Re: ReReRe: Bug report: measure numbering after repeat with alternative endings

[Mats Bengtsson]

Paul de Bruin wrote:
> Dear Mats, Bertalan and Graham,
>  
> Flooded with and surprised by instant reactions, I've given the 
> problem (or is it a bug?) some more thought. Two points of view are 
> possible:
>  
> 1. The Programmer's view. A measure of music, like any other object in 
> programming, should have a unique number. No ambiguity is logically 
> allowed or even thinkable. So, the first measure of the second piece 
> section should be 18.
>  
> 2. The Musician's view. Measure 16b (2a volta) is 
> essentially/musically the same as measure 16a (1a volta). Moreover, 
> this particular piece (bwv 1017) has not only a harsichord part but 
> also a violin part. There, there is no distinction between 16a and 
> 16b, and in the separate violon part there is no need to write it 
> out that way at all. Would I indeed follow that choice when coding 
> the violin part, I would end up with a difference in measure numbering 
> which is rather awkward during rehearsals. So, the first measure of 
> the second piece section should be 17 in both parts.

I could also imagine the conductors view at a rehearsal. She/he wants 
everybody to start at the same place and if there are two different 
measures with the same bar number, then you can be sure that at least 
one person in the orchestra will start at the wrong place. Admittedly, 
it's probably most common to specify that you should start at the "first 
ending"/"second ending" in such situations, but still it can be a source 
of confusion.

However, I just looked in a book on music typesetting ("Noter", by B. 
Tyboni, which unfortunately only is available in Swedish), which says 
that it is most common to do what you propose, namely to use the same 
measure numbers both in the first and second ending. Still, the 
formulation "most common" indicates that both strategies are used.

  /Mats