please bump lilypond to version 2.8.1 fontforge-20060406 fixes bug 127723, so you might depend lilypond on that and remove the "addwrite /root/.PfaEdit" workaround. current documentation revision is 4, so SRC_URI needs update.
Created attachment 86439 [details, diff] diff against lilypond-2.8.0-r1
Created attachment 86712 [details, diff] diff against lilypond-2.8.0-r1 lilypond-2.8.2 released. The documentation revision is 1 again.
Now it's up to 2.8.5-1. Someone, please update the ebuild! Waiting months for simple little incremental upgrades is really shaking my faith in Gentoo.
As far as I can tell, the updated ebuild in this bug report works quite well for lilypond version 2.8.6 with documentation revision 1. It is much better than 2.8.0.
Agreed - what is blocking this now. Even in stable the only things I can see is a fontforge upgrade. Is this all we are waiting for?
Is there any progress yet? This is really getting frustrating.
Now fontforge is marked stable (on x86) could we try for the version bump to 2.8.6 ?
OK - My last comment wasn't too helpful. I have successfully emerged 2.8.6 and checked that the output is correct. What I did: Used the patch enclosed in this bug to generate a lilypond-2.8.6.ebuild in my portage overlay. Added "app-text/mftrace" and "media-sound/lilypond" to /etc/portage/package.keywords. Succesfully emerged the ~x86 lilypond 2.8.6. As as test I grabbed at random "Trio Sonata IV" from this link http://www.mutopiaproject.org/cgibin/make-table.cgi?Composer=BachJS The output of lilypond SonataIV.ly from a quick inspection matches exactly that from the download offered on the site. Things look good for the 2.8.6 patch. I ask that this be applied to portage.
*** Bug 155058 has been marked as a duplicate of this bug. ***
Created attachment 101863 [details] New ebuild for 2.10.0 Ok, so it's not exactly related to 2.8.1, but since the other bug was marked a duplicate... here's the ebuild for lilypond 2.10.0. Enjoy.
Created attachment 104178 [details] Ebuild for lilypond 2.10.2 Latest and greatest. There are some problems with version detection in the config file, so use the patch for now if you have troubles.
Created attachment 104179 [details, diff] Patch for version detection issues Stick this in your /usr/local/portage/media-sound/lilypond/files/ directory.
The patching did not succeed :( >>> Unpacking lilypond-2.10.2.tar.gz to /var/tmp/portage/media-sound/lilypond-2.10.2/work * Applying lilypond-2.10.2-version.patch ... * Failed Patch: lilypond-2.10.2-version.patch ! * ( /usr/local/portage/media-sound/lilypond/files/lilypond-2.10.2-version.patch ) * * Include in your bugreport the contents of: * * /var/tmp/portage/media-sound/lilypond-2.10.2/temp/lilypond-2.10.2-version.patch-25526.out ^^^
The patching did not succeed :( >>> Unpacking lilypond-2.10.2.tar.gz to /var/tmp/portage/media-sound/lilypond-2.10.2/work * Applying lilypond-2.10.2-version.patch ... * Failed Patch: lilypond-2.10.2-version.patch ! * ( /usr/local/portage/media-sound/lilypond/files/lilypond-2.10.2-version.patch ) * * Include in your bugreport the contents of: * * /var/tmp/portage/media-sound/lilypond-2.10.2/temp/lilypond-2.10.2-version.patch-25526.out ^^^ find attached
Created attachment 104347 [details] patching of 2.10.2 failed
Created attachment 104371 [details, diff] A patch for lilypond-2.10.2.ebuild to hopefully fix the patching issue Patch your lilypond-2.10.2.ebuild with this, and see if that works. It looks like portage unpacks your packages into /var/tmp/portage/media-sound/lilypond, while mine go into /var/tmp/portage/lilypond.
Created attachment 104475 [details] Ebuild for Lilypond 2.10.3 Attached is a revised ebuild for 2.10.3. Differences include: - Reject guile 1.6.8 as a dependency since lilypond chokes on it. Instead, require 1.6.7 or 1.8.x. - Remove t1utils as a dependency, since it is really a dependency only of mftrace. - Require Ghostscript 8.x. - Add gtk2 USE flag for optional building of the GTK+ backend (experimental). - Eliminate recursive symlinks in the documentation, since they prevent a clean uninstall. - Fix broken paths in etf2ly and midi2ly. This ebuild uses the lilypond-2.10.2-version.patch supplied above.
2.10.3.ebuild cannot download the tar.gz from --23:56:12-- http://distfiles.gentoo.org/distfiles/lilypond-2.10.3.tar.gz 23:56:12 ERROR 404: Not Found. --23:56:12-- http://distro.ibiblio.org/pub/linux/distributions/gentoo/distfiles/lilypond-2.10.3.tar.gz 23:56:13 ERROR 404: Not Found. --23:56:13-- http://lilypond.org/download/v2.10.3/lilypond-2.10.3.tar.gz 23:56:13 ERROR 404: Not Found. and for the patched for patching issue 2.10.2 I'm still getting the same patching error :/
(In reply to comment #16) > Created an attachment (id=104475) [edit] > Ebuild for Lilypond 2.10.3 > > Attached is a revised ebuild for 2.10.3. Differences include: > > - Reject guile 1.6.8 as a dependency since lilypond chokes on it. Instead, > require 1.6.7 or 1.8.x. > - Remove t1utils as a dependency, since it is really a dependency only of > mftrace. > - Require Ghostscript 8.x. > - Add gtk2 USE flag for optional building of the GTK+ backend (experimental). > - Eliminate recursive symlinks in the documentation, since they prevent a clean > uninstall. > - Fix broken paths in etf2ly and midi2ly. > > This ebuild uses the lilypond-2.10.2-version.patch supplied above. > FYI I've added your ebuild to the pro-audio overly
sorry about the download errors -- everything's now fine, just my wordwrap cutted 1-2 from get_version_component_range to the next line. Still the patch does not work; even the patched patch. I suppose, it's only me with such problem? %)
Created attachment 105185 [details, diff] lilypond-sourcetree.patch The lilypond source tree has been moved on the lilypond.org web server, so the ebuild will need to be patched with the attached patch. (I have also used this same ebuild [renamed] to build 2.10.7.)
Created attachment 105651 [details] lilypond-2.10.8.ebuild
Created attachment 105653 [details, diff] files/lilypond-2.10.8-version.patch
Note: the dependency on bug 160124 is for building documentation; if you have USE="doc" you need to apply the patch at bug 160124 to ghostscript-gnu or ghostscript-gpl.
(In reply to comment #24) > Note: the dependency on bug 160124 is for building documentation; if you have > USE="doc" you need to apply the patch at bug 160124 to ghostscript-gnu or > ghostscript-gpl. Bother, got the bugs mixed up. That should be bug 160294.
"make web" has historically been the main source of Gentoo lilypond bugzilla entries. That's why the ~2.8.0 versions have downloaded the prebuilt doc tarball instead of running "make web". I understand that we want to accommodate French-speaking users with automatic language selection (if hosted on Apache), but French users can bookmark /usr/share/doc/lilypond-2.10.x/Documentation/user/lilypond/index.fr.html and they're good to go from there with the doc tarball. I'm just saying that having Lilypond generate its own documentation will double or triple the build time and the bugreports, with relatively little payoff.
Another note: sys-apps/findutils should be added to DEPEND for Ed's ebuild as well as mine, since xargs and find are both parts of that package. This is also a problem with the existing lilypond-2.8.0-r1 ebuild.
Created attachment 106145 [details] lilypond-2.10.8.ebuild re comment #27: Thanks. Added here.
(In reply to comment #26) > "make web" has historically been the main source of Gentoo lilypond bugzilla > entries. That's why the ~2.8.0 versions have downloaded the prebuilt doc > tarball instead of running "make web". I understand that we want to > accommodate French-speaking users with automatic language selection (if hosted > on Apache), but French users can bookmark > /usr/share/doc/lilypond-2.10.x/Documentation/user/lilypond/index.fr.html and > they're good to go from there with the doc tarball. > > I'm just saying that having Lilypond generate its own documentation will double > or triple the build time and the bugreports, with relatively little payoff. And? This is a source distro: we build everything we can. Also, building documentation is a good test of whether the build was successful. The documentation tarball could be installed as app-doc/lilypond-docs, perhaps?
(In reply to comment #29) > The documentation tarball could be installed as app-doc/lilypond-docs, perhaps? > That would be nice, although the thing that always seems to take the most time on my system is rebuilding all the fonts... every. single. time. :-)
1) It seems to be dev-scheme/guile, not dev/util: * dev-scheme/guile Available versions: 1.4-r3:1.4 1.4.1:1.4.1 1.6.4-r1:12 ~1.6.6:12 1.6.7:12 Installed: 1.6.7 Homepage: http://www.gnu.org/software/guile/ Description: Scheme interpreter 2) I get an error when compiling: ... MIDI-Ausgabe nach »typography-demo.midi«... Spur... Layout nach »typography-demo.ps« ausgeben... Writing typography-demo-1.signature Writing typography-demo-2.signature Writing typography-demo-3.signature Writing typography-demo-4.signature/var/tmp/portage/lilypond-2.10.8/work/lilypond-2.10.8/out/bin/../share/lilypond/2.10.8/scm/framework-ps.scm:345:9: In procedure scm_allocate_string in expression (ly:ttf->pfa file-name): /var/tmp/portage/lilypond-2.10.8/work/lilypond-2.10.8/out/bin/../share/lilypond/2.10.8/scm/framework-ps.scm:345:9: Argument 1 out of range: 16844822 make[2]: *** [out-www/typography-demo.pdf] Fehler 2 make[2]: Leaving directory `/var/tmp/portage/lilypond-2.10.8/work/lilypond-2.10.8/input' make[1]: *** [WWW] Fehler 2 make[1]: Leaving directory `/var/tmp/portage/lilypond-2.10.8/work/lilypond-2.10.8' make: *** [web] Fehler 2 !!! ERROR: media-sound/lilypond-2.10.8 failed. Call stack: ebuild.sh, line 1546: Called dyn_compile ebuild.sh, line 937: Called src_compile lilypond-2.10.8.ebuild, line 82: Called die Satux ~ # emerge --info Portage 2.1.1-r2 (default-linux/x86/2006.1, gcc-4.1.1, glibc-2.4-r4, 2.6.18-suspend2-r3 i686) ================================================================= System uname: 2.6.18-suspend2-r3 i686 Intel(R) Pentium(R) M processor 1.60GHz Gentoo Base System version 1.12.6 Last Sync: Fri, 12 Jan 2007 12:31:01 +0000 ccache version 2.4 [enabled] app-admin/eselect-compiler: [Not Present] dev-java/java-config: 1.3.7, 2.0.30 dev-lang/python: 2.4.3-r4 dev-python/pycrypto: 2.0.1-r5 dev-util/ccache: 2.4-r6 dev-util/confcache: [Not Present] sys-apps/sandbox: 1.2.17 sys-devel/autoconf: 2.13, 2.61 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2, 1.10 sys-devel/binutils: 2.17 sys-devel/gcc-config: 1.3.14 sys-devel/libtool: 1.5.22 virtual/os-headers: 2.6.17-r2 ACCEPT_KEYWORDS="x86" ACCEPT_LICENSE="" ALSA_CARDS="intel8x0" ALSA_PCM_PLUGINS="adpcm alaw asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa lfloat linear meter mulaw multi null plug rate route share shm softvol" ANT_HOME="/usr/share/ant-core" ARCH="x86" AUDIODEV="default" AUTOCLEAN="yes" CBUILD="i686-pc-linux-gnu" CCACHE_DIR="/var/tmp/ccache" CCACHE_SIZE="1024M" CFLAGS="-march=pentium-m -O2 -fomit-frame-pointer -fno-ident -pipe" CHOST="i686-pc-linux-gnu" CLASSPATH="." CLEAN_DELAY="5" CONFIG_PROTECT="/etc /usr/kde/3.5/env /usr/kde/3.5/share/config /usr/kde/3.5/shutdown /usr/share/X11/xkb /usr/share/config" CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/gconf /etc/java-config/vms/ /etc/revdep-rebuild /etc/splash /etc/terminfo /etc/texmf/web2c" CVS_RSH="ssh" CXXFLAGS="-march=pentium-m -O2 -fomit-frame-pointer -fno-ident -pipe" DISPLAY=":0.0" DISTDIR="/usr/portage/distfiles" EDITOR="/usr/bin/emacs" ELIBC="glibc" EMERGE_DEFAULT_OPTS="-av" EMERGE_WARNING_DELAY="10" FEATURES="autoconfig ccache distlocks metadata-transfer sandbox sfperms strict" FETCHCOMMAND="/usr/bin/wget -t 5 -T 60 --passive-ftp -P ${DISTDIR} ${URI}" GCC_SPECS="" GDK_USE_XFT="1" GDM_LANG="de_DE.utf8" GENERATION="2" GENTOO_MIRRORS="http://ftp.belnet.be/mirror/rsync.gentoo.org/gentoo/" GUILE_LOAD_PATH="/usr/share/guile/1.6" G_BROKEN_FILENAMES="1" G_FILENAME_ENCODING="UTF-8" HOME="/root" INFOPATH="/usr/share/info:/usr/share/binutils-data/i686-pc-linux-gnu/2.17/info:/usr/share/gcc-data/i686-pc-linux-gnu/4.1.1/info:/usr/share/info/emacs-22.0.92" INPUT_DEVICES="keyboard mouse synaptics" JAVAC="/etc/java-config-2/current-system-vm/bin/javac" JAVA_HOME="/etc/java-config-2/current-system-vm" JDK_HOME="/etc/java-config-2/current-system-vm" KDEDIRS="/usr/kde/3.5:/usr:/usr/local:/usr" KDE_IS_PRELINKED="1" KERNEL="linux" LANG="de_DE.utf8" LC_ALL="de_DE.utf8" LDFLAGS="-Wl,-O1 -Wl,--sort-common -Wl,--as-needed" LESS="-R -M --shift 5" LESSOPEN="|lesspipe.sh %s" LINGUAS="de" LOGNAME="root" MAKEOPTS="-j2" ...
(In reply to comment #31) > 1) It seems to be dev-scheme/guile, not dev/util: Yeah, it was moved on the 10th. I'll update the ebuild. > 2) I get an error when compiling: > make: *** [web] Fehler 2 That's a documentation build breakage. I'll put together lilypond-docs so you can set USE="-doc".
Created attachment 106813 [details] lilypond-2.10.8.ebuild category move: dev-util/guile -> dev-scheme/guile
Created attachment 106815 [details] lilypond-docs-2.10.9.ebuild
Note: 2.10.10 is out. No changes to ebuild needed; just do a straight version bump.
If the fix to bug 161949 works we can probably simplify the guile dep.
Created attachment 106874 [details] lilypond-2.10.10.ebuild Yes, it works once guile is fixed. This adds an extra bug dependency to this bug, but simplifies other stuff.
We'll wait for the guile bump...
Created attachment 107881 [details] lilypond-2.10.13.ebuild Now that guile-1.8.1 is in portage, we can dep on it. Note also lilypond-docs can be bumped to 2.10.13, but no changes needed to ebuild.
What is the 2.10.8-version.patch for?
(In reply to comment #40) > What is the 2.10.8-version.patch for? The version check (used for bison, gcc, guile, python) breaks on some version strings. The patch fixes it.
(In reply to comment #41) > The version check (used for bison, gcc, guile, python) breaks on some version > strings. The patch fixes it. I had it choke on mftrace-1.1.19, but using 1.2.9 it worked nicely. Any other specific cases?
(In reply to comment #42) > (In reply to comment #41) > > The version check (used for bison, gcc, guile, python) breaks on some version > > strings. The patch fixes it. > > I had it choke on mftrace-1.1.19, but using 1.2.9 it worked nicely. Any other > specific cases? No, but there will be next time lilypond depends on a package version x.y.z where z>0. It would be best to keep the patch in. That, or fix upstream. It's pretty stupid depending on higher versions than lilypond actually needs, just because their configure script is broken.
(In reply to comment #43) > No, but there will be next time lilypond depends on a package version x.y.z > where z>0. It would be best to keep the patch in. That, or fix upstream. Can I expect you have submitted this patch upstream?
(In reply to comment #44) > (In reply to comment #43) > > No, but there will be next time lilypond depends on a package version x.y.z > > where z>0. It would be best to keep the patch in. That, or fix upstream. > > Can I expect you have submitted this patch upstream? Was doing just as you posted ;) http://code.google.com/p/lilypond/issues/detail?id=254
(In reply to comment #45) > (In reply to comment #44) > > (In reply to comment #43) > > > No, but there will be next time lilypond depends on a package version x.y.z > > > where z>0. It would be best to keep the patch in. That, or fix upstream. > > > > Can I expect you have submitted this patch upstream? > > Was doing just as you posted ;) > > http://code.google.com/p/lilypond/issues/detail?id=254 > Actually, I grabbed that patch from a Lilypond mailing list... all it does is revert to an earlier version of the configure script (in regards to version checking). I don't know _why_ they broke it, but they definitely know about it.
Created attachment 108117 [details] files/lilypond-2.10.13-documentation-version.patch Fix for documentation build; some regression test files had a 2.11.x version string, but work with 2.10.13.
Created attachment 108118 [details] lilypond-2.10.13.ebuild
when USE="doc", emake web failed, because of error: can't find font: `aybabtu'
(In reply to comment #49) > when USE="doc", emake web failed, because of > error: can't find font: `aybabtu' That font's shipped (sources, as mf/aybabtu.pe.in) and should get built using fontforge. Maybe an error further up?
The previous was prolly a parallel build error. However now I get this error: Interpreting music... error: cannot find `emmentaler-20.otf' Music font has not been installed properly.
(In reply to comment #51) > error: cannot find `emmentaler-20.otf' > Music font has not been installed properly. Again, that's generated, this time by buildscripts/gen-emmentaler-scripts.py.
Created attachment 108272 [details] lilypond-2.10.13.ebuild Change >=dev-scheme/guile-1.8.1 dependency to >=dev-scheme/guile-1.6.7. This is because, until bug 163921 is fixed, most users will not be able to upgrade to guile-1.8.1. Contrary to what the Lilypond website says[1], lilypond-2.10.13 will install and work fine with guile 1.6.7, as long as the patch at bug 161949 is applied. 1. http://lilypond.org/doc/v2.10/Documentation/topdocs/INSTALL.html
Error building lilypond-2.10.13.ebuild. Any ideas? [ebuild N ] media-sound/lilypond-2.10.13 USE="gtk vim -debug -doc -emacs -profile" LINGUAS="-fr" 0 kB [1] Total: 1 package (1 new), Size of downloads: 0 kB Portage overlays: [1] /usr/portage/local/layman/ecatmur -------------------------------------------------------------------------- make[1]: Entering directory `/var/tmp/portage/media-sound/lilypond-2.10.13/work/lilypond-2.10.13/lily' cp -p /var/tmp/portage/media-sound/lilypond-2.10.13/work/lilypond-2.10.13/config.hh out/config.hh rm -f ./out/accidental-engraver.dep; DEPENDENCIES_OUTPUT="./out/accidental-engraver.dep ./out/accidental-engraver.o" i686-pc-linux-gnu-g++ -c -O2 -march=pentium-m -pipe -Woverloaded-virtual -DHAVE_CONFIG_H -DNDEBUG -I./include -I./out -I../flower/include -I../flower/./out -I../flower/include -O2 -march=pentium-m -pipe -O2 -finline-functions -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/accidental-engraver.o accidental-engraver.cc ./include/lily-guile-macros.hh: In function 'scm_unused_struct* scm_or_str2symbol(const char*)': ./include/lily-guile-macros.hh:36: error: 'scm_str2symbol' was not declared in this scope ./include/lily-guile.hh: In function 'bool ly_is_list(scm_unused_struct*)': ./include/lily-guile.hh:89: error: 'SCM_NFALSEP' was not declared in this scope ./include/lily-guile.hh: In function 'bool ly_is_procedure(scm_unused_struct*)': ./include/lily-guile.hh:90: error: 'SCM_NFALSEP' was not declared in this scope ./include/lily-guile.hh: In function 'bool ly_is_port(scm_unused_struct*)': ./include/lily-guile.hh:91: error: 'SCM_NFALSEP' was not declared in this scope ./include/lily-guile.hh: In function 'bool ly_is_equal(scm_unused_struct*, scm_unused_struct*)': ./include/lily-guile.hh:95: error: 'SCM_NFALSEP' was not declared in this scope ./include/lily-guile.hh: In function 'bool ly_scm2bool(scm_unused_struct*)': ./include/lily-guile.hh:98: error: 'SCM_NFALSEP' was not declared in this scope ./include/lily-guile.hh: In function 'scm_unused_struct* ly_bool2scm(bool)': ./include/lily-guile.hh:100: error: 'SCM_BOOL' was not declared in this scope ./include/lily-guile.hh: In function 'scm_unused_struct* ly_append2(scm_unused_struct*, scm_unused_struct*)': ./include/lily-guile.hh:104: error: 'scm_listify' was not declared in this scope ./include/lily-guile.hh: In function 'scm_unused_struct* ly_append3(scm_unused_struct*, scm_unused_struct*, scm_unused_struct*)': ./include/lily-guile.hh:108: error: 'scm_listify' was not declared in this scope ./include/lily-guile.hh: In function 'scm_unused_struct* ly_append4(scm_unused_struct*, scm_unused_struct*, scm_unused_struct*, scm_unused_struct*)': ./include/lily-guile.hh:112: error: 'scm_listify' was not declared in this scope make[1]: *** [out/accidental-engraver.o] Error 1 make[1]: Leaving directory `/var/tmp/portage/media-sound/lilypond-2.10.13/work/lilypond-2.10.13/lily' make: *** [all] Error 2 !!! ERROR: media-sound/lilypond-2.10.13 failed. Call stack: ebuild.sh, line 1611: Called dyn_compile ebuild.sh, line 968: Called qa_call 'src_compile' environment, line 3668: Called src_compile lilypond-2.10.13.ebuild, line 80: Called die -------------------------------------------------------------------------- dep lilypond media-sound/lilypond-2.10.13: =sys-devel/automake-1.10* sys-devel/automake-1.10 || >=app-text/ghostscript-esp-8.15 app-text/ghostscript-esp-8.15.3-r1 >=app-text/mftrace-1.1.19 app-text/mftrace-1.1.19 >=dev-lang/perl-5.8.0-r12 dev-lang/perl-5.8.8-r2 >=dev-lang/python-2.4 dev-lang/python-2.4.4 >=dev-scheme/guile-1.6.7 dev-scheme/guile-1.8.1-r1 >=media-gfx/fontforge-20060406 media-gfx/fontforge-20061220 >=media-libs/fontconfig-2.2.0 media-libs/fontconfig-2.4.2 >=media-libs/freetype-2 media-libs/freetype-2.1.10-r2 >=sys-apps/texinfo-4.8 sys-apps/texinfo-4.8-r5 >=sys-devel/autoconf-2.59 sys-devel/autoconf-2.61 >=sys-devel/flex-2.5.4a-r5 sys-devel/flex-2.5.33-r1 >=sys-devel/gcc-4.0 sys-devel/gcc-4.1.1-r3 >=sys-devel/make-3.80 sys-devel/make-3.81 gtk2? >=x11-libs/gtk+-2.4 x11-libs/gtk+-2.10.9 >=x11-libs/pango-1.6.0 x11-libs/pango-1.14.9 sys-apps/findutils sys-apps/findutils-4.3.2 sys-devel/bison sys-devel/bison-2.3 sys-devel/libtool sys-devel/libtool-1.5.22 virtual/tetex app-text/tetex-3.0_p1-r3
try some guile use flags, especially deprecated
Thank you! emerge =guile-1.6.7 solved the problem.
Another compile error: Converting to PNG...[/var/tmp/portage/media-sound/lilypond-2.10.14/work/lilypond-2.10.14/input/regression/out-www/lily-14eebdf8c3.eps]GPL Ghostscript 8.54 (2006-05-17) Copyright (C) 2006 artofcode LLC, Benicia, CA. All rights reserved. This software comes with NO WARRANTY: see the file PUBLIC for details. ] error: failed files: "lily-d01674577e lily-80103376d8" Ed, could my doc compile errors be because of bug 160294?
(In reply to comment #57) > Ed, could my doc compile errors be because of bug 160294? Possibly. You'll need to search up in the logs for the actual errors on those files, if they appear at all; could be easily a couple of thousand lines above.
Is there anyone on amd64 who can build doc? If it won't work I will have to remove it.
Created attachment 109124 [details] my working copy of media-sound/lilypond/lilypond-2.10.15.ebuild it works for 2.10.14 (except doc) but now i can't even build 2.10.15 without doc.
Created attachment 109126 [details] lilypond-2.10.15/temp/build.log
Created attachment 109337 [details] media-sound/lilypond/lilypond-2.10.16.ebuild If mftrace had been bumped I'd committed this ebuild. Comments and testing very welcome.
(In reply to comment #62) > Created an attachment (id=109337) [edit] > media-sound/lilypond/lilypond-2.10.16.ebuild > > If mftrace had been bumped I'd committed this ebuild. Comments and testing very > welcome. OK: * LANGS is just for doc, so should be just fr * the dependency on tetex should be included * mftrace doesn't need to be bumped, you just need to use files/lilypond-2.10.8-version.patch. Omitting the patch will just lead to more problems later on when another similar dependency comes around. * the dependency on findutils should be included * the dependency on gcc 4.0 should be included * I know you haven't included the doc USE flag yet, but the dependency on || ( >=app-text/ghostscript-gnu-8.54 >=app-text/ghostscript-gpl-8.54 ) should be included * I think it would be safer to use the Ghostscript NSCB fonts rather than what fc-match reports * any reason you've left out the vim/emacs USE flags? * did you manage to build OK without the python path hack?
i installed 2.10.16 from your ebuild and mftrace from the ebuild in the respective bug. lilypond and lilypond-invoke-editor seem to work. all other executables give me an error: abc2ly, convert-ly, etf2ly, lilypond-book, and musicxml2ly complain about missing import lilylib: Traceback (most recent call last): File "/usr/bin/convert-ly", line 41, in ? import lilylib as ly ImportError: No module named lilylib midi2ly complains about missing import midi: Traceback (most recent call last): File "/usr/bin/midi2ly", line 51, in ? import midi ImportError: No module named midi i can fix the errors concerning the lilylib import by creating a symlink /usr/share/lilypond/current that points to /usr/share/lilypond/2.10.16. this doesn't fix the missing "midi" import, though.
sorry, forgot to mention: that is mftrace-1.2.9, of course.
(In reply to comment #64) > Traceback (most recent call last): > File "/usr/bin/convert-ly", line 41, in ? > import lilylib as ly > ImportError: No module named lilylib > > midi2ly complains about missing import midi: > > Traceback (most recent call last): > File "/usr/bin/midi2ly", line 51, in ? > import midi > ImportError: No module named midi > > > i can fix the errors concerning the lilylib import by creating a symlink > /usr/share/lilypond/current that points to /usr/share/lilypond/2.10.16. this > doesn't fix the missing "midi" import, though. It's failing to interpolate the directories lilypond_datadir, lilypond_libdir into python/relocate-preamble.py.in.
Comment on attachment 105653 [details, diff] files/lilypond-2.10.8-version.patch Verified still needed
(In reply to comment #63) > * did you manage to build OK without the python path hack? Vios for me without the python path hack.
Created attachment 109350 [details, diff] files/lilypond-2.10.16-interpolate-sed.patch Try this.
(In reply to comment #63) > * LANGS is just for doc, so should be just fr as long as doc doesn't work it will remain commented. There does seem to be some support for more than just French. > * the dependency on tetex should be included > * the dependency on findutils should be included probably, but I haven't tested this yet. > * mftrace doesn't need to be bumped, you just need to use > files/lilypond-2.10.8-version.patch. Omitting the patch will just lead to more > problems later on when another similar dependency comes around. I think it is simpler just to bump mftrace for now. I'm also a little disappointed that your bug hasn't gotten a response yet. Perhaps you should try the mailing list too. > * the dependency on gcc 4.0 should be included I can't depend on gcc-4, cause that doesn't mean it will get build wih gcc-4. > * I know you haven't included the doc USE flag yet, but the dependency on || ( > >=app-text/ghostscript-gnu-8.54 >=app-text/ghostscript-gpl-8.54 ) should be > included will fix when doc works > * I think it would be safer to use the Ghostscript NSCB fonts rather than what > fc-match reports You mean "--with-ncsb-dir=/usr/share/fonts/default/ghostscript"? It seemed to make a difference sometimes when making doc, but then the next time it didn't. Not really sure what it does. > * any reason you've left out the vim/emacs USE flags? The emacs stuff didn't look like it did the right thing and with doc not working I haven't looked at it more closely. > * did you manage to build OK without the python path hack? Yes, I've posted to the list and apparently there was some bug which they then supposedly fixed. http://lists.gnu.org/archive/html/lilypond-devel/2007-01/msg00451.html
(In reply to comment #70) > (In reply to comment #63) > > * LANGS is just for doc, so should be just fr > as long as doc doesn't work it will remain commented. There does seem to be > some support for more than just French. The other language support is for interface pofiles; we install all of those regardless of LANGS. > > * mftrace doesn't need to be bumped, you just need to use > > files/lilypond-2.10.8-version.patch. Omitting the patch will just lead to more > > problems later on when another similar dependency comes around. > I think it is simpler just to bump mftrace for now. I'm also a little > disappointed that your bug hasn't gotten a response yet. Perhaps you should > try the mailing list too. Yeah, will do. > > * the dependency on gcc 4.0 should be included > I can't depend on gcc-4, cause that doesn't mean it will get build wih gcc-4. I think the right thing to do then is to choke in pkg_setup, or better find a way to pass g++-4.x to the build process. > > * I know you haven't included the doc USE flag yet, but the dependency on > > || ( >=app-text/ghostscript-gnu-8.54 >=app-text/ghostscript-gpl-8.54 ) > > should be included > will fix when doc works OK. > > * I think it would be safer to use the Ghostscript NSCB fonts rather than > > what fc-match reports > You mean "--with-ncsb-dir=/usr/share/fonts/default/ghostscript"? It seemed to > make a difference sometimes when making doc, but then the next time it didn't. > Not really sure what it does. It lets us control which font files lilypond uses for the NSCB fonts. The ghostscript fonts are known good, whereas some other font packages supply Century Schoolbook faces (e.g. urw-fonts) which might not work properly. > > * any reason you've left out the vim/emacs USE flags? > The emacs stuff didn't look like it did the right thing and with doc not > working I haven't looked at it more closely. Policy differs on whether to remove support files like those. It'd probably be simpler to leave off the USE flags and just let the syntax files be installed; they don't do any harm. That said, the vimfiles= on make install is needed. > > * did you manage to build OK without the python path hack? > Yes, I've posted to the list and apparently there was some bug which they then > supposedly fixed. > http://lists.gnu.org/archive/html/lilypond-devel/2007-01/msg00451.html Hm. I'm not so sure. It seemed to vio for me with that stanza removed. I'm testing the interpolation patch for now (will post to list) but when that's done I'll try again.
(In reply to comment #71) > It seemed to vio for me with that stanza removed. heh, you said vio again. Please inform me of its meaning.
Created attachment 109363 [details] media-sound/lilypond/lilypond-2.10.16.ebuild added gcc-4 detection code
Looks like none of the scripts, like convert-ly, work, out of the box. The build time and/or config-fed variables, @lilypond_datadir@ and @lilypond_libdir@ are not being populated. It's easily worked around, but still.... I'll try to get to a patch today.
(In reply to comment #72) > (In reply to comment #71) > > It seemed to vio for me with that stanza removed. > heh, you said vio again. Please inform me of its meaning. Sandbox violation. And they are happening for me, unfortunately. I'll try to work out why. (In reply to comment #74) > Looks like none of the scripts, like convert-ly, work, out of the box. The > build time and/or config-fed variables, @lilypond_datadir@ and > @lilypond_libdir@ are not being populated. It's easily worked around, but > still.... I'll try to get to a patch today. See above, comment 69. The patch has been committed, though I don't know whether that's to branch or just trunk; if the latter we'll need to keep the patch till it reaches branch. (In reply to comment #71) > (In reply to comment #70) > > I think it is simpler just to bump mftrace for now. I'm also a little > > disappointed that your bug hasn't gotten a response yet. Perhaps you should > > try the mailing list too. > Yeah, will do. OK, turns out the patch had been applied to trunk but not to branch; it's been applied on branch so we shouldn't need it from 2.10.17 on.
(In reply to comment #75) > (In reply to comment #72) > > (In reply to comment #71) > > > It seemed to vio for me with that stanza removed. > > heh, you said vio again. Please inform me of its meaning. > Sandbox violation. And they are happening for me, unfortunately. I'll try to > work out why. Worked it out. When --enable-documentation is given (indeed, when --disable-documentation is /not/ given), the build runs help2man, a little perl script that generates man pages from --help and --version output. Thus help2man actually runs the various lilypond programs/scripts, and does so before relocation is set up properly, so they see the installed lilypond python modules from earlier installs and try to recompile them. This should be fixable, I hope.
Created attachment 109587 [details] lilypond-2.10.16.ebuild 2.10.16 with documentation build This fixes two sandboxing bugs when building docs.
Created attachment 109854 [details] media-sound/lilypond/lilypond-2.10.16.ebuild missing eclass prevented previous version from passing gcc test, sorry about that.
(In reply to comment #63) > * the dependency on findutils should be included actually findutils is a system package, so we shouldn't depend on it.
*** Bug 167055 has been marked as a duplicate of this bug. ***
I'm getting the well-known sandbox errors with 2.10.17 unfortunately.
OK, 2.10.17 includes the version detection patch, so we don't need the bogus mftrace dependency any more. I'm building it with my ebuild now, to see whether my fixes can cope with the sandbox errors. Might not be able to report back till the end of the weekend, though.
Created attachment 110565 [details] lilypond-2.10.17.ebuild Installed fine w/o any sandbox vios. Marijn, you might want to see whether any of my fixes work for your problems.
Testing 2.10.19 from upstream
(In reply to comment #84) > Testing 2.10.19 from upstream > Looks OK so far. I don't have a lot of lilypond test cases yet, though.
(In reply to comment #85) > (In reply to comment #84) > > Testing 2.10.19 from upstream > > > Looks OK so far. I don't have a lot of lilypond test cases yet, though. > Take a look at Mutopia (http://www.mutopiaproject.org/).
Be careful when selecting test material from Mutopia. Most of the examples there are quite old, and won't work without much hackery. Lilypond itself includes many regression tests which are exercised whenever you build documentation. So, if you succeed in building the docs, then you shouldn't need more tests. I suppose you may want to look at the resulting PDFs to make sure nothing looks amiss.
A few more things to clean out of Ed's ebuild: - Remove sys-devel/make from DEPEND; it's a system package - Remove sys-devel/gettext from DEPEND; it's a system package - Remove dev-lang/perl from DEPEND; it's a system package - Remove the "doc" clause from DEPEND since it's no longer in IUSE - Remove "if has_version =guile-1.8*" test since we RDEPEND on guile-1.8 already
(In reply to comment #88) > A few more things to clean out of Ed's ebuild: Mine, or Marijn's? Both, I guess. > - Remove sys-devel/make from DEPEND; it's a system package > - Remove dev-lang/perl from DEPEND; it's a system package Yes. > - Remove sys-devel/gettext from DEPEND; it's a system package No, it's not. Flex is, though; so is bison, but we need the version dep. > - Remove the "doc" clause from DEPEND since it's no longer in IUSE I'm using that. > - Remove "if has_version =guile-1.8*" test since we RDEPEND on guile-1.8 > already Yeah, I'll drop the lower guile dep on mine; guile 1.8 seems to be working now.
Created attachment 111104 [details] lilypond-2.10.19.ebuild Fixes per previous comment.
The sandbox violations are still there with 2.1[01].20. I've added guile-1.8.1-r2 with the rational patch which 2.11.x needs so give it a spin if you want.
(In reply to comment #91) > The sandbox violations are still there with 2.1[01].20. I've added > guile-1.8.1-r2 with the rational patch which 2.11.x needs so give it a spin if > you want. 2.10.20 builds fine with docs for me first time through, using my ebuild (straight bump from 2.10.19). I'll give it another run with 2.10.20 installed and then start looking at 2.11.x.
the old sys.path.insert trick doesn't seem to work anymore.
(In reply to comment #93) > the old sys.path.insert trick doesn't seem to work anymore. That hasn't worked (or been needed) since 2.10.13. The solution I use is with LILYPOND_DATADIR=${S}/out/share/lilypond/current.
Created attachment 115140 [details] lilypond-2.11.21.ebuild building docs fails
Created attachment 115142 [details] lilypond-2.11.21/temp/build.log
(In reply to comment #96) > Created an attachment (id=115142) [edit] > lilypond-2.11.21/temp/build.log There's nothing there - did you mean to attach a different file?
Yes, I did. Due to the log being 6 times as large the maximum allowed non-patch attachment, I've put it into my dev-space: http://dev.gentoo.org/~hkbst/lilypond-2.11.21-build.log Thanks Ed
So, for two of those you're getting "/undefined in ttcf" (seems to be something to do with the MS-Gothic font?) and the other is failing because you don't have inkscape: Layout output to `lily-7be01a97b8-1.svg'...[/var/tmp/portage/media-sound/lilypond-2.11.21/work/lilypond-2.11.21/out/bin/../share/lilypond/current/fonts/svg/aybabtu.svg][/var/tmp/portage/media-sound/lilypond-2.11.21/work/lilypond-2.11.21/out/bin/../share/lilypond/current/fonts/svg/emmentaler-11.svg][/var/tmp/portage/media-sound/lilypond-2.11.21/work/lilypond-2.11.21/out/bin/../share/lilypond/current/fonts/svg/emmentaler-14.svg][/var/tmp/portage/media-sound/lilypond-2.11.21/work/lilypond-2.11.21/out/bin/../share/lilypond/current/fonts/svg/emmentaler-20.svg]] Invoking `LD_LIBRARY_PATH= inkscape -T -E lily-7be01a97b8-inkscape.eps lily-7be01a97b8-1.svg'...sh: inkscape: command not found `LD_LIBRARY_PATH= inkscape -T -E lily-7be01a97b8-inkscape.eps lily-7be01a97b8-1.svg' failed (32512) (...) Layout output to `lily-f9b2792266-1.eps'... Layout output to `lily-f9b2792266-2.eps'... Converting to `lily-f9b2792266-1.pdf'... Invoking `gs -dNOSAFER -dEPSCrop -dCompatibilityLevel=1.4 -dNOPAUSE -dBATCH -r1200 -sDEVICE=pdfwrite -sOutputFile="lily-f9b2792266-1.pdf" -c .setpdfwrite -f "lily-f9b2792266-1.eps"'...GPL Ghostscript 8.54 (2006-05-17) Copyright (C) 2006 artofcode LLC, Benicia, CA. All rights reserved. This software comes with NO WARRANTY: see the file PUBLIC for details. Error: /undefined in ttcf Operand stack: --nostringval-- Execution stack: %interp_exit .runexec2 --nostringval-- --nostringval-- --nostringval-- 2 %stopped_push --nostringval-- --nostringval-- --nostringval-- false 1 %stopped_push 1 3 %oparray_pop 1 3 %oparray_pop --nostringval-- 1 3 %oparray_pop 1 3 %oparray_pop .runexec2 --nostringval-- --nostringval-- --nostringval-- 2 %stopped_push --nostringval-- --nostringval-- --nostringval-- %loop_continue --nostringval-- 1 --nostringval-- --nostringval-- false 1 %stopped_push --nostringval-- false 1 %stopped_push --nostringval-- Dictionary stack: --dict:1126/1686(ro)(G)-- --dict:0/20(G)-- --dict:74/200(L)-- --dict:74/200(L)-- Current allocation mode is local Current file position is 5 GPL Ghostscript 8.54: Unrecoverable error, exit code 1 `gs -dNOSAFER -dEPSCrop -dCompatibilityLevel=1.4 -dNOPAUSE -dBATCH -r1200 -sDEVICE=pdfwrite -sOutputFile="lily-f9b2792266-1.pdf" -c .setpdfwrite -f "lily-f9b2792266-1.eps"' failed (256) (...) error: failed files: "lily-f9b2792266 lily-1c0265f933 lily-7be01a97b8"
lilypond-2.11.21 fails to compile on ppc: chmod 755 out/midi2ly /usr/bin/perl /var/tmp/paludis/media-sound/lilypond-2.11.21/work/lilypond-2.11.2 1/buildscripts/out/help2man out/midi2ly > out/midi2ly.1 cat lilypond-invoke-editor.scm | sed -e '#' -e '/@relocate-preamble@/r /var/tmp /paludis/media-sound/lilypond-2.11.21/work/lilypond-2.11.21/python/./out/relocat e-preamble.py' -e 's%@relocate-preamble@%%g' | sed -e '#' -e 's!@BASH@!/bin/sh! g' -e 's!@BUILD_VERSION@!1!g' -e 's!@DATE@!19APR07!g' -e 's!@FONTFORGE@!/usr/ bin/fontforge!g' -e 's!@GUILE@!/usr/bin/guile!g' -e 's!@MICRO_VERSION@!21!g' -e 's!@MAJOR_VERSION@!2!g' -e 's!@MINOR_VERSION@!11!g' -e 's!@PACKAGE@!LILYPON D!g' -e 's!@PATCH_LEVEL@!21!g' -e 's!@PATHSEP@!:!g' -e 's!@PERL@!/usr/bin/per l!g' -e 's!@PYTHON@!/usr/bin/python!g' -e 's!@SHELL@!/bin/sh!g' -e 's!@TARGET _PYTHON@!/usr/bin/python!g' -e 's!@TOPLEVEL_VERSION@!2.11.21!g' -e 's!@bindir@ !/usr/bin!g' -e 's!@datadir@!/usr/share!g' -e 's!@date@!19APR07!g' -e 's!@lil ypond_datadir@!/usr/share/lilypond/2.11.21!g' -e 's!@lilypond_docdir@!/usr/shar e/doc/lilypond!g' -e 's!@lilypond_libdir@!/usr/lib/lilypond/2.11.21!g' -e 's!@ local_lilypond_datadir@!/usr/share/lilypond/2.11.21!g' -e 's!@local_lilypond_li bdir@!/usr/lib/lilypond/2.11.21!g' -e 's!@localedir@!/usr/share/locale!g' -e ' s!@outdir@!./out!g' -e 's!@package@!lilypond!g' -e 's!@prefix@!/usr!g' -e 's! @program_prefix@!!g' -e 's!@program_suffix@!!g' -e 's!@sharedstatedir@!/usr/co m!g' > out/lilypond-invoke-editor chmod 755 out/lilypond-invoke-editor /usr/bin/perl /var/tmp/paludis/media-sound/lilypond-2.11.21/work/lilypond-2.11.2 1/buildscripts/out/help2man out/lilypond-invoke-editor > out/lilypond-invoke-edi tor.1 help2man: can't get `--help' info from out/lilypond-invoke-editor make[1]: *** [out/lilypond-invoke-editor.1] Error 1 make[1]: Leaving directory `/var/tmp/paludis/media-sound/lilypond-2.11.21/work/l ilypond-2.11.21/scripts' make: *** [all] Error 2 !!! ERROR in media-sound/lilypond-2.11.21: !!! In src_compile at line 3725 !!! emake failed !!! Call stack: !!! * src_compile (/var/tmp/paludis/media-sound/lilypond-2.11.21/temp/loadsav eenv:3725) !!! * ebuild_f_compile (/usr/libexec/paludis/src_compile.bash:51) !!! * ebuild_main (/usr/libexec/paludis/ebuild.bash:354) !!! * main (/usr/libexec/paludis/ebuild.bash:372) die trap: exiting with error. Install error: * In program paludis -i -1 lilypond: * When performing install action from command line: * When executing install task: * When installing 'media-sound/lilypond-2.11.21': * Install error: Install failed for 'media-sound/lilypond-2.11.21::local' Any ideas?
There is a problem with building both recent ebuilds 2.10.19 and 2.11.21 with media-gfx/fontforge-20070501. It seems that recent version of fontforge lacks the function RemoveATT. Downgrading to fontforge 20060703-r1 solves the problem. > out/feta-braces-i.otf-table /usr/bin/fontforge -script ../buildscripts/pfx2ttf.fontforge /usr/share/fonts/default/ghostscript/c059016l.pfb /usr/share/fonts/default/ghostscript/c059016l.afm ./out/ && /usr/bin/fontforge -script ../buildscripts/pfx2ttf.fontforge /usr/share/fonts/default/ghostscript/c059036l.pfb /usr/share/fonts/default/ghostscript/c059036l.afm ./out/ && /usr/bin/fontforge -script ../buildscripts/pfx2ttf.fontforge /usr/share/fonts/default/ghostscript/c059033l.pfb /usr/share/fonts/default/ghostscript/c059033l.afm ./out/ && /usr/bin/fontforge -script ../buildscripts/pfx2ttf.fontforge /usr/share/fonts/default/ghostscript/c059013l.pfb /usr/share/fonts/default/ghostscript/c059013l.afm ./out/ && true Copyright (c) 2000-2007 by George Williams. Executable based on sources from 13:10 1-May-2007-ML. RemoveATT: This scripting function no longer works. Try RemoveLookupSubtable or RemovePosSub.Called from... ../buildscripts/pfx2ttf.fontforge: line 23make[1]: *** [out/CenturySchL-Ital.otf] Error 1 make[1]: Leaving directory `/mnt/rh/gentoo/build/portage/media-sound/lilypond-2.11.21/work/lilypond-2.11.21/mf' make: *** [all] Error 2 !!! ERROR: media-sound/lilypond-2.11.21 failed. Call stack: ebuild.sh, line 1614: Called dyn_compile ebuild.sh, line 971: Called qa_call 'src_compile' environment, line 3801: Called src_compile lilypond-2.11.21.ebuild, line 70: Called die
Created attachment 118770 [details] lilypond-2.10.23.ebuild I've ripped out the doc building as I don't know how to do it in a sane way without causing sandbox violations. Please test/review. I hope to commit it in a few days.
committed as promised: fixed.