#1543 skipTypesetting doesn't behave well on its margins

[Anonymous]

Originally created by: *anonymous

Originally created by: ColinPKC...@gmail.com

Skiptypesetting seems to override commands *before* its use, in particular the \break in the following snippet, which produces different results for 2.12.3 and 2.13.51:
\version "2.12.3"
% \version "2.13.51"
bareSnippet = {
  e'1
  \break \bar "|:"
  \set Score.skipTypesetting = ##t
  s1
  \break \bar "||"
  \set Score.skipTypesetting = ##f
  c'1
}

\paper { line-width = 60\mm }
{
  \set Staff.instrumentName = #(lilypond-version)
    \bareSnippet
  }

In 2.13.51 neither \break is shown: only 1 line of output is produced.
2.12.3 produces 3 lines as expected.
Neither version sees the first \bar *before* \skiptypesetting is in effect, although 2.12.3 sees the \break, and both versions see the second \bar, *before* skiptypesetting is cancelled.

Apparently, \skiptypesetting is overriding several commands before its own invocation, so presumably it is a timing issue;).

It looks as though this was at least partially broken in 2.12.3 so I'm hesitantly calling it a regression.

[Google Importer]

Originally posted by: k-ohara5...@oco.net

Conversely, swap the first \break and \set skipTypesetting (LilyPond output does not change) and now 2.12.3 is "wrong" while 2.13.51 is "correct".

Events at the moment between measures all apply at that moment. For exampel the \bar"|:" can be before the \break but the repeat sign still correctly appears after the break.

If we really need to control the line breaks printed in drafts made with skipTypesetting, we can insert a \grace s8 to define the timing.

[Google Importer]

Originally posted by: n.putt...@gmail.com

> If we really need to control the line breaks printed in drafts made with skipTypesetting, we can insert a \grace s8 to define the timing.

A simpler method would be to unset forbidBreak.

Despite the change in behaviour, I'm not convinced this is serious enough to merit critical priority.

[Google Importer]

Originally posted by: PhilEHol...@googlemail.com

I agree with Neil's comment wrt critical - I use skipTypeSetting quite a bit, and its general behaviour is fairly erratic.  You quite often get warnings about spanner problems, and its parsing of lyrics is flaky.  My own view is that, if it did something about how it handled breaks before, that was coincidence, not intended behaviour.  If someone wants to drop everything to fix it, fine, but I don't think it's serious enough to panic.  It doesn't affect final output.  I'm pushing it back down to medium for those reasons.

Labels: -Priority-Critical Priority-Medium

[Google Importer]

Originally posted by: simon.al...@mail.de

(No comment was entered for this change.)

Summary: skipTypesetting doesn't behave well on its margins
Labels: -Priority-Medium

[Dan Eble]

\break \bar "|:"
\set Score.skipTypesetting = ##t

The prototype implementation of the \post function proposed in [#5740], when applied to this \set command, makes the line break appear in the visual score.