Re: [GLISS] non-timed or non-musical events "z" "y"

[David Kastrup]

Graham Percival <address@hidden> writes:

> On Thu, Sep 13, 2012 at 07:18:50AM +0200, David Kastrup wrote:
>> Graham Percival <address@hidden> writes:
>> 
>> > Here's a few imaginary ways of writing the same music:
>> >   \at 4 \< \at 1 \! c'1
>> >   c'1 y4 y\< y z\!
>> >   c'1-<< y4 y\< y z\! >>
>> >   c'1-{ s4 s\< s2 z\! }
>> >
>> > Out of these, I'm most comfortable with the last one.
>> 
>> c-d-.
>> 
>> Which one gets the accent?
>
> Neither; lilypond exits with an error because there's no explicit
> -{} and because there's a pitch within the "extra timed events"
> portion.

Why do we need an explicit -{}?  What with

c-{ d e }-.

How long is this construct in music?  Which one gets the accent?  How is
this clearer than

<< c-. { d e } >> ?

Can I write

{ d e }-c ?

If not, why not?  Why is this not symmetric?  Can I write

{ d e }-c-. ?

> The idea is to line up cresc/dynamics/etc that occur within a note (or
> chord), not to replace the current method of writing multiple voices.

You are confusing multiple voices with parallel music.  << >> does not
(without \\) introduce parallel voices.  It is just a reasonably
transparent way of arranging time within a voice.

You want to replace this reasonably transparent way with something
messing around with post-event syntax which is bad enough on its own.
That's awful since it

a) obfuscates the fact that parallel music has the same level of
   priority
b) does not at all correspond to the logical structure of music and thus
   can't be reasonably translated into MusicXML or even LilyPond's own
   Music expressions.

-- 
David Kastrup