Re: extended auto beaming

[Erik Sandberg]

On Saturday 14 August 2004 01.59, Werner LEMBERG wrote:
> > ah, i see. Sounds sensible. Now what about this example?
> >
> > -------
> >
> > |   --|
> >
> > o     o
> >
> > not 100% clear how it would it be notated, but something like
> > c8[[ c16]]
> > would work
>
> I would prefer
>
>   c8[ c16[]]
>
> since the second beam is attached to the second note only. 

That's ambiguous, what would this [] actually tell you? You already know that 
it has 2 flags so you don't need to know it.

> This makes 
> the algorithm much simpler: `[' starts a beam, `]' ends it, and this
> can be nested arbitrarily. 

No, in addition you will have to define defaults when deeper [] are left out, 
and you have to define what happens in cases like:

c8[ c16[ c16 c8 c16 c8 c16] c8]

my suggestion might seem difficult at first glance, but that's because I wrote 
it technically.. what it basically says, written out in clear text, is that 
each [..] marks that the notes within brackets are grouped, within the given 
level of beams.

The musical semantics of the notation
------
| -| |
o  o o
is really that the 16th somehow is grouped with the first eighth more than 
with the second one, thus there is a semantical relevance in saying
c8[[ c16] c8]

This also corresponds to a weakness of music notation, namely that
--------
| | -| |
o o  o o
can be interpreted both as
c8[ c8[ c16] c8]
and as
c8[[ c8 c16] c8]

> There is just a single problem, namely to 
> decide the direction (left or right) of beams attached to a single
> note:
>
>   -----------------              -----------------
>
>   |    --|        |            |      |--      |
>   |
>   |      |        |      vs.     |      |        |
>
>   o      o        o            o      o        o
>
> Perhaps this:
>
>   c8[ c16[<] c8]                  c8[ c16[>] c8]

This hurts IMHO; postfixes of this kind should be commutative, so c16][< would 
be an obfuscated way of writing the same thing. Not to mention things like
{ c8[ c16]][ c16][[ c8] }

Erik