James E. Bailey [Mon, 23 Mar 2009 19:53:57 +0000 (20:53 +0100)]
modified: Documentation/user/tutorial.itely Removed the section on polyphonic music in 2.3.5 and replaced it with a short reason and a link to section 3.2.
John Mandereau [Fri, 20 Mar 2009 08:56:21 +0000 (09:56 +0100)]
Make fc-match prefer Type 1 over TTF for NCSB files
This fixes font selection on some Debian systems with a Canorus
package that installs NCSB in TTF format, which makes fc-match prefer
TTF over Type 1 (reported by David Kastrup). Setting the foundry
field to "urw" restores fc-match selection of Type 1.
Trevor Daniels [Fri, 13 Mar 2009 09:35:19 +0000 (09:35 +0000)]
Auto-beaming changes
- add beam grouping rules for 3/8, 4/8, 6/4, 6/16,
9/4, 9/16, 12/4, 12/16 time signatures
- amend beam ending rules to limit lengths of
beamed runs of 16th and 32nd notes in some
time signatures only (these may need some
fine-tuning with experience and feedback)
- except for 4/4 time, where the previous beaming
is preserved
John Mandereau [Fri, 13 Mar 2009 07:40:09 +0000 (08:40 +0100)]
Fix Info documentation with images
This patch mainly applies suggestions from David Kastrup and Jan.
* fix symlinks creation from prefix/share/info to
prefix/share/doc/lilypond, enable it even when installing in
nonstandard prefix, and clean it up in the makefiles,
* add info and info-install toplevel targets,
* improve documentation (make help, stdout prints and Application
Usage),
* fix flats graphics in Introduction for Info format.
John Mandereau [Thu, 12 Mar 2009 12:26:04 +0000 (13:26 +0100)]
Docs: merge duplicate instances of compilation instructions
Of course, this does not prevent us to split compilation instructions
between instructions for everybody (self-builders, packagers and
developers), which should still be present in Application Usage, and
instructions only for LilyPond developers, which IMHO shall be only in
the Contributors' Guide.
-) implement \eyeglasses as markup command, using the postscript stored
in the eyeglassesps variable
-) Use this markup command in musicxml2ly, too
-) Add regtest for \eyeglasses (in a \markup and as a markup attached to
a note)
-) Remove the eyeglasses example from the postscript command and
replace it by two rings (the stereo symbol) written in postscript
Until now, \bar"." produced a single dot instead of the bar line, which
was however not used anywhere and also not documented...
Using "." for the bar line style is consistent with other types like
".|" or "|.", where a | indicates a think line and . indicates a thick
line.
-) Add function for different fraction than the actual tuplet fraction
-) Add wrapper function that appends a note with given duration after
the standard tuplet text (denominator or fraction)
-) Add function to print tuplet text with both fraction values with a
note duration attached
-) Add snippet to be included in the Docs
-) Add regtests for all these cases
-) Remove definitions from musicxml2ly and make it use these new functions
-) Add "scripts.snappizzicato" glyph to the feta font
-) Add \snappizzicato articulation function
-) Add regtest for snappizzicato
-) Remove custom definition from musicxml2ly and use the new \snappizzicato
-) Add the new articulation to our chart of articulations
Also clean up and improve some Git instructions, and rename section
"Texinfo crash course" to "Texinfo introduction and usage policy" to
make it clear that it contains some policy.
Formatting issue: using @smallexample and @exampleindent 0 to allow
long lines fit in PDF output.
Michael Käppler [Sat, 28 Feb 2009 09:20:38 +0000 (10:20 +0100)]
This fixes an endless loop in duration.cc:67.
The loop occurred when Duration::Duration(Rational r, bool scale) was called
with r.num() >= 2 * r.den(), f.e. the duration of a longa: r.num = 2,
r.den = 1. If k < 0, the left-shifting-operator << returns an undefined value
instead of right-shifting bits. The fix introduces a new function
shift_left() in misc.hh, which behaves different.
use correct left to right ordering in keySignature list
The keySignature property list of the KeySignature so far was using
reversed order (e.g. the rightmost accidental came first). This was
quite confusing when manually setting non-standard key signatures like
\set Staff.keySignature = #`(((0 . 3) . ,SHARP)
((0 . 5) . ,FLAT)
((0 . 6) . ,FLAT))
In this example, you got bes, as, fis in that reversed order rather than
in the order in which you list the alterations. Now, you'll get fis, as,
bes in that (expected) order.
Also update the snippets and add convert-ly check to warn users about the
reversion (the keySignature property was hardly ever used, though).