.
Ok, but my question was on implementation. Would you envisage trying to flesh out these properties and the contexts in which they apply, and then include these in a future version of the Decisions? Or do you favour a different approach, in which case what broadly would that be?
Tim