2007/5/3, Mats Bengtsson <[EMAIL PROTECTED]>:
You are right that the handling of inter-score line distance, for
example, is lost since
each score line is generated as a separate EPS/PDF file which is then
included into
the LaTeX document using \includegraphics{...}.
However, this is not necessarily a bad choice. The problem is
complicated since you
have many different situations to handle.
I'm very well aware of that.
If you have a short musical
example, then
you typically want to include it as a figure somewhere in the document
and LilyPond
should clearly not do any page breaking or other page layout.
True, and used this way lilypond-book works excellently.
Even if
you have a
long score that fills several pages, you may want to use LaTeX to
typeset the titling
at the top of the score, so it's not obvious what page height LilyPond
should use for the
first page, if we added the possibility to include scores page by page
instead of
score line by score line in the text document.
Good point. However, if you do import a lilypond file with headers,
the headers will be included in the import. If you assume a match
between the available vertical space within a page of the LaTeX
document and the space declared available within the lilypond file,
*and* you keep the relevant LaTeX pages free of other stuff eating up
vertical space, you *could* allow Lilypond to do page-filling when
called from lilypond-book as well.
So, how do you know which scores can be page-stretched and which
should have the current behaviour? I think ragged-last-bottom is a
good indicator: If set to ##f, we could safely assume that the user
wants all kinds of page-based stretching.
I hope you realize that you can play tricks by defining the
\betweenLilyPondSystem
function in LaTeX if you want to modify the spacing between systems.
...typically like this:
\newcommand{\betweenLilyPondSystem}[1]{\vfill}
I also hope you
know that you can include the full score in a \book block if you want
the full score
to be included as a single EPS/PDF into the document, which means that
LilyPond
does all the layout decisions as usual (which of course is limited to
the case where the
score is not longer than a single page).
That's quite a limitation.
BTW, for an example of a document where I chose not to use
lilypond-book because the page-based output is vastly better (IMHO),
check out my edition of Edvard Grieg's "Album for Mandssang" (album
for male song):
http://www.lulu.com/content/737011
(In the "useless trivia" section, even the cover for that project was
typeset using Lilypond, although only as a way to start Guile to write
postscript.)
Also, please note that this isn't meant so much as criticism of
lilypond-book. It's just that Lilypond itself has taken a great leap
forward with the new page breaking algorithms, especially with
ragged-last-bottom = ##f. (The vastly improved collision avoidance is
what switched me to 2.11, though.) Go Joe! :-)
--
Arvid
_______________________________________________
lilypond-user mailing list
lilypond-user@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-user
