%define festivalversion 1.96 Name: festival Summary: A free speech synthesis and text-to-speech system Version: %{festivalversion} Release: 1.4%{?dist} URL: http://www.cstr.ed.ac.uk/projects/festival/ Group: Applications/Multimedia License: MIT-style # The main festival source %define baseURL http://festvox.org/packed/festival/%{festivalversion} Source0: %{baseURL}/festival-%{festivalversion}-beta.tar.gz # Fedora package notes Source10: festival-1.96-README-fedora # Our local site config files. Source50: festival-1.96-0.7-fedora-siteinit.scm Source51: festival-1.96-0.7-fedora-sitevars.scm # Set defaults to American English instead of British English - the OALD # dictionary (free for non-commercial use only) is needed for BE support # Additionally, prefer the smaller (and I think nicer sounding) nitech hts # voices. Patch1: festival-1.96-nitech-american.patch # Whack some buildroot references Patch2: festival_buildroot.patch # We don't want ESD to be the default, since it's not always # available. Patch4: festival-1.96-lower-esd-priority.patch # Look for siteinit and sitevars in /etc/festival Patch8: festival-1.96-etcsiteinit.patch # Alias old cmu names to new nitech ones Patch9: festival-1.96-alias_cmu_to_nitech.patch # Look for system-installed speech tools, not sitting next to us in the # build area. An improvement to be made: this should use INCDIR instead of # hard-coding /usr/include. Patch10: festival-1.96-find-system-speech-tools.patch # Build main library as shared, not just speech-tools Patch11: festival-1.96-main-shared-build.patch # Don't use the speech-tools libs as actual makefile dependencies. # There's really no need. Instead, just list the libs to use. Patch12: festival-1.96-trust-us-on-the-buildreqs.patch # This makes festival use /usr/lib[arch]/festival/etc for its # arch-specific "etc-path", rather than /usr/share/festival/etc/system_type. # Then I use sed to replace the token with actual arch-specific libdir. # A better way would be to actually make this a flexible makefile parameter, # but that's something to take up with upstream. Patch31: festival-1.96-kludge-etcpath-into-libarch.patch BuildRequires: speech-tools-devel = 1.2.96 BuildRequires: ncurses-devel esound-devel Requires: festival-system Requires: festival-voice # This is hard-coded as a requirement because it's the smallest voice (and, # subjectively I think the most pleasant to listen to and so a good # default). # # Ideally, this would be a "suggests" instead of a hard requirement. # # Update: with the new nitech versions of the voices, slt-arctic is no # longer the smallest. But... AWB has a strong scottish accent, and JMK a # kind of odd canadian one, so they're not great candidates for inclusion. # And I find RMS a bit hard to understand. BDL isn't much smaller than SLT, # and since I like it better, I think I'm going to keep it as the default # for a price 12k. So, in case anyone later questions why this is the # default, there's the answer. :) Requires: festvox-slt-arctic-hts BuildRoot: %(mktemp -ud %{_tmppath}/%{name}-%{festivalversion}-%{release}-XXXXXX) %package system Summary: The festival command-line tool and its infrastructure Group: Applications/Multimedia %package lib Summary: The shared library for the Festival speech synthesis system # this is here to make sure upgrades go cleanly. In other cases, # the auto-deps should handle this just fine. Requires: speech-tools Group: System Environment/Libraries Requires(post): /sbin/ldconfig Requires(postun): /sbin/ldconfig %package devel Summary: Development files for the Festival speech synthesis system Version: %{festivalversion} Group: Development/Libraries # Note: rpmlint complains incorrectly about # "no-dependency-on festival" Requires: speech-tools-devel >= 1.2.96 Requires: festival-lib %description Festival is a general multi-lingual speech synthesis system developed at CSTR. It offers a full text to speech system with various APIs, as well as an environment for development and research of speech synthesis techniques. It is written in C++ with a Scheme-based command interpreter for general control. %description system The festival command-line tool and its basic infrastructure, including utilities, the required Scheme libraries, configuration files, and so on. You will also need to install at least one voice in order to actually generate speech. %description lib The shared library used by the Festival text-to-speech and speech synthesis system. %description devel Development files for the Festival speech synthesis system. Install festival-devel if you want to use Festival's capabilities from within your own programs, or if you intend to compile other programs using it. Note that you can also interface with Festival in via the shell or with BSD sockets. %prep %setup -q -n festival %patch1 -p1 -b .nitech %patch2 -p1 -b .buildrootrefs %patch4 -p1 -b .esd %patch8 -p1 -b .etc %patch9 -p1 -b .cmu2nitech # patch9 creates a new file; patch helpfully makes a "backup" of the # non-existent "original", which then has bad permissions. zap. rm -f lib/alias_cmu_to_nitech.scm.cmu2nitech %patch10 -p1 -b .findspeechtools %patch11 -p1 -b .shared %patch12 -p1 -b .trustus %patch31 -p1 -b .libarch # finish the kludge for arch-specific "etc" (misc. binaries) for f in src/arch/festival/festival.cc; do sed -i -e 's,{{HORRIBLELIBARCHKLUDGE}},"%{_libdir}",' $f done # the Fedora readme cp %{SOURCE10} README.fedora %build # build the main program export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$(pwd)/src/lib # instead of doing this, maybe we should patch the make process # so it looks in the right place explicitly: export PATH=$(pwd)/bin:$PATH %configure # -fno-shared-data is used 'cause the upstream makefile does. # -fPIC 'cause we're building shared libraries and it doesn't hurt. # And make -w because debugging this setup is incredibly crazy without. make -w \ FTLIBDIR="%{_datadir}/festival/lib" \ CFLAGS="$RPM_OPT_FLAGS -fPIC -fno-shared-data" \ CXXFLAGS="$RPM_OPT_FLAGS -fPIC -fno-shared-data" %install # "make install" for this package is, um, "interesting". It seems geared for # local user-level builds. So, rather than doing that and then patching it # up, do the right parts by hand as necessary. mkdir -p $RPM_BUILD_ROOT%{_datadir}/festival/lib # binaries: make INSTALLED_BIN=$RPM_BUILD_ROOT%{_bindir} make_installed_bin_static install -m 755 bin/text2wave $RPM_BUILD_ROOT%{_bindir} # install the shared library mkdir -p $RPM_BUILD_ROOT%{_libdir} cp -a src/lib/libFestival.so* $RPM_BUILD_ROOT%{_libdir} # this is just nifty. and it's small. install -m 755 examples/saytime $RPM_BUILD_ROOT%{_bindir} # man pages mkdir -p $RPM_BUILD_ROOT%{_mandir}/man1 cp -a doc/*.1 $RPM_BUILD_ROOT%{_mandir}/man1 # lib: the bulk of the program -- the scheme stuff and so on pushd lib mkdir -p $RPM_BUILD_ROOT%{_datadir}/festival/lib for f in *.scm festival.el *.ent *.gram *.dtd *.ngrambin speech.properties ; do install -m 644 $f $RPM_BUILD_ROOT%{_datadir}/festival/lib/ done mkdir -p $RPM_BUILD_ROOT%{_datadir}/festival/lib/multisyn/ install -m 644 multisyn/*.scm $RPM_BUILD_ROOT%{_datadir}/festival/lib/multisyn/ popd # "etc" -- not in the configuration sense, but in the sense of "extra helper # binaries". pushd lib/etc # not arch-specific mkdir -p $RPM_BUILD_ROOT%{_datadir}/festival/lib/etc install -m 755 email_filter $RPM_BUILD_ROOT%{_datadir}/festival/lib/etc # arch-specific mkdir -p $RPM_BUILD_ROOT%{_libdir}/festival/etc install -m 755 */audsp $RPM_BUILD_ROOT%{_libdir}/festival/etc popd # the actual /etc. :) mkdir -p $RPM_BUILD_ROOT%{_sysconfdir}/festival # use our version of this file rm $RPM_BUILD_ROOT%{_datadir}/festival/lib/siteinit.scm install -m 644 %{SOURCE50} $RPM_BUILD_ROOT%{_sysconfdir}/festival/siteinit.scm install -m 644 %{SOURCE51} $RPM_BUILD_ROOT%{_sysconfdir}/festival/sitevars.scm # copy in the intro.text. It's small and makes (intro) work. in the future, # we may want include more examples in an examples subpackage mkdir -p $RPM_BUILD_ROOT%{_datadir}/festival/examples/ install -m 644 examples/intro.text $RPM_BUILD_ROOT%{_datadir}/festival/examples # header files mkdir -p $RPM_BUILD_ROOT%{_includedir}/festival cp -a src/include/* $RPM_BUILD_ROOT%{_includedir}/festival %clean rm -rf $RPM_BUILD_ROOT %post lib -p /sbin/ldconfig %postun lib -p /sbin/ldconfig %files %defattr(-,root,root) %doc README.fedora %files system %defattr(-,root,root) %doc ACKNOWLEDGMENTS COPYING NEWS README %dir %{_sysconfdir}/festival %config(noreplace) %{_sysconfdir}/festival/siteinit.scm %config(noreplace) %{_sysconfdir}/festival/sitevars.scm %{_bindir}/festival %{_bindir}/festival_client %{_bindir}/festival_server %{_bindir}/festival_server_control %{_bindir}/text2wave %{_bindir}/saytime %dir %{_datadir}/festival %dir %{_datadir}/festival/lib %{_datadir}/festival/lib/*.scm %{_datadir}/festival/lib/festival.el %{_datadir}/festival/lib/*.ent %{_datadir}/festival/lib/*.gram %{_datadir}/festival/lib/*.dtd %{_datadir}/festival/lib/*.ngrambin %{_datadir}/festival/lib/speech.properties %{_datadir}/festival/lib/etc %dir %{_datadir}/festival/lib/multisyn %{_datadir}/festival/lib/multisyn/*.scm %dir %{_datadir}/festival/examples %{_datadir}/festival/examples/intro.text %dir %{_libdir}/festival %dir %{_libdir}/festival/etc %{_libdir}/festival/etc/* %{_mandir}/man1/* %files lib %defattr(-,root,root) %doc COPYING %{_libdir}/libFestival.so.* %files devel %defattr(-,root,root) %doc COPYING %{_libdir}/libFestival.so %dir %{_includedir}/festival %{_includedir}/festival/* %changelog * Mon Jun 25 2007 Matthew Miller 1.96-1.4 - move voices to their own packages * Mon Jun 25 2007 Matthew Miller 1.96-1.3 - Strictly speaking, festival doesn't require any voices to run; that's just if you want to actually use it for its main purpose. Making voices be a requirement is useful in that you can yum install festival and have it all work -- but unfortunately, introduces a buildreq loop with festlex-cmu. Therefore, breaking this loop by making an overall "festival" package which contains meta-requires, and moving the bulk of the package to "festival-system". This may be slightly over-engineered, but having festlex-cmu a separate package has enough benefits that I want to try it. (One large benefit is that we no longer have to deal with the weirdness that comes from having multiple Versions in one package.) * Wed Apr 4 2007 Matthew Miller 1.96-1.2 - split docs into subpackage. - split dictionaries into subpackages -- particularly useful for non-english distributions, once we actually have some non-english voices. - make voices provide festival-voice-male or festival-voice-female * Thu Mar 29 2007 Matthew Miller 1.96-1.1 - split speech tools into its own source package. - move these patches to the speech-tools package: # festival-1.96-bettersonamehack.patch # festival-1.96-speechtools-1.2.96-beta+awb.patch # festival-1.96-speechtools-linklibswithotherlibs.patch # festival-1.96-speechtools-ohjeezcxxisnotgcc.patch # festival-1.96-speechtools-rateconvtrivialbug.patch # festival-1.96-speechtools-shared-build.patch - these patches applied partly to speech_tools but also to festival proper: # festival-1.96-kludge-etcpath-into-libarch.patch # festival-1.96-main-shared-build.patch # festival-1.96-speechtools-buildesdmodule.patch (renamed to festival-1.96-lower-esd-priority.patch) # festival_buildroot.patch (note this one still needs to be modernized) - numbering error in setup macros was making them only work by coincidence. fixed. :) - update the find-speechtools patch to look for the separate stuff from speech-tools-devel - add festival-1.96-trust-us-on-the-buildreqs.patch to help make build with system-wide speech-tools libs. - libFestival wasn't getting linked against libest*. Oops! (That's in the previous package release too -- not a side-effect of splitting packages.) * Tue Mar 20 2007 Ray Strode 1.96-1 - rebuild * Mon Mar 19 2007 David Zeuthen 1.96-0.11 - Forgot to add the .scm files * Mon Mar 19 2007 David Zeuthen 1.96-0.10 - Update to Matthew Miller's much improved package (#232105) - Move the buildroot patch around * Sun Mar 18 2007 Matthew Miller 1.96-0.9 - fix the library link patch to use -lncurses instead of -ltinfo -- the later is all that's really needed, but the former works on older distros too. * Fri Mar 16 2007 Matthew Miller 1.96-0.8 - festival-devel depends on the libraries package, not base festival. this raises an multilib question: need to obsolete festival.i386 on x86_64. Right now, there's no mechanism for doing that. Fortunately, all the changes in packaging happen to make it so that the current version doesn't conflict with the old release, so one will get unused cruft but not breakage when upgrading. - Bite teh proverbial bullet and make libFestival build shared. - update speech-tools soname patch to work in the more general case needed by the festival main build - make said shared-lib a subpackage to avoid multiarching the whole thing - split festival-devel and speechutils-devel in anticipation of future plan of actually decoupling these packages. - note that rpmlint complains about "missing" deps on the devel packages. it should be fixed to recognize requiring a -lib/libs package is sufficent or better. - add saytime script. Because, really, what else is this package *for*? - add the intro.text so (intro) works. 196 more bytes won't kill us. :) - remove $PATH from LD_LIBRARY_PATH used in build. (What the heck?) - add defattr to all subpackages. I don't think it's strictly necessary since putting it in the first package seems sufficient, but that's probably not behavior to count on. - make descriptions and summaries use more consistant language * Thu Mar 15 2007 Matthew Miller 1.96-0.7 - Upstream baseurl now includes version. (Thanks Alan Black @ cmu) - Update siteinit patch to also incorporate sitevars - Add sitevars as a config file - Ship our own siteinit and sitevars as sources - In default sitevars, reference /usr/local/share/festival/lib as another place to look for voices (it's okay if that doesn't exist). Hopefully, this will encourage people who want to install non-RPM-packaged voices to keep from doing it in /usr/share. - Fix wrong references to slt voice in other nitech voices - Fix wrongly commented-out (require 'f2bf0lr) in awb, clb, and rms voices. - Stop untarring source files and use the setup macro properly. - Get rid of silly DATA.TMP directories for installing voices and dictionaries. - Stop making ../speechtools link. Currently solved by patching to look in the current directory; could also do this by moving everything up a directory. - TODO: festival-buildroot.patch could stand to be updated. May not even be needed anymore. - Drop the 8k versions of the diphone voices, since there's not really any point. If you want smaller, use one of the arctic_hts voices instead. And overall, this saves us about 4.5M. * Wed Mar 14 2007 Matthew Miller 1.96-0.6 - Fix copy-paste error in JMK description (thanks Matthias Clasen) - Remove "nitech-us-" from the names of those voice packages to make the package names shorter. (This will also be more convenient if we switch to the cmu versions in the future.) - made aliases so old cmu_us_*_arctic_hts voice names still work. - Look for /etc/festival/siteinit.scm (and move siteinit.scm there!) - Mark siteinit.scm as a config file - Remove some non-useful stuff from speech-tools-utils. - Move main dir from /usr/share/festival to /usr/share/festival/lib at request of upstream. Also, we can drop the FHS (well, "fsstnd" -- it's old) patch and just pass FTLIBDIR to make. Which, hey, we were already doing. Yay redundancy. - clean up CFLAGS and CXXFLAGS. "-fpermissive" was hiding bad stuff. - update speech tools with patch from AWB to fix 64-bit build issue with EST_DProbDist - there's still some compiler warnings which should be addressed upstream. - The nitech hts voices don't properly proclaim_voice, making them not show up for gnome-speech and thus making orca crash. See details in the comments in bug #232105. * Tue Mar 13 2007 Matthew Miller 1.96-0.5 - use festvox- instead of festival-voice for voice packages -- matches upstream tarballs, and is shorter. Also, use shorter form of the date-based version. - get the README.htsvoice from the nitech voices -- it contains license info. - build (but don't enable by default) ESD support in speech-tools (bug #198908) - fix coding error noted in bug #162137 -- need to push this upstream. - link speech tools libraries with -lm, -ltermcap, -lesd and with themselves (bug #198190, partially) - holy sheesh. Use g++ for CXX, not gcc. Fixes bug #198190 completely. * Tue Mar 13 2007 Matthew Miller 1.96-0.4 - subpackages! Split out speech-tools, docs, voices. - long descriptions for the individual voices, carefully gathered from readmes and web sites. - ooh. replace cmu_hts voices with the updated versions from upstream . Two new voices, and updated versions of the others. (The voices packaged at festvox.org are apparently based on older versions of these, which in turn are from the CMU upstream.) - TODO: make aliases for the cmu voices. - arguably, voices should be made in to their own src.rpms. They don't need anything from here to build. That's for a future version. (At that time, the gigantic multisyn voices could be added.) The CMU dict needs festival installed to build, but I don't think it needs the source, so dicts could be subpackages too. And the docs are also a good candidate for separation. speech-tools, though, is incestuously used in the festival build process and I think it makes sense to keep that bundled. - TODO: check through the speechtools-utils for what should actually be packaged; fix the include path for siod (and anything else that needs it). - TODO: package festival.el so it just works with emacs. - TODO: reinvent festival_server_control as a proper init script - TODO: put the festival server in sbin, maybe? - Another question: should we drop the 8k diphone voices? Any point? - Changed "X11-like" to "MIT-style" (which is what X11 is) to make rpmlint happy. - make %{festivalversion} macro to deal with all of the changes to version in subpackages. Kludgy, but there's RPM for you. * Tue Mar 13 2007 Matthew Miller 1.96-0.3 - oh! The "etc-path" is important after all. Map that into /usr/lib[arch]/festival via a kludge. - make cmu_us_slt_arctic_hts the default voice, in preparation for splitting the voice packages. (thankfully, there's already a fallback mechanism -- cool!) * Mon Mar 12 2007 Matthew Miller 1.96-0.2 - clean up accidental backup file left in updated awb_arctic_hts tarball - remove /usr/share/festival/etc (see bug #228315) - move unpackage voices to the prep section where it belongs - other minor spec file readability changes - "make install" for this package is, um, "interesting". It seems geared for local user-level builds. So, rather than doing that and then patching it up, do the right parts by hand as necessary. (The previous version of the spec file did a convoluted mix of both.) - don't install static libs. - took out the massive hack that munges EST_*.h to speech_tools/EST_*.h in the installed header files -- programs should instead use -I/usr/include/speech_tools, shouldn't they? Put this back if I'm wrong. - TODO -- autogenerated speech_tools docs - festvox_ellpc11k.tar.gz, the spanish voice, wasn't getting installed anyway due to a license question. Since it's also gone upstream, removing. * Fri Mar 09 2007 Matthew Miller 1.96-0.1 - Preliminary update to 1.96 - Update to new cmu_us_*_arctic files -- they're changed upstream, although they don't appear to be versioned. Awesome. The current versions are those found in the same directory with the 1.96 files. - ditto festlex_CMU.tar.gz - add macro for speechtoolsversion - minor update to festival-1.96-american.patch. - update shared build patch and rename to make more obvious that it applies to the speechtools portion of the package. - gcc 4 build patches now upstream. - localhost-connections patch now upstream. - note that festvox_ellpc11k.tar.gz and festvox_kallpc8k.tar.gz are no longer in the directory tree upstream; drop? * Fri Jan 19 2007 Miroslav Lichvar - 1.95-6 - link with ncurses - add dist tag - make scriptlets safer * Wed Jul 12 2006 Jesse Keating - 1.95-5.2.1 - rebuild * Fri Feb 10 2006 Jesse Keating - 1.95-5.2 - bump again for double-long bug on ppc(64) * Tue Feb 07 2006 Jesse Keating - 1.95-5.1 - rebuilt for new gcc4.1 snapshot and glibc changes * Sun Jan 22 2006 Ray Strode - 1.95-5 - get gnopernicus working again. Patch from Fernando Herrera (bug 178312) - add a lot of compiler flags and random cruft to get festival to build with gcc 4.1 * Fri Dec 09 2005 Jesse Keating - rebuilt * Mon Oct 10 2005 Florian La Roche - another try to get it to compile again * Tue Apr 28 2005 - 1.95-3 - require info packages so the post does not fail - remove /usr/bin/VCLocalRule from buildroot since it is an extranious file that does not need to be installed * Wed Apr 27 2005 Miloslav Trmac - 1.95-2 - Fix build with gcc 4 (#156132) - Require /sbin/install-info for scriptlets (#155698) - Don't ship %{_bindir}/VCLocalRules (#75645) * Fri Feb 25 2005 - 1.95-1 - patch from Matthew Miller to update to 1.95. Full changelog below * Mon Feb 7 2005 Matthew Miller 1.95-0.mattdm8 - put speech-tools binaries in /usr/libexec/speech-tools so as to not clutter /usr/bin. Another approach would be to make speech-tools a separate package and to make these utilities a subpackage of that. - macro-ize /usr/bin, /usr/lib, /usr/include * Sun Feb 6 2005 Matthew Miller 1.95-0.mattdm6 - worked on this some more - made actually work -- put back rest of fsstnd patch which I had broken - made kludge for lack of sonames in shared libraries -- I think I did the right thing - put back american as the default -- british dicts are non-free. * Wed Jan 5 2005 Matthew Miller 1.95-0.mattdm1 - preliminary update to 1.95 beta - add really nice CMU_ARCTIC HTS voices, which is the whole point of wanting to do this. (They have a free license.) - switch to festvox.org north american upstream urls - keep old doc files -- there's no new ones yet. - add comment to specfile about reason for lack of OALD (British) voices -- they've got a more restrictive license. - change license to "X11-style", because that's how they describe it. - remove exclusivearch. I dunno if this builds on other archs, but I also don't know why it wouldn't. - fancier buildroot string, 'cause hey, why not. - more "datadir" macros - remove most of Patch0 (fsstnd) -- can be done by setting variables instead. there's some bits in speechtools still, though - update Patch3 (shared-build) - don't apply patches 20 and 21 -- no longer needed. - disable adding "FreeBSD" and "OpenBSD" to the dictionary for now. Probably a whole list of geek words should be added. Also, the patch was applied in an icky kludgy way. * Thu Jul 29 2004 Miloslav Trmac - 1.4.2-25 - Update for gcc 3.4 * Wed Jul 28 2004 Miloslav Trmac - 1.4.2-24 - Use shared libraries to reduce package size - Don't ship patch backup files * Tue Jun 15 2004 Elliot Lee - rebuilt * Tue May 4 2004 Jonathan Blandford 1.4.2-21 - Remove the spanish voices until we get clarification on the license * Sat Apr 10 2004 Warren Togami - BR libtermcap-devel #104722 * Fri Feb 13 2004 Elliot Lee - rebuilt * Mon Aug 25 2003 Bill Nottingham 1.4.2-19 - clean up buildroot references (#75643, #77908, #102985) - remove some extraneous scripts - fix build with gcc-3.3 * Thu Jun 12 2003 Elliot Lee 1.4.2-17 - Rebuild * Wed Jan 22 2003 Tim Powers - rebuilt * Thu Jan 9 2003 Tim Powers 1.4.2-15 - redirect install-info spewage * Tue Jan 7 2003 Jens Petersen 1.4.2-14 - put info files in infodir - add post and postun script to install and uninstall info dir file entry - drop postscript and info files from docs * Wed Dec 11 2002 Tim Powers 1.4.2-13 - rebuild * Thu Aug 15 2002 Bernhard Rosenkraenzer 1.4.2-12 - Adapt to current libstdc++ * Tue Jul 23 2002 Tim Powers 1.4.2-10 - build using gcc-3.2-0.1 * Wed Jul 3 2002 Trond Eivind Glomsrød 1.4.2-9 - Add some missing helpprograms (# 67698) * Fri Jun 21 2002 Tim Powers - automated rebuild * Mon Jun 10 2002 Trond Eivind Glomsrød 1.4.2-7 - Fix some rpmlint errors * Mon Jun 10 2002 Bernhard Rosenkraenzer 1.4.2-6 - Fix ISO C++ compliance * Mon Mar 18 2002 Tim Powers - rebuilt * Thu Mar 14 2002 Trond Eivind Glomsrød 1.4.2-2 - Get rid of CVS directiories in doc dir - Fix broken symlinks for components from speech_tools * Wed Mar 6 2002 Trond Eivind Glomsrød - 1.4.2 - Lots of fixes to make it build, more needed - Cleanups - Update URL - Fix docs inclusion - Drop prefix - Use %%{_tmppath} * Wed Aug 2 2000 Bernhard Rosenkraenzer - Add defattr (Bug #15033) * Tue Jul 25 2000 Bernhard Rosenkraenzer - fix build on current 7.0 * Mon Jul 17 2000 Bernhard Rosenkraenzer - fix build on current 7.0 * Thu Jul 6 2000 Bernhard Rosenkraenzer - FHSify * Mon Jun 19 2000 Bernhard Rosenkraenzer - Fix build on non-x86 * Sun Apr 22 2000 Bernhard Rosenkraenzer - initial packaging