Re: Tuplet brackets do not respect edge skips

Hi Jean,
Yes this explains my confusion. I was assuming that all of these techniques were equivalent. I will take the time to learn how the different approaches alter the spacing and I will choose what I need.

To answer what I meant about bifurcating the bracket:
You are correct that understanding the underlying code makes this behavior more expected but here is what I thought before your explanation. “No one would ever want the output resulting from puting skips in the tuplets, it’s basically inconsistent music notation.” If the bracket is supposed to avoid the skips (this thought is my fallacy!) Then the result should be s4 * 2/3 \times 2/3 {c'4} s4 * 2/3 \times 2/3 {c'4 c'4} s4 * 2/3 meaning I would expect the brackets to wrap around the non-skips and avoid even the inner skips, which is not the current behavior. What it felt like from the user end was that lilypond was deliberately coercing the bracket to avoid the skips, which I as a user am totally able to do manually, so I didn’t like that feeling. But now understanding that s, \omit Rest, and \hide Rest all work at different times in the engraving process, I now understand that the behavior is not “inconsistent” or “coercive” like I thought. The behavior is merely a cascade from the point in time that the spacing is "ignored". I am sorry that I did not know this, because now it appears somewhat obvious that the commands are not equivalent.

regards,
greg

On Wed, Jul 5, 2023 at 3:35 AM Jean Abou Samra <jean@abou-samra.fr> wrote:

Hi,

There are three different ways to have some invisible pause:

s

\once \hide Rest r

\once \omit Rest r

None of these are equivalent. A hidden rest is basically just like a normal rest, with all the typesetting done with this normal rest in mind, except that the rest is skipped while doing the final drawing. An omitted rest is a rest, but with its drawing removed earlier during the typesetting phase, so that, for example, its extent does not factor in other things; but it's still a rest that exists. And a skip is the most radical option: from the typesetting point of view, it creates nothing at all. There is no grob associated with a skip.

I don't understand what you mean by a bracket “bifurcating” itself. What should that look like, visually?

To me, the behavior is not so different from beams: you can sometimes see things like
\version "2.24.1"

\fixed c' <<
  { d16[ s d s d s d s d s d s d s d] s }
  \\
  { s16 d[ s d s d s d s d s d s d s d] }
>>
but I have never seen
\version "2.24.1"

\fixed c' <<
  { d16[ s d s d s d s d s d s d s d \once \omit Rest r] }
  \\
  { \once \omit Rest r16[ d s d s d s d s d s d s d s d] }
>>
I can see why “your” behavior can make sense, I'm just explaining why the handling of skips in tuplet brackets is relatively consistent with some other things.

By calling it a bug, I didn't mean to sound aggressive.

No worries, I didn't interpret it so.

Jean

From:	Gregory Evans
Subject:	Re: Tuplet brackets do not respect edge skips
Date:	Wed, 5 Jul 2023 09:11:52 -0400