Re: Add the command \offset to LilyPond (issue 8647044)

[dak]

On 2013/10/15 18:04:43, janek wrote:
> Hi,

> an afterthought.

> On 2013/10/06 01:15:12, david.nalesnik wrote:
> > The examples below represent my efforts to test the effects of
multiple
> > applications of \offset.  You can see that some accumulation is
possible.
> > [...]
> > %%%%
> > \relative c' {
> >   %% TESTS FOR ACCUMULATION %%
> >
> >   % default
> >   <c e g b>1\arpeggio
> >
> >   \override Arpeggio.positions = #'(-3.5 . 0.5)
> >   <c e g b>1\arpeggio
> >
> >   % values created by override are offset
> >   \offset #'positions #'(-2 . 2) Arpeggio
> >   <c e g b>1\arpeggio

> I'm not sure if i haven't missed something, but to your realize that
in this
> case the offset isn't applied on top of the override (as the comment
suggests),
> but replaces it?  This is self-evident in the example below:

> \relative c' {
>    % default
>    <c e g b>1\arpeggio

>    \override Arpeggio.positions = #'(-5 . 5)
>    <c e g b>1\arpeggio

>    \offset #'positions #'(-1 . 2) Arpeggio
>    <c e g b>1\arpeggio
> }

> with current master (a82d8622e6b1be36169de7d2fe1f9aa88618933b,
containing offset
> patch) the last arpeggio is shorter than the second, while it should
be longer
> in both directions.

Well, that's the slightly icky thing.  The \offset is with respect to
the situation you'll get when doing \revert Arpeggio.positions
afterwards.  So if you want to make a _relative_ offset to the first
setting, you have to do
\temporary \offset #'positions #'(-1 . 2) Arpeggio

I'm not overly enthusiastic about how the \offset command ends up when
entered as an override.  Maybe we should allow
\offset #'(-1 . 2) Arpeggio.positions here?  It seems that the offset
list should not be confusable with a symbol? or whatever is used in the
front?  So we could make that first argument optional, and complain when
the last argument is not a symbol list making up for it?

https://codereview.appspot.com/8647044/