Comment #71 on issue 3330 by [email protected]: bracketify-stencil moves grob's
refpoint
http://code.google.com/p/lilypond/issues/detail?id=3330
Regarding comment #70: this is not easily fixable since the information
whether there is a footer or not is not accessible at the time the footnote
stencil room is being calculated.
That's one consequence of the page breaker being badly designed in that it
runs in several phases that need to produce the same results, and the
design of the first phase does not retain the necessary information for
making this particular decision correctly.
The consequence for spacing is that after this patch the page break
decisions are the same as before, but the additional space that is reserved
needlessly is distributed through the whole page rather than placed below
the footnotes without good reason.
So I am leaning towards going through. The open question is what to do
about \vspace. A convert-ly rule turning \vspace #x into \combine \null
vspace #x in order to preserve the "I am also a proper line wanting
baselineskip" semantics for preexisting code seems rather awkward. For the
existing code/regtests, I prefer adding appropriate skippage (usually,
adding #1 to the value used in \vspace should be about correct). One could
do this hardcoded in convert-ly as well on the assumption that \vspace
should usually only appear in "vertical lists" as people are likely fuzzy
about whether it adds stuff above or below the baseline when within a line
(it's above, I think). That's nicer than \combine \null \vspace but also
less reliable.
--
You received this message because this project is configured to send all
issue notifications to this address.
You may adjust your notification preferences at:
https://code.google.com/hosting/settings
