Re: macOS 64-bit

[Daniel Johnson]

Here is my summary of what has been discussed regarding MacOS builds 
over the last couple months. I feel like we are largely rehashing the 
same questions/answers.


1. Lilypond’s current build system is GUB. GUB is a cross-platform 
all-in-one build system which produces binaries for Linux, Windows, BSD, 
and MacOS.
2. MacOS Mojave periodically issues warnings that 32-bit applications 
will not run on the next version of MacOS.
3. Precompiled Lilypond binaries for MacOS are only available in 32-bit 
versions.
4. The above is true because in order to create a version of Lilypond 
that has GUI aspects and/or can be invoked from the Finder, it must be 
compiled against the XCode SDK.
  a. Older, 32-bit only versions of the SDK are GPL3-compatible; this 
is why 32-bit MacOS binaries have been available up to this point.
  b. Versions of the SDK with 64-bit support are not license-compatible 
with GPL3, which is the license Lilypond uses.
  c. These more modern versions of the SDK require compilation on Apple 
hardware.
  d. Therefore, 64-bit versions of Lilypond for MacOS cannot be 
compiled with GUI/Darwin support.
5. Lilypond must be compiled with GCC rather than clang, due to 
differences in how the C++ standard is interpreted regarding templates. 
There does not seem to be a reasonable solution to this issue.
6. MacOS provides a CLI binary called gcc, but it is just a thin wrapper 
around clang and therefore cannot be used.
7. Lilypond performs best (and most stably) when compiled against Guile 
1.8. Current versions of Guile are 2.x; this is the version shipped by 
default by pretty much every package manager for MacOS.
8. There are two major CLI package managers for MacOS, which are not 
interchangeable with each other: MacPorts and Homebrew. I am ignoring 
Fink for now. Both MacPorts and Homebrew supply GCC binaries.
9. Homebrew has dropped TeXLive support entirely. Lilypond requires TeX 
and supporting libraries in order to build. You can, however, install 
MacTeX using homebrew casks. With some supplemental software you can 
make Lilypond build using MacTeX.


For the record, I have successfully built 64-bit Lilypond on a Homebrew 
base. In addition to adding a few contributed taps, I had to manually 
download, build and install the following:

- TeX Gyre font
- Flex 2.5.37
- Extractpdfmark
- Lilypond itself

I did not attempt to build documentation.

Full details here: 
http://danieljohnson.name/blog/installing-lilypond-on-macos


___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: 64-bit version of Lilypond?

[Daniel Johnson]

> On Feb 23, 2019, at 5:11 PM, Carl Sorensen  wrote:
> I have no idea how best to go after creating an OSX 64-bit binary in gub.  I 
> think we can neither provide the Xcode binaries nor find a link that will 
> allow gub to download the binaries.  It may be that the only way forward on 
> OSX 64-bit is to provide a MacPorts or Homebrew solution for users to build 
> their own, which would be a shame.
> 
> Carl
If it’s any help, I recently built Lilypond with mostly-Homebrew, and blogged 
about it: http://danieljohnson.name/blog/installing-lilypond-on-macos
It’s far from pretty but it got me a native 64-bit Lilypond, without any of the 
GUI extras.

Daniel
___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel

MacOS 64-bit build

[Daniel Johnson]

Hi all,

Apologies if this is off-topic or excessively noobish.

As an educational process, I've been working on a script to get Lilypond 
2.19.82 to build on MacOS Mojave with dependencies mostly handled by 
Homebrew. I had to build extractpdfmark by hand, but the rest of the 
dependency tree is Homebrew. I am using guile 1.8.8, gcc 7.4.0, and flex 
2.6.4.  Here's my configure line:


PATH="/usr/local/opt/texinfo/bin:/usr/local/opt/flex/bin:/usr/local/opt/bison/bin:$PATH" 
\
  CC=/usr/local/bin/gcc-7 CXX=/usr/local/bin/gcc-7 
CPP=/usr/local/bin/cpp-7 \

  CPPFLAGS="-I/usr/local/include -I/usr/local/opt/flex/include" \
  LDFLAGS="-L/usr/local/lib -L/usr/local/opt/flex/lib 
-L/usr/local/opt/bison/lib" \
  ./configure --with-texgyre-dir=/Library/Fonts --disable-documentation 
--disable-optimising \

  --disable-pipe --disable-debugging --disable-guile2 --disable-profiling

On build, I get a long string of errors originating in out/lexer.cc:

out/lexer.cc:6272:46: error: cannot convert 'std::istream {aka 
std::basic_istream}' to 'std::istream* {aka 
std::basic_istream*}' in assignment

    YY_CURRENT_BUFFER_LVALUE->yy_input_file = yyin;
  ^~~~
out/lexer.cc: In constructor 'yyFlexLexer::yyFlexLexer(std::istream*, 
std::ostream*)':
out/lexer.cc:6387:75: error: 'std::basic_istream<_CharT, 
_Traits>::basic_istream() [with _CharT = char; _Traits = 
std::char_traits]' is protected within this context
 yyFlexLexer::yyFlexLexer( std::istream* arg_yyin, std::ostream* 
arg_yyout )

^
In file included from 
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/iostream:40:0,

 from out/lexer.cc:120:
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/istream:606:7: note: 
declared protected here

   basic_istream()
   ^
out/lexer.cc:6387:75: error: 'std::basic_ostream<_CharT, 
_Traits>::basic_ostream() [with _CharT = char; _Traits = 
std::char_traits]' is protected within this context
 yyFlexLexer::yyFlexLexer( std::istream* arg_yyin, std::ostream* 
arg_yyout )

^
In file included from 
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/iostream:39:0,

 from out/lexer.cc:120:
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/ostream:384:7: note: 
declared protected here

   basic_ostream()
   ^
out/lexer.cc:6389:9: error: no match for 'operator=' (operand types are 
'std::istream {aka std::basic_istream}' and 'std::istream* {aka 
std::basic_istream*}')

  yyin = arg_yyin;
 ^~~~
In file included from 
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/iostream:40:0,

 from out/lexer.cc:120:
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/istream:625:7: note: 
candidate: std::basic_istream<_CharT, _Traits>& 
std::basic_istream<_CharT, 
_Traits>::operator=(std::basic_istream<_CharT, _Traits>&&) [with _CharT 
= char; _Traits = std::char_traits]

   operator=(basic_istream&& __rhs)
   ^~~~
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/istream:625:7: note:   
no known conversion for argument 1 from 'std::istream* {aka 
std::basic_istream*}' to 'std::basic_isttream&&'
out/lexer.cc:6390:10: error: no match for 'operator=' (operand types are 
'std::ostream {aka std::basic_ostream}' and 'std::ostream* {aka 
std::basic_ostream*}')

  yyout = arg_yyout;
  ^
In file included from 
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/iostream:39:0,

 from out/lexer.cc:120:
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/ostream:402:7: note: 
candidate: std::basic_ostream<_CharT, _Traits>& 
std::basic_ostream<_CharT, 
_Traits>::operator=(std::basic_ostream<_CharT, _Traits>&&) [with _CharT 
= char; _Traits = std::char_traits]

   operator=(basic_ostream&& __rhs)
   ^~~~
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/ostream:402:7: note:   
no known conversion for argument 1 from 'std::ostream* {aka 
std::basic_ostream*}' to 'std::basic_ostream&&'
out/lexer.cc: In member function 'virtual void 
yyFlexLexer::switch_streams(std::istream*, std::ostream*)':
out/lexer.cc:6436:11: error: no match for 'operator=' (operand types are 
'std::ostream {aka std::basic_ostream}' and 'std::ostream* {aka 
std::basic_ostream*}')

   yyout = new_out;
   ^~~
In file included from 
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/iostream:39:0,

 from out/lexer.cc:120:
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/ostream:402:7: note: 
candidate: std::basic_ostream<_CharT, _Traits>& 
std::basic_ostream<_CharT, 
_Traits>::operator=(std::basic_ostream<_CharT, _Traits>&&) [with _CharT 
= char; _Traits = std::char_traits]

   operator=(basic_ostream&& __rhs)
   ^~~~
/usr/local/Cellar/gcc@7/7.4.0/include/c++/7.4.0/ostream:402:7: note:   
no known conversion for argument 1 from 'std::ostream* {aka 
std::basic_ostream*}' to 'std::basic_ostream&&'

out/lexer.cc: At global scope:
out/lexer.cc:32:25: error: prototype for 'size_t 

Re: Error compiling Lilypond 2.19.81

[Daniel Johnson]

Gentoo ships a patch that has enabled me to build against Guile 2.2.3 
and run successfully:


http://mirrors.evowise.com/gentoo-portage/media-sound/lilypond/files/lilypond-2.19.80-remove-ly-protect.patch


On 4/11/18 5:03 PM, Karlin High wrote:

On 4/11/2018 3:48 PM, Peter Engelbert wrote:
In the meantime, I also noticed that for some reason Lilypond 2.18.2 
is not

in the Debian Stretch (9) repository for ARM, but it IS in the Debian
Jessie (8) repository. So I was able to install 2.18.2 for the 
raspberry pi
by adding the Jessie repository. Is it’s absence in the Stretch 
repository

an oversight, or is there some reason for it?


LilyPond depends on Guile 1.8; more recent versions are incompatible 
so far. Newer Debian versions have removed Guile 1.8. Here are a few 
threads from the lilypond-devel list that may be of interest.


LilyPond in Debian


compiling lilypond in debian stretch with self-compiled guile-1.8




___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: Build error. MergeFonts: Can't find font: feta11.pfb

[Daniel Johnson]

In Gentoo it was discovered that there is a buggy version of t1utils which 
segfaults and causes this exact error.

Sent from my iPhone

> On Mar 11, 2018, at 2:18 PM, Werner LEMBERG  wrote:
> 
> 
>> Invoking "t1asm parmesan-noteheads14.pt1 parmesan-noteheads14.pfb"...
>> mf2pt1: You'll need either to install t1utils and rerun mf2pt1 or find
>> another way to convert parmesan-noteheads14.pt1 to
>> parmesan-noteheads14.pfb
>> […]
>> 
>> t1utils (including t1asm) are installed in the correct version, so I
>> don’t know why it doesn’t do its job …
> 
> The perl script `scripts/build/out/mf2pt1' does the following.
> 
>  ...
>  if (!execute_command 0, ("t1asm", $pt1file, $pfbfile)) {
>  die "${progname}: You'll need either to install t1utils \
>and rerun $progname or find another way \
>to convert $pt1file to $pfbfile\n"; 
>exit 1;
>  }
> 
> So: Can you manually execute t1asm?  Does it have the right
> permissions?  Is it in the path used by the lilypond build?
> 
> Can you manually execute mf2pt1?  Have a look into `mf/README' for
> some help how to do it.
> 
> 
>Werner
> ___
> lilypond-devel mailing list
> lilypond-devel@gnu.org
> https://lists.gnu.org/mailman/listinfo/lilypond-devel

___
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel

Lilypond fails to compile in GCC 4.4 - patch attached

[Daniel Johnson]

GCC 4.4 tends to be extra-picky about explicitly including header
files in C++.  Attached is a patch to resolve the issue.


0001-Add-header-includes-for-GCC-4.4.patch
Description: application/download
___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: RemoveATT no longer available in fontforge CVS

[Daniel Johnson]

Werner,

Were you able to resolve this?  I saw on fontforge-devel that George
recommended either RemoveLookup or RemoveLookupSubtable.

--Daniel

On 4/7/07, Werner LEMBERG [EMAIL PROTECTED] wrote:



George,


a lilypond build needs RemoveATT which apparently no longer exists in
the fontforge CVS -- I get this error message (which looks
grammatically incorrect to me, BTW):

  RemoveATT: This scripting function no works.

What's the new way to write the script below?


 Werner


==


[EMAIL PROTECTED]@

Open($1);
MergeKern($2)


# The AFM files of `New Century Schoolbook' family as distributed within
the
# urw-fonts-1.0.7pre41.tar.bz2 archive contain a bunch of ligatures which
# shouldn't be active by default:
#
#   T + M - trademark
#   N + o - afii61352
#   i + j - ij
#   I + J - IJ
#
# This font bundle is shipped by Fedora Core 6 and other GNU/Linux
# distributions; we simply remove those ligatures.

SelectIf(trademark, trademark, \
 afii61352, afii61352, \
 ij, ij, \
 IJ, IJ);
RemoveATT(Ligature, *, *);

Generate($3 + $fontname + .otf);

# EOF

___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: Controlling vertical spacing exactly - mini HOWTO

[Daniel Johnson]

 Hi Graham  Max (and everyone),

 I've been reading through chapter 11 on vertical spacing and getting
 some vertical spacing examples ready for the manual.

 And on reflection I think I've now figured out why it's been tricky
 for me to get a good conceptual handle on vertical spacing. So I'd
 like to bounce two ideas off of you guys and see if the way that I'm
 thinking about the problem makes sense.

 1. First, it seems that there are two *different* models for
 controlling vertical spacing, which we might term permissive and
 strict, respectively. The first model corresponds roughly to chapers
 11.3.1 and 11.3.3, and concerns overrides to VerticalAxisGroup and
 settings in the paper block like

   \override Staff.VerticalAxisGroup #'minimum-Y-extent = #'(-3 . 3)

   \paper {
  between-system-space = 1.5\cm
  between-system-padding = #1
  ragged-bottom=##f
  ragged-last-bottom=##f
   }

snip

Hi Trevor,

There is a way I've found to make the VerticalAxisGroup stuff strict.  It
roughly involves the following:

\override Staff.VerticalAxisGroup #'minimum-Y-extent = ##f
\override Staff.VerticalAxisGroup #'Y-extent = #'(-3 . 3)

Sometime in the 2.7 dev cycle this started to cause misaligned beams, and
someone was kind enough to point me in the direction of a fix, which I've
incorporated into a Scheme function.  I don't know if this bug still
exists, but I've just gotten used to including this function whenever I
want to set absolute staff distances (for two-staff choral scores, for
example). I can verify that it works with the 2.10 series.

Usage:
\include set-y-extent.ly
\new Staff {
  \setYextent #'(-3 . 3)
  a' b' c'' d''
}

And the function itself:

setYextent = #(define-music-function (parser location yExtent)
 (pair?)
  (make-music
'SequentialMusic
'elements
(list (make-music
'ContextSpeccedMusic
'property-operations
'()
'context-type
'Staff
'element
(make-music
  'OverrideProperty
  'grob-property-path
  (list (quote minimum-Y-extent))
  'grob-value
  #f
  'symbol
  'VerticalAxisGroup
  'pop-first
  #t))
  (make-music
'ContextSpeccedMusic
'property-operations
'()
'context-type
'Staff
'element
(make-music
  'OverrideProperty
  'grob-property-path
  (list (quote Y-extent))
  'grob-value
  (lambda (g)
   (begin
 (ly:axis-group-interface::height g)
 yExtent))
  'symbol
  'VerticalAxisGroup
  'pop-first
  #t)



___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: proposal: second style for quartertone accidentals

[Daniel Johnson]

Han-Wen Nienhuys wrote:

Maximilian Albert escreveu:
  

. So it seems that the use of Accidental #'style is deprecated and one
should set the alteration-alist directly. (But changing the style
property still works in v2.11.13, even if I don't know why, given that I
couldn't find the code handling it). What is the reason for this? To me
the former notation seems much more intuitive and easy to use.



The recent microtone improvements needed a much more flexible way to map
pitches onto symbols, and it seems superfluous to have two mechanisms for
setting glyphname at the same time.  It would be possible  to have a 
mechanism to set the alist based on the style property, but I thought it 
would be overkill.
  
Among other things I use Lilypond for, I transcribe Byzantine chant, 
which, rather than being based on quarter-tones, is based on 
sixth-tones.  (Actually it's based on twelfth-tones, but it's rare to 
encounter an odd number of twelfths.)  The arrow accidentals are perfect 
for this since a natural with a down-arrow can indicate a lowering by a 
sixth-tone, while a flat with an up-arrow can indicate a lowering by two 
sixth-tones, etc.


I'd like to be able to translate this accurately into midi output.  Is 
there already a mechanism in place for associating a microtonal 
relationship with the elements in an alteration-alist, or is this just a 
pipe dream?


--Daniel


___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: NoteEdit?

[Daniel Johnson]

I finally succeeded in building Canorus on my Gentoo box, though I haven't
had time to play with it yet.

Some issues:
1) It claims it needs Qt 4.1, but it actually needs 4.2. (Build will fail
on a missing header with 4.1.)
2) I had to modify some of the cmake-generated makefiles to add the -lutil
flag when linking.  This is definitely an issue on Gentoo and from my
Googling I suspect it will be an issue on SuSE as well.
3. make install installed the Century Schoolbook and Emmentaler fonts in
a subdir of /usr/local/share/fonts. Lilypond now picks up that version of
at least one of the fonts instead of its own, and thereafter GS chokes on
a bad /findfont.  I haven't resolved this yet other than to (temporarily)
remove the Canorus font directory.

 Canorus is still in alpha phase. Version 1.0 is planned for the end of
 the next year and should implement all of the NoteEdit features. Many
 exotic features are there though (like multiple viewports on the same
 score, scripting language and plugins support, support for harmonic
 function markings, Qt4, cross-platform and the best music font I've seen
 so far - Lily's emmentaler :) ).

 Regards.
 - Matevž
 Canorus Development team
 http://canorus.berlios.de


___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: Easy Note Heads and colors

[Daniel Johnson]

 Unfortunately I couldn't find any more info about brew-ez-stencil --
 even grepping the source didn't turn up anything useful.  (?!)

Try looking at brew_ez_stencil in lily/easy-notation.cc.  The relevant
block is as follows:

  SCM idx = scm_from_int (pit-get_notename ());
  SCM names = me-get_property (note-names);
  SCM charstr = SCM_EOL;
  if (scm_is_vector (names))
charstr = scm_vector_ref (names, idx);
  else
{
  char s[2] = a;
  s[0] = (pit-get_notename () + 2) % 7 + 'a';
  s[0] = toupper (s[0]);
  charstr = scm_makfrom0str (s);
}

If you \override NoteHead #note-names with a vector of note names, that
should take care of the problem.  NOTE: I am neither a C++ nor Scheme
programmer, and I haven't tested the above.


___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: Inital barnumber?

[Daniel Johnson]

Graham Percival wrote:


On 23-May-06, at 2:29 PM, Christian Conkle wrote:

I'm working on what the Lilypond docs so nicely term a musicological 
document with lilypond-book excerpts from a Bach fugue. (fis minor 
from WTCI) I'm using \set Score.currentBarNumber to keep the bar 
numbering correct, and it would be nice to have Lilypond put a bar 
number at the beginning of the excerpt.  Is this possible?


I believe that this _is_ possible, but it is much more difficult than 
it should be.  The docs say that

barNumberVisibility (procedure)
Procedure that takes an int and returns whether the corresponding 
bar number should be printed


so if you wrote a procedure to return #t if the currentBarNumber was 
1, it should work.  I'm not certain how to go about writing such a 
procedure, though.


I sent this email to -devel as well, as a feature request for a simple 
property (printFirstBarNumber ?) which does this.


Cheers,
- Graham

See this message:

http://lists.gnu.org/archive/html/lilypond-user/2005-09/msg00662.html

(This applied to the 2.7 codebase; I hope it still works.)


___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: \encoding

[Daniel Johnson]

Cameron Horsburgh wrote:

Hi folks,

The 2.4 series news file tells me that

The encoding of a file can be switched with \encoding. The strings for
markup texts are translated automatically.

Is this still relevant?

Cam


With the introduction of UTF-8 support in the 2.5 series, this became 
obsolete.



___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: please spread release announcement

[Daniel Johnson]

Good news:  As of yesterday, there is an ebuild for 2.8.0 in Gentoo 
portage.  Previously, the latest versions were 2.4.2 and 2.5.2.  It's 
good not to have to roll my own ebuilds anymore.


All Gentoo Lilypond users who update their systems regularly will see 
this if they have the unstable flag set for this application.  
Unfortunately the latest version marked stable in Portage is still 
2.0.3.  I hope this changes soon.


--Daniel


___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

VerticalAxisGroup minimum-Y-extent

[Daniel Johnson]

I've run convert-ly (2.7.18) on some files that are about a month old,
and it made the following change:

\set Staff.minimumVerticalExtent = ##f
is changed to
\override Staff.VerticalAxisGroup #'minimum-Y-extent = ##f

However, this does not force the staff-distance the way setting
minimumVerticalExtent did.  Consider the following example:

%%% BEGIN LILYPOND CODE %%%
\version 2.7.18

text = \lyricmode { Hel -- lo world }

\score {
   \context ChoirStaff 
   \context Staff = women {
   %\set Staff.minimumVerticalExtent = ##f
   \override Staff.VerticalAxisGroup #'minimum-Y-extent = ##f
   \set Staff.verticalExtent = #'(-6 . 2)
   \clef treble
   \context Voice = sop { c4 c c2 }
   }
   \context Lyrics = sopLyrics \lyricsto sop \text
   \context Staff = men {
   %\set Staff.minimumVerticalExtent = ##f
   \override Staff.VerticalAxisGroup #'minimum-Y-extent = ##f
   \set Staff.verticalExtent = #'(-2 . 6)
   \clef bass
   \context Voice = bas { c4 c c2 }
   }

}
%%% END LILYPOND CODE %%%

In 2.7.14 and earlier (and in 2.6.4), the lyrics would be placed exactly
centered between the two staves, regardless of collision with notes. 
However, in 2.7.17 the lyrics are pushed down by the low notes in the
upper staff.

Am I completely missing something, or is minimum-Y-extent not working as
advertised?

--Daniel




___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Build error for 2.7.16

[Daniel Johnson]

I am trying to build 2.7.16 on a Gentoo box, and the build is dying with
the following error:

i686-pc-linux-gnu-g++ -c -O3 -march=pentium4 -funroll-loops
-fomit-frame-pointer -pipe -DHAVE_CONFIG_H  -DNDEBUG
-DSTRING_UTILS_INLINED -I./include -I./out -I../flower/include
-I../flower/./out -I../flower/include  -O3 -march=pentium4
-funroll-loops -fomit-frame-pointer -pipe -O2 -finline-functions -g
-pipe -pthread -I/usr/include/freetype2   -I/usr/include/pango-1.0
-I/usr/include/freetype2 -I/usr/include/glib-2.0
-I/usr/lib/glib-2.0/include   -Wno-pmf-conversions -W -Wall -Wconversion
-o out/font-config-scheme.o font-config-scheme.cc
font-config-scheme.cc: In function `void display_fontset(FcFontSet*)':
font-config-scheme.cc:32: error: invalid conversion from `FcChar8*' to
`const
   char*'
make[1]: *** [out/font-config-scheme.o] Error 1
make[1]: Leaving directory
`/var/tmp/portage/lilypond-2.7.16/work/lilypond-2.7.16/lily'
make: *** [all] Error 2

Any clues on how to proceed?

--Daniel


___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Lily 2.7.7 and MIDI

[Daniel Johnson]

Hello all --

I know that MIDI output was broken under 2.7.5 (and 2.7.6?).  Is it 
functional under 2.7.7?  I'm running the 2.7 series now and I'd prefer 
not to have to do a parallel installation of 2.6 just for the sake of MIDI.


--Daniel


___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: 'make web' fails (and takes forever)

[Daniel Johnson]

Han-Wen Nienhuys wrote:

 David Bobroff wrote:


 [NB I recall that it was taking a Real Long Time to convert to PNG in
 these last two instances (i.e. for mm-rests2.ly and multi-measure-rest-
 center.ly).  Could the memory leak be leading to the failure?]


 Yes. Luckily, it only bites you if you have long lilypond-book
 documents. Still, I'm puzzled that we're seeing the problem only now,
 since the leak has been in Lily/GUILE for ages.

This has been a problem on my end for at least the last ten releases. 
If I want to generate HTML documentation, I have to make web, let it
fail, make web, let it fail again, etc.  Eventually it'll finish cleanly.

I also noticed that make web fails if I have the MS Mincho font
installed on my Gentoo box.  Since this font attempts to cover a great
portion of the defined Unicode universe, I suspect that Lilypond is
trying to use it for some symbol for which Emmentaler is the desired
font.  Next time I generate the documentation, I'll re-install MS Mincho
first and send in a report if the build fails again.

--Daniel


___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: Make web fails

[Daniel Johnson]

Hmmm... acting on a suspicion, I re-ran make web.  It got a little
further this time and then died again.  I ran it a third time and it
seems to have gotten past that roadblock.

I seem to recall that either lilypond-book or TeX itself has a tendency
to run out of available temp space, and then it dies without a
meaningful error message.  Is that what is happening here?

Daniel Johnson wrote:

Hello all...

I've been unable to make web for a number of releases now.  The make
consistently fails in the same place.  I am currently trying to make web
for 2.5.29 on a Gentoo box with the following installed:

Ghostscript: 8.51
teTeX: 2.0.2
Texinfo: 4.8
Imagemagick: 6.2.2-3
Netpbm: 10.20

I've uploaded my output to the following URLs:
http://www.theologian.org/ly/lilypond.log [dvips log]
http://www.theologian.org/ly/make-web-fail.log [last ~900 lines of
screen output]

Here is the tail-end of the make web failure:
  

snip


___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: slowdown in 2.5.15?

[Daniel Johnson]

Is there a particular .scm file in CVS which I can add to my 2.5.15 
install without having to do a complete rebuild?

Getting your feeble attempt reminds me that a programmers hunch is
usually wrong. After rewriting a 1.5 page clumsy function down to a
1.0 page function, I found that your file still took ages, and that
the actual problem was an else I accidentally misplaced somewhere
else.
fixed in CVS.
 


___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Make web on 2.5.13

[Daniel Johnson]

It's been a couple of versions in the 2.5 tree since I've been able to 
make web.  (Last working make web was for version 2.5.11.)  I'm 
including the dump from 2.5.13 -- anybody have any ideas (other than 
removing lily-1812551792 before making web)?

--Daniel
Processing `lily-1812551792.ly'
Parsing...
warning: ly:set-point-and-click called
Interpreting music... programming error: Grob `NoteHead' has no 
interface for property `text'
Continuing; crossing fingers
programming error: Grob `NoteHead' has no interface for property `text'
Continuing; crossing fingers
programming error: Grob `NoteHead' has no interface for property `text'
Continuing; crossing fingers
programming error: Grob `NoteHead' has no interface for property `text'
Continuing; crossing fingers
programming error: Grob `NoteHead' has no interface for property `text'
Continuing; crossing fingers
programming error: Grob `NoteHead' has no interface for property `text'
Continuing; crossing fingers
programming error: Grob `NoteHead' has no interface for property `text'
Continuing; crossing fingers
[5]
Preprocessing graphical objects...
Calculating line breaks... [3][5]
Writing lily-1812551792-systems.tex
Writing lily-1812551792-systems.texi
Layout output to `lily-1812551792-1.eps'...
Backtrace:
In 
/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/share/lilypond/scm/output-ps.scm:
290: 24  (let* (# # # #) (string-append #  setfont  #))
294: 25* [decode-byte-string latin1 4]
In 
/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/share/lilypond/scm/encoding.scm:
141: 26  (let* (# # #) (do # # #))
141: 27* [get-coding-vector latin1]
131: 28  [caddr ...
131: 29* [get-coding latin1]
112: 30  (let ((entry #)) (if entry (cons # #) (if # # #)))
   ...
114: 31  [cons latin1.enc ...
114: 32* [force #promise #procedure #f ()]
In unknown file:
  ?: 33* [#procedure #f ()]
In 
/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/share/lilypond/scm/encoding.scm:
 76: 34* [get-coding-from-file latin1.enc]
 65: 35  (let* ((coding #) (command #) (vector #) ...) (list command 
vector ...))
 65: 36* [read-encoding-file latin1.enc]
 15: 37  (let* ((path #) (unused #) (raw #) (string #) ...) (cons 
command vector))
 19: 38* [match:substring #f 1]
In unknown file:
  ?: 39  (let* ((matchnum #) (start #) (end #)) (and start end ...))
  ?: 40* [match:start #f 1]
  ?: 41  (let* ((matchnum (if # # 1)) (start (car #))) (if (= start -1) 
#f start))
  ?: 42* [car ...
  ?: 43* [vector-ref #f 2]

unnamed port: In procedure vector-ref in expression (vector-ref match 
matchnum):
unnamed port: Wrong type argument in position 1: #f
../../scripts/lilypond-book.py: warning: `lilypond' failed (status 2) 
(ignored)

../../scripts/lilypond-book.py: error: Process lilypond --backend eps  
--formats=ps,png --header=texidoc -I ./input/test -e '(ly:set-option 
(quote internal-type-checking) #t)' -I 
/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/Documentation/user  
-I 
/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/Documentation/user  
-I 
/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/Documentation/user/out-www  
-I /var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/input  -I 
/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/input/regression  
-I /var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/input/test  -I 
/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/input/tutorial  -I 
/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/mf/out  -I 
/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/mf/out 
lily-248913041 lily-410343011 lily-1725215974 lily-1812551792 
lily-1833613947 lily-30533 lily-1924324177 lily-1118511755 
lily-1559121341 lily-1898926172 lily-1936770277 lily-1656058444 
lily-1529620773 lily-1258123945 lily-351588894 lily-1006575972 
lily-1135070995 lily-1681340668 lily-2128551185 exited unsuccessfully.
Removing `./out-www/introduction.texi'
Removing `./out-www/lilypond.texi'
Traceback (most recent call last):
 File ../../scripts/lilypond-book.py, line 1468, in ?
   main ()
 File ../../scripts/lilypond-book.py, line 1464, in main
   ly.exit (1)
 File 
/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/share/lilypond/python/lilylib.py, 
line 132, in exit
   raise _ ('Exiting (%d)...') % i
Exiting (1)...
make[3]: *** [out-www/lilypond.texi] Error 1
make[3]: Leaving directory 
`/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/Documentation/user'
make[2]: *** [WWW] Error 2
make[2]: Leaving directory 
`/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13/Documentation'
make[1]: *** [WWW] Error 2
make[1]: Leaving directory 
`/var/tmp/portage/lilypond-2.5.13/work/lilypond-2.5.13'
make: *** [web] Error 2


___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Re: ps2pdf is called wrongly

[Daniel Johnson]

Sebastian Spaeth wrote:
Hello,
I just installed lilypond 2.5.11 on my amd64 and it works fine. 
Thanks! There is one thing that does not work:
Whenever I call lilypond -f pdf (or use the default as that creates 
pdf as well) lilypond hangs indefinitly: if I kill it with ctrl-c I get:

Konvertierung nach »CatStevens-FatherAndSon.pdf«...Error invoking 
`ps2pdf  -sPAPERSIZE=a4 CatStevens-FatherAndSon.ps  /dev/null 21 
'. Return value 2

I use ghostscript-7.07.1 the one from aladdin, and ps2pdf will also 
hang if I call it like this on the command line.

Have you tried upgrading to ghostscript 8.x?  Newer versions of Lilypond 
are known to have certain problems with the 7.x series.  Unfortunately 
ESP ghostscript is stuck at 7.07, but GPL/Aladdin is up to 8.5 (with 
8.15 being latest stable).

___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel

Questions regarding TeX backend

[Daniel Johnson]

Hello all --
First, the background:
I am creating sheet music for a project that requires that titling and 
lyric text be set in Times (or a Times lookalike). Using lilypond 2.5.8, 
I was able to use the ptmr font when invoking lilypond -b tex.  However, 
since version 2.5.10 (possibly also 2.5.9) dvips is unable to locate the 
feta fonts, so in my Postscript output there are lots of Courier glyphs 
instead of clefs, noteheads, etc. (though the text fonts are correct; 
and all fonts look right in the DVI file).  I've tried manually invoking 
dvips with the arguments given in the documentation for lilypond-book 
but that doesn't help. Here's my invocation:

dvips -Ppdf -u+lilypond.map -u+ec-mftrace.map -q -t letter -o 
${LILYFILE}.ps ${LILYFILE}.dvi

I've tried using ptmr and friends using the postscript backend but 
lilypond complains that it can't find the corresponding Type 1 fonts for 
ptmr.  I've also tried installing the ptmr .pfb files via KDE's font 
manager; lilypond stopped complaining, but the font is still incorrect 
in the PS/PDF output (it's showing up as monospace).

Now, the questions:
1. Is the TeX backend broken? Are there any command-line options that 
fix the problem?
2. By what nomenclature does one select text fonts using the postscript 
backend? Where does lilypond (or pango?) look for these font files?  
Must these fonts be visible to TeX? Can anyone give a working example?

With regard to question 2, it seems to me that section 7.3.2 in the 
manual should probably be updated too, since font handling now seems to 
vary depending upon the backend.

Thanks in advance,
--Daniel
___
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel