Re: page break failure

[Alexander Kobel]

Hugh Myers wrote:
> Second point taken. I hope you realize that 2.13 has bug fixes for
> 2.12--- yes? Without those, there is no project...

Yes it has (as well as enhancements), but during developement - as is an 
unstable release - they are allowed to break other things to a certain 
degree. (Very loosely speaking. There is a really strict policy about 
this, but it may happen. And tests are done more precisely for stable 
releases.)

> I don't suppose you took a look?
No, probably not. Graham spends quite some time for LilyPond, and does 
not care about ill-posed questions or bug reports. (Well, he cares, 
suffers in his inner, thinks about what went wrong with documentation, 
but does not try to get a clue or answer out of them.)
Which is the best thing he can do, since this means that he takes the 
time to get you better documentation, engaged developers, or releases. 
Which he manages, at least to a large part, amonst other things.

> Did you notice that in included tablature? If you had,
> then you would perhaps have sent me a different 'Most spacing bugs can
> be reproduced...'
Well. It certainly can be simpler than what you've written. And it's 
awkward trying to debug anything in this mess if you didn't write it 
yourself.
In your example, all the textSpan stuff is completely irrelevant, as 
well as the chords themselves (other than their vertical extent below 
the staves, i.e.). So that's a bunch of lines you can just spare out. I 
was surprised to see that the repeat mattered, though.

Finally, make sure that your code causes no warnings beyond those really 
belonging to the problem you're asking about. There was something about 
a volta spanner, which might or might not be a cause of your problem 
(since it works fine without the repeat), but I can get the bug without 
warnings.
If there /are/ warnings, however, chances are that anybody looking at 
your code immediately thinks "your fault".

> Also note that I made no claim of 'BUG'. I am new at
> this and would more likely expect the problem to have been caused by
> something I did, not a bug.

Actually, it looks like one. I remember a similar bug report about 
chords not long ago, although I did not follow it with high interest. 
The attached code reproduces the problem, and is far simpler. (Though 
even this is not as minimal as I'd wish to have it.)

For a workaround: You can include \pageBreak commands at your choice to 
get a, well, page break. Put them in, e.g. after the last line correctly 
fit unto the page, and everything's fine again.
But you're right in your initial guess that you should not have to tweak 
this yourself.


Cheers,
Alexander

\version "2.13.7"
\header {
  title = "Take some space"
  subtitle = "Sorry about this clutter:"
  subsubtitle = "I could not get the collision without it."
  composer = "The repeat / volta also seems to matter"
  arranger = "Look at the bottom system!"
}

x = \repeat unfold 3 <e'' e>4
notes = {
  \time 3/4
  \repeat volta 2 { \repeat unfold 7 \x }
  \alternative {
    { \x \x }
    { \x \x }
  }
  \repeat unfold 50 \x
  %% With manual page breaks, everything is fine again.
%  \repeat unfold 30 \x \pageBreak \repeat unfold 20 \x
}

\score {
  <<
    \new Staff \notes
    \new TabStaff \notes
  >>
}