Re: SPECS: privoxy.spec - up to 3.0.10 - not defined %_docdir issue
On Tuesday 14 of October 2008, 07:40, Jakub Bogusz wrote: On Tue, Oct 14, 2008 at 02:35:29AM +0200, Tomasz Pala wrote: On Mon, Oct 13, 2008 at 18:29:51 +0200, Jakub Bogusz wrote: 2. Why %doc macro puts documentation files in /usr/share/doc/%{name}-%{version} directory instead of /usr/share/doc/%{name} ? [...] Ability to install new version of package (usually library) without removing old one. This ability breaks well known convention that documentation should be placed in `${datarootdir}/doc/${PACKAGE}' (according to `info automake': 2.2.3 Standard Directory Variables) Some programs have internal help, howtos, manual, etc viewers. These viewers rely on ${datarootdir}/doc/${PACKAGE} documentation location. I'm not going to fix such every program because of broken PLD packaging scheme (even such fixes are trivial). There is other issue - %doc macro should only do one thing - mark files as documentation - unfortunately it does 3 things at once: marks files as docs (this is OK), compress documentation files (it is not always desirable) and copy files to %{_defaultdocdir}/%{name}-{version} (it is most confusing and harmful for me). IMHO %doc macro should accept some options for better controlling its behavior. After initial mail I've checked my system - none of multiple version libraries I have had installed ship %doc. Non-library packages would almost certainly conflict on other files. Currently on my system (conflicting packages skipped): libmpcdec-1.2.2-1.athlon libmpcdec-1.2.6-2.athlon libstdc++-4.1.2-1.athlon libstdc++-3.3.6-1.athlon -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: gksu.spec - findlang commented out
On Sunday 12 of October 2008, 23:09, Elan Ruusamäe wrote: On Sunday 12 October 2008 23:03, shadzik wrote: Author: shadzik Date: Sun Oct 12 20:03:45 2008 GMT Module: SPECS Tag: HEAD Log message: - findlang commented out ... @@ -83,12 +83,13 @@ rm -f $RPM_BUILD_ROOT%{_libdir}/nautilus/extensions-2.0/*.la -%find_lang %{name} +#%find_lang %{name} macro is still expanded, however as it's expanded in shell context, it's commented out for shell... ... and why have you removed %find_lang - it should be mentioned in your commitlog. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: privoxy.spec - up to 3.0.10 - not defined %_docdir issue
On Sunday 12 of October 2008, 19:15, twittner wrote: Author: twittner Date: Sun Oct 12 17:15:39 2008 GMT Module: SPECS Tag: HEAD [...] - rel. 0.1 - rpm -E '%_docdir' shows that this macro is not defined 1. Ditto: $ rpm -E '%_docdir' %_docdir $ q rpm-build-macros rpm-build-macros-0:1.478-1 noarch th-x86-64.pld-linux.org What should be fixed? 2. Why %doc macro puts documentation files in /usr/share/doc/%{name}-%{version} directory instead of /usr/share/doc/%{name} ? What we gain by adding -%{version} to directory name? (my answer: nothing but unnecessary troubles).Many programs expects that their documentation is located in /usr/share/doc/%{name} - privoxy, fluid from fltk-devel are good examples. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: PLD Live 2.0 beta3
On niedziela 31 sierpnia 2008, 18:17, Patryk Zawadzki wrote: Hi, qwiat did a hell of a good job resurrecting the PLD LiveCD idea. The ISO is available here: ftp://ep09.pld-linux.org/people/qwiat/ Please report all bugs to: https://bugs.launchpad.net/pld-linux Consider creating PLD-Live package (or sth like that) So far, I've spotted that file with MD5 sum is not suitable for md5sum -c processing, user live account is protected by password (`pld', but I had to get this by guesswork). Send kudos to qwiat. Looks and works nice. Maybe, in spite of it is beta release, it is worth mentioning on www.pld-linux.org -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: Build problem
On sobota 30 sierpnia 2008, 16:51, Zsolt Udvari wrote: Hi all! I've got a little(?) problem with building libtorrent-rasterbar. configure runs fine but the make will recive the next message: + /usr/bin/make Making all in include make[1]: Entering directory `/var/rpmbuild/BUILD/libtorrent-rasterbar-0.13.1/include' [...] `/var/rpmbuild/BUILD/libtorrent-rasterbar-0.13.1/src' make: *** [all-recursive] Error 1 error: Bad exit status from /root/tmp/rpm-tmp.25296 (%build) RPM build errors: Bad exit status from /root/tmp/rpm-tmp.25296 (%build) ^^^ Error: package build failed. (no more info) What's wrong? Building packages from root account (regardless of compilation errors). It's generally not good thing: potentially danger, can hide some errors. This is mentioned (among other useful information) in http://cvs.pld-linux.org/cgi-bin/cvsweb/PLD-doc/devel-hints-en.txt?rev=1.32 I've attached the spec file and a patch. Zsolt -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
[rc-scripts] Broken start-stop-daemon?
rc-scripts-0.4.1.24-1.i686 + /sbin/start-stop-daemon -q --start --nicelevel +0 -u postgres --exec /usr/bin/pg_ctl -- -s -w -D /var/lib/postgresql/8.3/main start pg_ctl: cannot be run as root Please log in (using, e.g., su) as the (unprivileged) user that will own the server process. + return 1 Downgrade to rc-scripts-0.4.1.23-2.i686 is a solution. Good night. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: cmake.spec - even more :/
On Thursday 31 of July 2008, 08:54, Elan Ruusamäe wrote: On Thursday 31 July 2008 07:45, qboosh wrote: +# - rpmldflags/rpmcppflags are not passed through %%cmake macro at all +# (is there any standard way???) there isn't (this has gone over devel-en list already at least twice). however i've tried to use in few places our standard, ie the ones we use in autoconf: # override __cmake to add -j4 in your ~/.rpmmacros for parallel make %__cmake/usr/bin/cmake %cmake { \ CC=%{__cc} \ CXX=%{__cxx} \ CFLAGS=%{rpmcflags} \ CXXFLAGS=%{rpmcxxflags} \ %{__cmake} \ } another place i recall, where these compiler options are not accepted is apache modules via apxs, they use cc and flags from apache compile time. The same story with ruby packages - most of them use ruby script which takes values of CC, CXX, C(XX)FLAGS from Config module (file /usr/lib/ruby/1.8/*-linux/rbconfig.rb). IMO there is no way other than patching these setup scripts one by one to accept CC, CXX, C(XX)FLAGS env. var. or taking flags via commandline switchs - IMO such patching doesn't pay, there is only one practical inconvenience which is raleted to passing CC/CXX values when ccache is used without symlinking compilers commands to ccache. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: Packaging .py files
On Wednesday 16 of July 2008, 15:07, Patryk Zawadzki wrote: Hi, Following up on a Jabber discussion with arekm I'd like to start packaging .py files along with their bytecode counterparts. Then put them in separate packages, named python-FOO-src or sth like that. [...] -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
[Th][FTP] where are newly built packages?
Hi, 1. I cannot find recently built packages on Th FTP - f.e. qt4-4.4.0*-1.i686.rpm . I checked ftp://master-ftp.pld-linux.org. 2. There are a few packages under ftp://master-ftp.pld-linux.org/dists/th/ready/i686/RPMS/ - AFAIK this directory should be emty for the time being. What's going on? -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: [th] builder fails: error: QtDBus-4.4.0-1.athlon: req /usr/lib/qt4/plugins/script not found
On Wednesday 07 of May 2008, 17:40, Kamil Dziedzic wrote: Could somebody fix that? For now on I can't. In my environment building [EMAIL PROTECTED] failed after: + [ -a /home/users/tw/tmp/qt4-4.4.0-root-tw/usr/include/qt4/Qt/Q3CanvasPolygonalItem ] + ifecho Qt3Support-devel /usr/include/qt4/Qt/Q3CanvasPolygonalItem + mkdevfl phonon + MODULE=phonon + shift + echo %defattr(644,root,root,755) + phonon-devel.files + ifecho phonon-devel /usr/lib/libphonon*.so Error generation phonon-devel files list! /home/users/tw/tmp/qt4-4.4.0-root-tw/usr/lib/libphonon*.so: no such file or direcotry! error: Bad exit status from /home/users/tw/tmp/rpm-tmp.99904 (%install) RPM build errors: Bad exit status from /home/users/tw/tmp/rpm-tmp.99904 (%install) 11634.15s real 18997.28s user 1143.41s system Error: package build failed. (no more info) There are some missing BRs? (I don't have any kde4 related stuff installed on my computer). http://buildlogs.pld-linux.org/index.php?dist=tharch=x86_64ok=0name=Virt ualBox Thanks;) -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: libdumbtts.spec - up to 0.2.0 - added -Makefile.patch which...
On Wednesday 16 of April 2008, 18:46, Kamil Dziedzic wrote: [...] $Log$ -Revision 1.3 2008-04-14 23:04:46 darekr -- md5sum, use -C for make +Revision 1.4 2008-04-15 18:30:31 twittner +- up to 0.2.0 +- added -Makefile.patch which allows to pass CC, CFLAGS, LDFLAGS Revision 1.2 2008-04-12 08:28:33 gotar - cosmetics Why did you remove 1.3 revision? Mistake? Probably I forgot to check status before committing - I'm sorry - it's already fixed, thnx for pointing out it. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES (rpm-4_4_9): rpm-lualeak.patch - more _free() compatible, ...
On Monday 17 of March 2008, 21:12, glen wrote: Author: glen Date: Mon Mar 17 20:12:18 2008 GMT Module: SOURCES Tag: rpm-4_4_9 Log message: - more _free() compatible, however no luck still Files affected: SOURCES: rpm-lualeak.patch (1.1.2.1 - 1.1.2.2) Diffs: Index: SOURCES/rpm-lualeak.patch diff -u SOURCES/rpm-lualeak.patch:1.1.2.1 SOURCES/rpm-lualeak.patch:1.1.2.2 --- SOURCES/rpm-lualeak.patch:1.1.2.1 Mon Mar 17 20:25:50 2008 +++ SOURCES/rpm-lualeak.patch Mon Mar 17 21:12:13 2008 @@ -6,7 +6,7 @@ lua_pushliteral(L, LUA_PATH); lua_pushstring(L, _lua_path); -free(_lua_path); -+free(_lua_path); _lua_path = NULL; ++if (_lua_path) free(_lua_path); _lua_path = NULL; } } lua_rawset(L, LUA_GLOBALSINDEX); @@ -15,7 +15,7 @@ if (Stat(_lua_init, st) != -1) (void)rpmluaRunScriptFile(lua, _lua_init); -free(_lua_init); -+free(_lua_init); _lua_init = NULL; ++if (_lua_init) free(_lua_init); _lua_init = NULL; man 3 free free() frees the memory space pointed to by ptr, which must have been returned by a previous call to malloc(), calloc() or realloc(). Otherwise, or if free(ptr) has already been called before, undefined behaviour occurs. If ptr is NULL, no operation is performed. ^^ IMO construction like `if(foo) free(foo);' has little value (maybe is useful for performance reasons - conditional instruction vs. calling function - I don't know). I didn't write any piece of C code for ages, but it seems to me that you would like achieve rather sth like: if(_lua_init) { free(_lua_init); _lua_init = NULL; } } } #undef _LUADOTDIR CVS-web: http://cvs.pld-linux.org/cgi-bin/cvsweb.cgi/SOURCES/rpm-lualeak.patch?r1=1. 1.2.1r2=1.1.2.2f=u ___ pld-cvs-commit mailing list [EMAIL PROTECTED] http://lists.pld-linux.org/mailman/listinfo/pld-cvs-commit -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: heath of th packages
On Sunday 06 January 2008, 17:28, Elan Ruusamäe wrote: http://ep09.pld-linux.org/~pldth/mvpkg.log please have look at that log and if you find your package there please fix it or point out that can be moved without the missing arch packages listed there as WARN 1. ERR: fwbuilder-2.1.16-2 has only src.rpm built error: Failed build dependencies: antlr conflicts with fwbuilder-2.1.16-2.src I can't help another way (for now) - I've put Conflict. 2. ERR: python-pexpect-2.2-1 moving would remove archs: ['ppc', 'x86_64', 'athlon', 'i486', 'i686'] It is OK - I've changed it to noarch - look at next entry (just the opposite) 3. ERR: qjackctl-0.3.2-1 moving would remove archs: ['x86_64'] Fix provided - builds on carme x86_64, rel. bumped sent to builders (BTW: I don't get mails with buildlogs from th-x86_64) -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: qt4 tools
On Friday 04 January 2008, 14:13, Jakub Bogusz wrote: AICS some packages expect Qt4 tools to have -qt4 suffix rather than qt4- prefix - so I'd rename them to be compatible with the rest of the world. Also, I'm going to add suffixed version of lrelease, so it could be called without specifying full %{_libdir}/qt4/bin path. Any comments? Go-ahead. Last time a was forced to patching qsvn due to this PLD idiosyncrasy and spent too much time digging into cmake macros - they also use qmake-qt4. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: rpm bug?
On Thu 19. of July 2007, 16:04, Elan Ruusamäe wrote: On Thursday 19 July 2007 11:44:14 Marcin Król wrote: not sure how to put this in proper words, but querying binheader results this, one should query srcheaders. Anyone brave enough to make requried changes into builder script? :) well. somebody mentioned that builder script should not depend on any higher language liker perl so... But why? Better have broken tool than use suitable language? -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: make-3.81 in AC
On Wed 4. of July 2007, 23:46, Elan Ruusamäe wrote: it seems already built: $ autotag make.spec make.spec:auto-ac-make-3_81-1 hghw why it was never moved to ac-main/ac-updates Here is explanation: http://src.ac.pld-linux.org/~buildsrc/queue.html I upgraded make on Ac-branch and sent it to builders right after reading Jerome's e-mail. [...] -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
[Th] Re: upgrade failed OK: espeak.spec
On Sun 20. of May 2007, 12:31, PLD th-ppc builder wrote: espeak.spec (HEAD): OK [...] error: Failed dependencies: /usr/lib/libespeak.so.1 is needed by espeak-devel-1.25-1.ppc package upgrade failed This error occurs only on Th. Of curse, /usr/lib/libespeak.so.1 is not packaged but only created in %post by ldconfig. Does anyone know, how to fix this? [...] -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: [Ac] hdparm hangs often during starting (udev issue?)
On Sun 11. of February 2007, Tomasz Wittner wrote: Hello, Nearly every two booting I'm forced to press reset button on my workstation (or sometimes is possible Alt+PrintScreen+S/B) - starting hangs during hdparm invokation. Because hdparm is called right after starting udev I suspect that udev (v. 079) doesn't finish its job and hdparm maybe hangs because of queued kernel/udev events[1] are not handled yet. Newer udev (v. 104) has binary called udevsettle which is used in start_udev (PLD Th) or /etc/init.d/udev (Debian unstable f.e.) scripts. As is said in [1]man udevsettle is used for delaying script execution until queued kernel/udev events are handled. I've just upgraded udev to 104 (and uClibc to 0.28.1) on my desktop (Ac) - both built from HEAD - everything seems to work properly. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
[Th] espeak - error during instalation
poldek:/all-avail install espeak-1.19-0.1 Processing dependencies... espeak-1.17-2 obsoleted by espeak-1.19-0.1 espeak-1.19-0.1 marks espeak-libs-1.19-0.1 (cap espeak-libs = 1.19-0.1) espeak-libs-1.17-2 obsoleted by espeak-libs-1.19-0.1 error: espeak-1.19-0.1: req /usr/share/espeak-data/voices/!v not found On Ac everything goes OK - what does cause that error? -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: [Th] espeak - error during instalation
On Sat 27. January 2007 14:29, Jakub Bogusz wrote: On Sat, Jan 27, 2007 at 01:34:40PM +0100, Tomasz Wittner wrote: poldek:/all-avail install espeak-1.19-0.1 Processing dependencies... espeak-1.17-2 obsoleted by espeak-1.19-0.1 espeak-1.19-0.1 marks espeak-libs-1.19-0.1 (cap espeak-libs = 1.19-0.1) espeak-libs-1.17-2 obsoleted by espeak-libs-1.19-0.1 error: espeak-1.19-0.1: req /usr/share/espeak-data/voices/!v not found On Ac everything goes OK - what does cause that error? Broken symlink? `find /usr/share/espeak-data/ -type l' shows nothing. I rather suspect that this error is caused by funny `!' characters in filename paths. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
[Ac] Re: TEST build ERRORS: espeak.spec
On Sat 27. January 2007 15:22, you wrote: espeak.spec (AC-branch): FAILED [...] /bin/sh ../libtool --tag=CXX --mode=compile athlon-pld-linux-g++ -DHAVE_CONFIG_H -I. -I..-Wall -pedantic -fno-exceptions -O2 -march=athlon -MT intonation.lo -MD -MP -MF .deps/intonation.Tpo -c -o intonation.lo intonation.cpp athlon-pld-linux-g++ -DHAVE_CONFIG_H -I. -I.. -Wall -pedantic -fno-exceptions -O2 -march=athlon -MT intonation.lo -MD -MP -MF .deps/intonation.Tpo -c intonation.cpp -fPIC -DPIC -o .libs/intonation.o athlon-pld-linux-g++ -DHAVE_CONFIG_H -I. -I.. -Wall -pedantic -fno-exceptions -O2 -march=athlon -MT intonation.lo -MD -MP -MF .deps/intonation.Tpo -c intonation.cpp -o intonation.o /dev/null 21 mv -f .deps/intonation.Tpo .deps/intonation.Plo make[2]: *** No rule to make target `numbers.c', needed by `numbers.lo'. Stop. make[2]: Leaving directory `/home/users/builder/rpm/BUILD/espeak-1.19-source/src' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/users/builder/rpm/BUILD/espeak-1.19-source' make: *** [all] Error 2 error: Bad exit status from /var/tmp/rpm-tmp.74144 (%build) [...] End-PLD-Builder-Info The next problem - it builds on my computer (Ac, autoconf-2.59) and on Th builders but it doesn't on Ac builders (an error is listed above). I'm getting mad. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: scripts: profile - QTDIR QMAKESPEC killed especially setting Q...
On Wed 17. January 2007 00:42, Arkadiusz Miskiewicz wrote: On Wednesday 17 January 2007 00:22, twittner wrote: Author: twittner Date: Tue Jan 16 23:22:38 2007 GMT Module: scripts Tag: HEAD Log message: - QTDIR QMAKESPEC killed especially setting QMAKESPEC to /usr/share/qt/mkspecs/linux-g++ is really inappropriate when qt-3.3.x and qt-4.x can exist together W qt4.spec IMO je warto czyścić przed budowaniem. A translation attempt: IMO is worth to unset these (QTDIR, QMAKESPEC) in qt4.spec before building. My suggestion: by default, start building with completely cleaned environment, then set only necessary env. variables [1]. Of course there should be possibility to turn off cleaning setting in rpm such environment (via rpm macro). [1] I remember that I was against such solution some time ago, but I wasted plenty of time trying to build qt4 due to QMAKESPEC setting in my shell environment. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
ruby - load paths are broken (?)
Hi, Today I tried to upgrade ruby-FXRuby (I haven't touched ruby for quite a long time) and notice that ruby-FXRuby examples aren't working at all - both Ac and Th are affected [1]: [EMAIL PROTECTED] /usr/src/examples/ruby-FXRuby-1.6.3]$ ruby hello.rb hello.rb:3:in `require': no such file to load -- fox16 (LoadError) from hello.rb:3 I've checked load paths - they seem completely broken: [EMAIL PROTECTED] /usr/src/examples/ruby-FXRuby-1.6.3]$ irb irb(main):001:0 $: = [/1.8, /1.8/i686-linux, , /usr/lib/ruby/1.8, /usr/lib/ruby/1.8/i686-linux, .] irb(main):002:0 AFAIK they should look like (ok - it's Debian with its idiosyncrasies): # irb1.8 irb(main):001:0 $: = [/usr/local/lib/site_ruby/1.8, /usr/local/lib/site_ruby/1.8/i486-linux, /usr/local/lib/site_ruby/1.8/i386-linux, /usr/local/lib/site_ruby, /usr/lib/ruby/1.8, /usr/lib/ruby/1.8/i486-linux, /usr/lib/ruby/1.8/i386-linux, .] [1] $ q ruby ruby-1:1.8.5p2-3 athlon athlon.ac.pld-linux.org $ q ruby ruby-1:1.8.5-2 athlon athlon.th.pld-linux.org -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ruby - load paths are broken (?)
On Sat 23. December 2006 13:47, Tomasz Wittner wrote: [...] [1] $ q ruby ruby-1:1.8.5p2-3 athlon athlon.ac.pld-linux.org $ q ruby ruby-1:1.8.5-2 athlon athlon.th.pld-linux.org Last working version that I've found: q --changelog ruby $Log: ruby.spec,v $ Revision 1.135 2006/08/27 19:05:49 qboosh - updated: ruby 1.8.5, onigd 2.5.7, doc-stdlib 0.10.1 $ q ruby ruby-1:1.8.5-1 athlon asus $ ruby -e '$:.each {|x| print #{x}\n}' /usr/lib/ruby/site_ruby/1.8 /usr/lib/ruby/site_ruby/1.8/i686-linux /usr/lib/ruby/site_ruby /usr/lib/ruby/1.8 /usr/lib/ruby/1.8/i686-linux . $ -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: [Ac] kaddressbook - sigsegv while adding IM
On Thu 2. November 2006 21:24, Tomasz Wittner wrote: Hi, kaddressbook segfaults while adding IM address. $ qf `which kaddressbook` kdepim-kaddressbook-9:3.5.4-2 athlon.ac.pld-linux.org I've tested kaddressbook from Debian/unstable: $ dpkg-query -W kaddressbook kaddressbook4:3.5.5.dfsg.1-1 and it doesn't crashes in the same situation. BTW: I saw that kde*.spec on AC-branch have been upgraded by glen to 3.5.5 version (but they have dotted releases). Upgrading kde to 3.5.5 didn't help: [KCrash handler] #5 0xb79b85d6 in KPluginInfo::name () from /usr/lib/libkutils.so.1 #6 0xb7f5520a in IMAddressWidget::slotProtocolChanged () from /usr/lib/libkaddressbook.so.0 #7 0xb7f54d2b in IMAddressWidget::init () from /usr/lib/libkaddressbook.so.0 #8 0xb7f54806 in IMAddressWidget::IMAddressWidget () from /usr/lib/libkaddressbook.so.0 #9 0xb7f52362 in IMEditorWidget::slotAdd () from /usr/lib/libkaddressbook.so.0 #10 0xb7f53760 in IMEditorWidget::qt_invoke () [...] $ qf `which kaddressbook` kdepim-kaddressbook-9:3.5.5-1 athlon.ac.pld-linux.org -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
[Th] There are more than one package which provide /usr/lib/debug ...
From poldek: There are more than one package which provide /usr/lib/debug: a) filesystem-debuginfo-3.0-12 b) galago-daemon-0.5.0-2 c) tcl-tls-1.5.0-2 Please remember that /usr/lib{,64} contains debug directory which should go into -debuginfo subpackage thus sth like: %files %{_libdir}/* is wrong. I've already corrected galago-daemon.spec and tcl-tls.spec in cvs. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: smtpdaemon and sendmail wrapper
On Tue 31. October 2006 21:41, Elan Ruusamäe wrote: imho they are different things 1. smtpdaemon means something that listens port 25 2. sendmail wrapper is package that provides /usr/lib/sendmail are there any program that really needs smtpdaemon in this context? i'd propose to use /usr/lib/sendmail Provides/Requires pairs when program needs /usr/lib/sendmail binary (and drop smtpdaemon if none actually needs it). IMO, for simplicity and clarity: `P/O: smtpdaemon' should be renamed to `P/O: smtpd' and what acts as smtp client (has sendmail(1) command) should have `P/O: smtp'. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: smtpdaemon and sendmail wrapper
On Tue 31. October 2006 21:41, Elan Ruusamäe wrote: imho they are different things 1. smtpdaemon means something that listens port 25 2. sendmail wrapper is package that provides /usr/lib/sendmail are there any program that really needs smtpdaemon in this context? i'd propose to use /usr/lib/sendmail Provides/Requires pairs when program needs /usr/lib/sendmail binary (and drop smtpdaemon if none actually needs it). P/O: smtpdaemon (imo better P/O: smtpd) is needed by MTAs for obsoleting each other.(?) -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
[Ac] kaddressbook - sigsegv while adding IM
Hi, kaddressbook segfaults while adding IM address. $ qf `which kaddressbook` kdepim-kaddressbook-9:3.5.4-2 athlon.ac.pld-linux.org I've tested kaddressbook from Debian/unstable: $ dpkg-query -W kaddressbook kaddressbook4:3.5.5.dfsg.1-1 and it doesn't crashes in the same situation. BTW: I saw that kde*.spec on AC-branch have been upgraded by glen to 3.5.5 version (but they have dotted releases). -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: smtpdaemon and sendmail wrapper
On Tue 31. October 2006 21:41, Elan Ruusamäe wrote: imho they are different things 1. smtpdaemon means something that listens port 25 2. sendmail wrapper is package that provides /usr/lib/sendmail are there any program that really needs smtpdaemon in this context? i'd propose to use /usr/lib/sendmail Provides/Requires pairs when program needs /usr/lib/sendmail binary (and drop smtpdaemon if none actually needs it). Thank gln for pointing about that - f.e. msmtp can acts as /usr/lib/sendmail but it isn't daemon and doesn't listens on 25 port - IMO msmtp (really: msmtp-sendmail package) shouldn't provides/obsoletes smtpdaemon. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: udev.spec - up to udev-100 - 60-cdrom_id.rules added
On Tue 26. September 2006 07:37, Fryderyk Dziarmagowski wrote: --- shadzik [EMAIL PROTECTED] wrote: Author: shadzik Date: Mon Sep 25 23:41:15 2006 GMT Module: SPECS Tag: HEAD Log message: - up to udev-100 - 60-cdrom_id.rules added Adding it will cause a clash with other rules and break nodes permissions. AFAIK, he doesn't subscribe pld-devel-en (only -pl). -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
[AC-branch] rpm.spec - rpm-find_lang.patch doesn't applied.
I beg to inform that rpm-find_lang.patch (glen's work) from AC-branch doesn't applied: + /home/users/inter/my/devel/rpm-ac/SOURCES/rpm-find_lang.patch 3 out of 3 hunks FAILED -- saving rejects to file scripts/find-lang.sh.rej error: Bad exit status from /var/tmp/rpm-tmp.10182 (%prep) Regards, -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: [Fwd: bzr.spec - podbicie wersji do 0.10]
On Thu 14. September 2006 22:28, Elan Ruusamäe wrote: On Thursday 14 September 2006 23:14, Michal Kochanowicz wrote: Topposting also is not okay, but you don't seem to care. i can't seem to find anymore that setting in kmail where's the preferrence where to start quoting. Try menu Settings-icon Composer-tab General-checkit Use smart quoting I niech mu ktoś napisze po angielskiemu tą mądrość o o belce w oku ;) (PL) po angielskiemu, a nawet angielskawemu: You can see blade in sb's eye but can't see beam in your own. (Polish proverb translated litterally at the request of mkochano ;) -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: pld-builder.spec - added initscript (template) for performi...
On Tue 4. April 2006 00:00, Mariusz Mazur wrote: On Monday 03 April 2006 18:00, twittner wrote: Author: twittner Date: Mon Apr 3 16:00:29 2006 GMT Module: SPECS Tag: HEAD Log message: - added initscript (template) for performing administrator's tasks for pld-builder such as mounting /proc in builder chroots, [EMAIL PROTECTED] PLD_Builder]$ grep mount *.py rpm_builder.py:chroot.run(test ! -f /proc/uptime mount /proc, 'root') 1. What about umounting proc inside chroots when system is shutting down?: $ sudo umount /mnt/hda9 umount: /mnt/hda9: device is busy umount: /mnt/hda9: device is busy I have inside /mnt/hda9 chroots with mounted /proc 2. What about installing/removing crontables with pld-builder jobs on demand? - I have pld-builder installed on my desktop. Every minute cron runs ca 5 python interpreters performing builder tasks at the same time - it breaks playing sound. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: python.spec - added python info pages as separate subpackage
On Mon 27. March 2006 10:59, wrobell wrote: On Sun, 2006-03-26 at 15:01 +0200, twittner wrote: Author: twittner Date: Sun Mar 26 13:01:42 2006 GMT Module: SPECS Tag: HEAD Log message: - added python info pages as separate subpackage [...] +BuildRequires: emacs = 21 is there any other way? AFAIK no - see Python-2.4.2/Doc/tools/mkinfo file. installing emacs just to build info docs... There is no problem - on builders is installed everything. If you want to build Python for yourself and don't want to install Emacs, you can pass `--without info' now - I've just added an appropriate bcond. BTW: Python info pages are used by Emacs context help - it is quite convenient. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: gawk.spec - created -devel
On Fri 24. March 2006 23:54, glen wrote: Author: glen Date: Fri Mar 24 22:54:27 2006 GMT Module: SPECS Tag: HEAD Log message: - created -devel What is the use of it? [...] +%files devel +%defattr(644,root,root,755) +%{_includedir} %_includedir belogns to FHS -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
STBR AC
Hi, Please, send to ac builders: wwwoffle.spec (HEAD) hk_classes.spec (HEAD) knoda.spec (HEAD) -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: python qt
On Mon 23. of January 2006 19:03, Tomasz Wittner wrote: [...] It shows that on amd64 also /usr/lib64/python2.4/site-packages/gtk-2.0 is missing. Oops - this path is appended by content of pygtk.pth file and gtk-2.0 directory existence. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: python qt
On Sat 21. of January 2006 12:05, Elan Ruusamäe wrote: so could somebody confirm or fix this? No, I can't, because, among other things, I don't have amd64. Also, I've never touched python-PyQt.spec, thus I don't understand, why have you sent this mail with CC to me ;) - I subscribe and read all topics on pld-devel-en. [...] -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: AC: TODO list
On Sat 31. of December 2005 11:15, Elan Ruusamäe wrote: On Saturday 31 December 2005 11:40, Tomasz Wittner wrote: On Sat 31. of December 2005 10:31, Andrzej Krzysztofowicz wrote: one more problem (ecasound.spec): error: line 36: Unknown tag: %ruby_mod_ver_requires_eq chicken-egg problem use %{?ruby_mod_ver_requires_eq} instead (like php-pecl*.spec do) $ rpm -E '%ruby_mod_ver_requires_eq' Requires: ruby-modules(ver) = 1.8 rpmbuild(macros) = 1.277 provides this macro and is placed as BuildRequires. build requires can't be satisfied if .spec is not parseable... Thanks a lot. [...] -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: AC: TODO list
On Sat 31. of December 2005 11:26, Andrzej Krzysztofowicz wrote: Andrzej Krzysztofowicz wrote: Elan =?iso-8859-1?q?Ruusam=E4e?= wrote: On Saturday 31 December 2005 11:40, Tomasz Wittner wrote: On Sat 31. of December 2005 10:31, Andrzej Krzysztofowicz wrote: one more problem (ecasound.spec): error: line 36: Unknown tag: %ruby_mod_ver_requires_eq chicken-egg problem use %{?ruby_mod_ver_requires_eq} instead (like php-pecl*.spec do) Problem is that srcbuilder has old, small envinronmrnt which is not upgraded. And it builds everything with --nodeps. ... and in this case (and similar) produces broken srpms by exapanding %{?ruby_mod_ver_requires_eq} to nothing? .. and rpm-build-macros cannot be upgraded at the moment because of broken ftp indexes. So two solutions: - fix it not to require newer macros to build src.rpm or - wait for sb with knowledge of ftp automatics (I don't know it) IMO the latter one is correct. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: AC: TODO list
On Sat 31. of December 2005 12:17, Tomasz Wittner wrote: On Sat 31. of December 2005 11:26, Andrzej Krzysztofowicz wrote: [...] Problem is that srcbuilder has old, small envinronmrnt which is not upgraded. And it builds everything with --nodeps. .. and in this case (and similar) produces broken srpms by exapanding %{?ruby_mod_ver_requires_eq} to nothing? It seems that I'm wrong - src.rpm keeps original spec and doesn't perform macros expansion. .. and rpm-build-macros cannot be upgraded at the moment because of broken ftp indexes. So two solutions: - fix it not to require newer macros to build src.rpm or - wait for sb with knowledge of ftp automatics (I don't know it) IMO the latter one is correct. Please, send to builders sth small - rake.spec f.e. I wonder what's happened (IMO everything should be OK). -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: raggle.spec - Ho Ho Ho, Meeerrryyy Raggle 0.4.4!
On Sat 31. of December 2005 13:08, glen wrote: Author: glen Date: Sat Dec 31 12:08:09 2005 GMT Module: SPECS Tag: HEAD Log message: - Ho Ho Ho, Meeerrryyy Raggle 0.4.4! [...] Not so merry (due to lack of ruby-modules on src builder) - the same with rake: ERRORS: raggle.spec From: PLD ac-src builder [EMAIL PROTECTED] To: [EMAIL PROTECTED] CC: [EMAIL PROTECTED] Date: Sat Dec 31 13:08:20 2005 raggle.spec (HEAD): FAILED --- raggle.spec:HEAD: Build-Time: user:0.62s sys:0.30s real:3.60s (faults io:84 non-io:27283) *** buildlog for raggle.spec request from: glen started at: Sat Dec 31 13:08:16 2005 building SRPM using: cd rpm/SPECS; nice -n 10 ./builder -nu --nodeps -bs -r HEAD -Tp auto-ac- -tt raggle.spec 21 U raggle.spec # $Revision: 1.14 $, $Date: 2005/12/31 12:08:03 $ Searching for tag auto-ac---... sh: ruby: not found error: line 16: Version required: Requires: ruby-modules(ver) = Building target platforms: noarch Building for target noarch Error: package build failed. (no more info) exit status 1280 error: No files produced. sh: ruby: not found error: line 16: Version required: Requires: ruby-modules(ver) = Begin-PLD-Builder-Info Build-Time: user:0.62s sys:0.30s real:3.60s (faults io:84 non-io:27283) End-PLD-Builder-Info -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: AC: TODO list
On Sat 31. of December 2005 14:07, Andrzej Krzysztofowicz wrote: building SRPM using: cd rpm/SPECS; nice -n 10 ./builder -nu --nodeps -bs -r HEAD -Tp auto-ac- -tt rake.spec 21 U rake.spec # $Revision: 1.20 $, $Date: 2005/12/31 10:36:33 $ Searching for tag auto-ac---... sh: ruby: not found error: line 15: Version required: Requires: ruby-modules(ver) = Error: package build failed. (no more info) exit status 1280 error: No files produced. sh: ruby: not found error: line 15: Version required: Requires: ruby-modules(ver) = Begin-PLD-Builder-Info Build-Time: user:0.59s sys:0.29s real:2.73s (faults io:84 non-io:27097) I see - rpm-build-macros-1.277 is installed whereas ruby-modules not. Please, install ruby-modules on src builder. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: ruby packages (was Re: AC: TODO list)
On Sat 31. of December 2005 14:47, Elan Ruusamäe wrote: On Saturday 31 December 2005 15:07, Andrzej Krzysztofowicz wrote: building SRPM using: cd rpm/SPECS; nice -n 10 ./builder -nu --nodeps -bs -r HEAD -Tp auto-ac- -tt rake.spec 21 [...] sh: ruby: not found error: line 15: Version required: Requires: ruby-modules(ver) = Begin-PLD-Builder-Info Build-Time: user:0.59s sys:0.29s real:2.73s (faults io:84 non-io:27097) fixed that. note that for the rpm macros to output something sane, ruby-modules package needs to be installed, as the rbconfig is otherwise not available. maybe move rbconfig to main ruby package? On the one hand for this purpose I've moved directory skeleton from ruby-modules to ruby package (I thought rather about 3rd party modules) and I've add %ruby_ver_requires_eq in addition to %ruby_mod_ver_requires_eq. # cp /usr/lib/ruby/1.8/i686-linux/rbconfig.rb . # rpm -ev ruby-modules --nodeps # install -m644 -D rbconfig.rb /usr/lib/ruby/1.8/i686-linux/rbconfig.rb # rpm -E '%ruby_version' 1.8 It seems that rbconfig doesn't depend on any others modules. But, on the other hand, additional ~6MB (ruby-modules) on builder doesn't hurt too much and is needed f.e. for generating documentation - most of ruby packages use rdoc for this purpose and rdoc depends on ruby-modules. So you gain no big profit from moving rbconfig to ruby package - BR and R ruby-modules are needed allmost for every ruby package (even I've such assumption - I'm not going to check every ruby package). So, do as you wish (I'm not convinced of it). $ poldek -e ruby $ rpm -E %ruby_mod_ver_requires_eq Requires: ruby-modules(ver) = ERROR $ poldek -u ruby $ rpm -E %ruby_mod_ver_requires_eq Requires: ruby-modules(ver) = ERROR Nice. $ poldek -u ruby-modules $ rpm -E %ruby_mod_ver_requires_eq Requires: ruby-modules(ver) = 1.8 btw, why not creating rpm-rubyprov package? i think the deps can be produced same way as php/perl are. i guess (by brief looking as i don't know the language) pattern like /require\s+'([^']+)';/ should do it. there's include yet, maybe sth else (I don't know ruby too - I've just started reading some tutorials and the like). Here is disscusion about this: http://www.redhat.com/archives/rpm-list/2003-January/msg00092.html http://www.redhat.com/archives/rpm-list/2003-January/msg00096.html -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: webapps and /cgi-bin alias
On Wed 28. of December 2005 22:58, Elan Ruusamäe wrote: some of you probably know that it's is troublesome to make alias underneath /cgi-bin directory. with %apache_config macros it was possible to insert slot so that the /cgi-bin/subdir alias is defined before global /cgi-bin alias is defined. Site note: boa server doesn't handle multiple ScriptAlias directives - this is next problem. but those are obsoleted by better webapps ;) I haven't seen webapps in action yet (I've just read its README). but webapps framework configs are all read after apache internal configs are read. so there's no way to load application's /cgi-bin alias override. Then read (all or some of these, grouped in separated directories) before. there are five ways to solve this. 1. do not use /cgi-bin alias for applications maybe use Redirect to application address to keep some kind of compatability for users depending on the url. 2. do not define /cgi-bin alias globally in apache. in ideal world all cgi programs should have their apache config and nothing is packaged to /home/services/httpd/cgi-bin directory at all. and therefore that alias is not needed. but that httpd/cgi-bin location would be for users, ie their own cgi-bin applications needed to be placed somewhere too. 3. hack apache global /cgi-bin alias and make it as webapp, thus it defines it's url in same way as webapps do. and somehow make it appear as last. i don't want to change this as sort-order independant webapps looked pretty good already. 4. change apache config so the /cgi-bin alias is always loaded last after include webapps.d/*.conf line. not sure if this will make config valid without cgi module present. because there are two modules providing cgi interface: mod_cgi and mod_cgid and ifmodule directive might not work 5. maybe something i've been missing? Don't perform any automatic configuration during instalation web aware packages and allow adminisrator to do her/his job (put exemplary configs in %doc). currently known packages of this kind are: - cvsweb - man2html-cgi -- glen -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: AC: TODO list
On Wed 28. of December 2005 18:21, Arkadiusz Miskiewicz wrote: Current TODO list, based on i686: [...] bd: ruby-ecasound-2.4.3-2: req ruby = 1:1.8.2 version mismatch - twittner $ cvs log ecasound.spec | grep -q twittner || echo Absent Absent $ I don't know what's it for - simply, I've removed this. Almost every ruby dependent package should require ruby-modules with major version of ruby due to directory layout (and maybe for some others reasons) - think of /usr/lib/ruby/{,site_ruby/}1.8 . I've already made appropriate changes in ruby.spec, rpm-macros-build.spec and ruby dependent specs - please, build ruby.spec and rpm-build-macros first, next ruby packages which are already on ftp - poldek:/all-avail search -r ruby shows 42 packages. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: ruby.spec - up to 1.8.4 - Merry Christmas!
On Sun 25. of December 2005 20:39, Arkadiusz Miskiewicz wrote: On Sunday 25 December 2005 20:34, Tomasz Wittner wrote: Cool! Commit then :) Thank you! But at first I must add `R: ruby-modules' to each of ruby-*.spec and R: ruby-tools due to `rdoc' usage to some of them - it will big change. Ok but don't wait too long (2-3 days at most ;)). ruby package was huge crap and it needed split badly. OK - splitting in ruby.spec is finished and fortunately it builds on all archs (thanks [twice] to glen for sending it), but note warnings on 64-bit archs: `cast from pointer to integer of different size'. I can't help it. I enclose list with ruby-dependent specs, with corrected BR/R in accordance to changes in ruby.spec and bumped releases. I assume they should be rebuilt to acquire proprer B/Requires (requires are more important since ruby package doesn't contain longer ruby modules). But take into account, before sending these specs to builders, that I'm unable to check each of one hundred packages (so I did not put STBR in my commitlogs) and I'm not going to fix if some of them are broken. -- Tomasz Wittner amrita2.spec amrita.spec cgikit.spec cvsspam.spec fastcst.spec freeride.spec glark.spec iowa.spec kdebindings.spec mysql-ruby.spec raggle.spec rails.spec rake.spec rbbr.spec redland-bindings.spec rpa.spec ruby-ActionMailer.spec ruby-ActionPack.spec ruby-ActionWebService.spec ruby-ActiveRecord.spec ruby-ActiveSupport.spec ruby-arrayfields.spec ruby-bdb.spec ruby-Borges.spec ruby-breakpoint.spec ruby-Criteria.spec ruby-DBI.spec ruby-dev-utils.spec ruby-dict.spec ruby-Diff-LCS.spec ruby-dl2.spec ruby-ecore.spec ruby-edje.spec ruby-epeg.spec ruby-epsilon.spec ruby-ERuby.spec ruby-esmart.spec ruby-evas.spec ruby-exif.spec ruby-extensions.spec ruby-facets.spec ruby-Fam.spec ruby-fcgi.spec ruby-Gen.spec ruby-gettext.spec ruby-Glue.spec ruby-guid.spec ruby-Iowa.spec ruby-jabber4r.spec ruby-jabber-rpc.spec ruby-json.spec ruby-kansas.spec ruby-levenshtein.spec ruby-libexif.spec ruby-LOG4R.spec ruby-mmap.spec ruby-mount.spec ruby-Multiton.spec ruby-mysql.spec ruby-NArray.spec ruby-Ncurses.spec ruby-Needle.spec ruby-Net-SFTP.spec ruby-Net-SSH.spec ruby-Nitro.spec ruby-odbc.spec ruby-Og.spec ruby-open4.spec ruby-OrderedHash.spec ruby-PageTemplate.spec ruby-pluginfactory.spec ruby-Quanty.spec ruby-Racc.spec ruby-rbogl.spec ruby-rdf-redland.spec ruby-redcloth.spec ruby-rena.spec ruby-RMagick.spec ruby-RMail.spec ruby-RubyGems.spec ruby-runt.spec ruby-sqlite.spec ruby-syntax.spec ruby-Text-Format.spec ruby-TMail.spec ruby-transaction-simple.spec ruby-vorbisfile.spec ruby-web.spec ruby-wikify.spec ruby-xattr.spec ruby-xhtmldiff.spec ruby-XTemplate.spec ruby-ZenTest.spec ruwiki.spec setup.rb.spec sqlite3-ruby.spec sqlite-ruby.spec tml.spec WideStudio.spec ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: ruby.spec - up to 1.8.4 - Merry Christmas!
On Sun 25. of December 2005 01:15, Arkadiusz Miskiewicz wrote: On Sunday 25 December 2005 00:42, aredridel wrote: Author: aredridelDate: Sat Dec 24 23:42:53 2005 GMT Module: SPECS Tag: HEAD Log message: - up to 1.8.4 - Merry Christmas! We want -libs for christmas in ruby ;-) No, I don't want ;). Here is my proposal: http://twittner.host.sk/files/ruby/ruby.spec -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: ruby.spec - up to 1.8.4 - Merry Christmas!
On Sun 25. of December 2005 19:32, Arkadiusz Miskiewicz wrote: On Sunday 25 December 2005 18:03, Tomasz Wittner wrote: On Sun 25. of December 2005 01:15, Arkadiusz Miskiewicz wrote: On Sunday 25 December 2005 00:42, aredridel wrote: Author: aredridelDate: Sat Dec 24 23:42:53 2005 GMT Module: SPECS Tag: HEAD Log message: - up to 1.8.4 - Merry Christmas! We want -libs for christmas in ruby ;-) No, I don't want ;). Here is my proposal: http://twittner.host.sk/files/ruby/ruby.spec Cool! Commit then :) Thank you! But at first I must add `R: ruby-modules' to each of ruby-*.spec and R: ruby-tools due to `rdoc' usage to some of them - it will big change. I think, I should wait some time for others people's opinions, reservations, etc. In the meantime i will add aforementioned requires to ruby-*.spec (for now, only on my disk, of course). -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: xplanet.spec - %{_includedir} macro used
On Tue 13. of December 2005 10:53, Andrzej Krzysztofowicz wrote: Elan =?iso-8859-1?q?Ruusam=E4e?= wrote: On Monday 12 December 2005 19:06, you wrote: Author: czuczDate: Mon Dec 12 17:06:18 2005 GMT Module: SPECS Tag: HEAD Log message: - %{_includedir} macro used what for? it's not safe thing to do. if somebody redefines _includedir in .spec, you change will break Yesterday, I expressed similar reservations on pld-devel-pl@ ... ps: adapter is not always right! Exactly. It is to be reverted. [...] - CPPFLAGS=-I/usr/include/freetype2 -I/usr/X11R6/include -I`pwd` \ + CPPFLAGS=-I%{_includedir}/freetype2 -I/usr/X11R6/include -I`pwd` \ See http://cia.navi.cx/stats/author/czucz or his last commits on [EMAIL PROTECTED] He made more such changes (ca in 20 specs). He reverted his changes only in xplanet.spec. Does he understand what he doing? - I'm not sure. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
python.spec - unnecessary dependencies
$ sudo rpm -Uvh RPMS/python-examples-2.4.1-1.athlon.rpm error: Failed dependencies: python-devel = 1:2.4.1-1 is needed by python-examples-2.4.1-1.athlon IMO python-examples doesn't require python-devel. The same issue with python-doc which requires python - what's it for? I for watching examples and documentation I don't need interpreter nor python development headers and libraries. There are any reasons for such requires? -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SOURCES: chkconfig-rc.d.patch (NEW), chkconfig-nostatic.patch (NEW...
On Sat 22. of October 2005 15:31, Tomasz Wittner wrote: On Sat 22. of October 2005 15:27, glen wrote: Author: glen Date: Sat Oct 22 13:27:05 2005 GMT Module: SOURCES Tag: HEAD Log message: - adjust for PLD Files affected: SOURCES: chkconfig-rc.d.patch (NONE - 1.1) (NEW), chkconfig-nostatic.patch (NONE - 1.1) (NEW) cvs ci -r DEVEL chkconfig-rc.d.patch chkconfig-nostatic.patch doesn't work? Sorry, I see - chkconfig-1.3.20 occupied DEVEL brach only temporarily. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: php-pear-HTML_AJAX.spec - BR: sed 4.0, use __sed macro
On Thu 18. of August 2005 22:15, Elan Ruusamäe wrote: On Thursday 18 August 2005 23:09, darekr wrote: Author: darekr Date: Thu Aug 18 20:09:42 2005 GMT Module: SPECS Tag: HEAD Log message: - BR: sed 4.0, use __sed macro what for? By what reasons We use %{__make}? - it is just /usr/bin/make . I've found some justifications: http://archive.netbsd.se/?ml=viewcvs-deva=2002-05m=406158 https://www.redhat.com/archives/fedora-test-list/2003-August/msg01306.html (PLD also is mentioned ;) Personally, on my FreeBSD box I have %_sed definied as /usr/local/bin/gsed because PLD relies on GNU sed which is differ from /bin/sed which comes with fbsd base system so using macros instead plain commands invokation was more convenient for me (I don't use fbsd at the moment). there's a lot of sed usage without macros so far that i know. put it into adapter.awk? Do it, if you please, but mayby wait some time for other's people opinions. [...] -sed -i -e '[EMAIL PROTECTED]@#%{php_pear_dir}#g' -e s#'HTML_AJAX'#'HTML'.DIRECTORY_SEPARATOR.'AJAX'#g \ +%{__sed} -i -e '[EMAIL PROTECTED]@#%{php_pear_dir}#g' -e s#'HTML_AJAX'#'HTML'DIRECTORY_SEPARATOR.'AJAX'#g \ $RPM_BUILD_ROOT%{php_pear_dir}/%{_class}/AJAX/Server.php -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: mono.spec - handle rpm mono provides/requires - rel 2
On Thu 11. of August 2005 21:42, Bartosz Taudul wrote: On Thu, Aug 11, 2005 at 09:38:21PM +0300, Elan Ruusamäe wrote: why thse are in /usr/bin, while other similiar scripts are in /usr/lib/rpm? Mono places them there. If it's good or not is open to discussion. I suppose that /usr/bin/mono-find-{provides,requires} are not called directly by admin or users but only from other programs (from rpm in this case). In such situation, in accordance with FHS 2.3 they should go to %{_libdir}/rpm/ http://www.pathname.com/fhs/pub/fhs-2.3.html#USRLIBLIBRARIESFORPROGRAMMINGANDPA quote /usr/lib : Libraries for programming and packages Purpose /usr/lib includes object files, libraries, and internal binaries that are not intended to be executed directly by users or shell scripts. /quote wolf -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: makewhatis.spec (NEW) - added - linux port of FreeBSD makew...
On Sat 9. of July 2005 10:10, Patrys :: Patryk Zawadzki wrote: Dnia 08-07-2005, pią o godzinie 23:02 +0200, Tomasz Wittner napisał(a): On Fri 8. of July 2005 18:34, Elan Ruusamäe wrote: On Friday 08 July 2005 19:17, twittner wrote: +Version: 2005.07.08 +Release: 0.1 +License: BSD uh! such versioning just begs for epoch bump afterwards! I can't imagine such situation (but I'm not brilliant ;) - version is taken from date '+%Y.%m.%d' . Please, provide an example when bumping epoch is necessary - show me output from: rpmvercmp .MM.DD-relX .mm.dd-relY We use versioning as follows: (1.)0.20050708-0.1 Then reverting to another snapshot or changing the versioning scheme allows for smooth upgrades within the same epoch. I really don't know what's it all about: $ rpmvercmp 2005.07.08-1 2005.07.09-0.1 2005.07.08-1 2005.07.09-0.1 It works as expected without bumping epoch. For example if author decides to give the next release a number of 2.0 instead of a date. I'm author of FreeBSD makewhatis Linux port and I don't plan to change versioning schema because it is my versioning schema, which is convient to me. I've taken makewhatis's sources from FreeBSD cvs (makewhatis.*, queue.h, stringlist.* and each exists only with its cvs revision), adapted and slightly refined (help printing, ac/am). I hope that everything was explained. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: makewhatis.spec (NEW) - added - linux port of FreeBSD makew...
On Sat 9. of July 2005 12:26, Arkadiusz Miskiewicz wrote: On Saturday 09 of July 2005 11:15, Fryderyk Dziarmagowski wrote: On Sat, 9 Jul 2005 10:46:15 +0200 Tomasz Wittner [EMAIL PROTECTED] wrote: your versioning is your personal problem. but in PLD you _must_ follow PLD's versioning schema (already pointed by Patrys). No comment - I have a good laugh. Huh? We follow author versioning unless these are snapshots where in this case these aren't snapshots it seems. Yes, it isn't snapshot. I make tarball after finishing and testing my work not in the half way. Now, I agree, my versioning schema may leads people to confusion - this thread shows that, but I found such versioning convenient for me - I needn't hesitate when and why one version should be 0.3.4 and next 1.2.5 - version is generated from `date '+%Y.%m.%d'` by my private release.sh script. You can see rcorder.spec (another my rework) and pure-config - they have similar versioning schema. Finally, I don't force people to using these programs. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: makewhatis.spec (NEW) - added - linux port of FreeBSD makew...
To all, But the main problem is that our cool makewhatis from man-1.5p called from: /etc/cron.weekly/makewhatis (makewhatis.crondaily in cvs/SOURCES) #!/bin/sh LOCKFILE=/var/lock/makewhatis.lock # [... commnets ...] [ -f $LOCKFILE ] exit 0 touch $LOCKFILE makewhatis -u -w rm -f $LOCKFILE exit 0 works in this way: # /etc/cron.daily/makewhatis chmod: cannot access `/var/cache/man/whatis': No such file or directory chmod: cannot access `/var/cache/man/whatis': No such file or directory chmod: cannot access `/var/cache/man/whatis': No such file or directory chmod: cannot access `/var/cache/man/whatis': No such file or directory chmod: cannot access `/var/cache/man/whatis': No such file or directory chmod: cannot access `/var/cache/man/whatis': No such file or directory chmod: cannot access `/var/cache/man/whatis': No such file or directory chmod: cannot access `/var/cache/man/whatis': No such file or directory and called from: /etc/cron.weekly/makewhatis (makewhatis.cronweekly in cvs/SOURCES) #!/bin/sh LOCKFILE=/var/lock/makewhatis.lock # [... commnets ...] [ -f $LOCKFILE ] exit 0 touch $LOCKFILE makewhatis -w rm -f $LOCKFILE exit 0 produces: root 0 2005-07-09 17:35 /var/cache/man/whatis [EMAIL PROTECTED] 17:36 ~]# ls -l /var/cache/man/whatis /usr/share/man/whatis -rw--- 1 root root 0 2005-07-09 17:36 /usr/share/man/whatis -rw-r--r-- 1 root root 0 2005-07-09 17:35 /var/cache/man/whatis But I don't care about broken makawhatis: # makewhatis-bsd -o /var/cache/man/whatis /usr/share/man /usr/X11R6/man /usr/local/man makewhatis-bsd: /usr/share/man/man1/xpmroot.1: No such file or directory # l /usr/share/man/man1/xpmroot.1 lrwxrwxrwx 1 root root 11 2004-09-17 01:13 /usr/share/man/man1/xpmroot.1 - fvwm-root.1 Dead link. # l /var/cache/man/whatis -rw-r--r-- 1 root root 891K 2005-07-09 17:43 /var/cache/man/whatis # umask 077 [EMAIL PROTECTED] 17:45 ~]# whatis XtSetValues XtSetValues(3x), XtVaSetValues(3x), XtSetSubvalues(3x), XtVaSetSubvalues(3x), XtGetValues(3x), XtVaGetValues(3x), XtGetSubvalues(3x), XtVaGetSubvalues(3x) - obtain and set widget resources $ whatis oki4drv oki4drv(1) - User space based driver for OKIPAGE 4w, OKI400w, OKIPAGE 4w PLUS and OKIPAGE 8w Lite printers oki4drv(8) - User space based driver for OKIPAGE 4w, OKI400w, OKIPAGE 4w PLUS and OKIPAGE 8w Lite printers grep: /usr/share/man/whatis: Permission denied ^^_ this file is produced by makewhatis of course I leave for someone else - you can switch to makewhatis-bsd or repair makewhatis. -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: makewhatis.spec (NEW) - added - linux port of FreeBSD makew...
On Sat 9. of July 2005 18:26, Paweł Sakowski wrote: On Fri, 2005-07-08 at 23:02 +0200, Tomasz Wittner wrote: I can't imagine such situation (but I'm not brilliant ;) - version is taken from date '+%Y.%m.%d' . I have one (off-topic) question: why invent your own private date format when there's ISO 8601? Interesting (http://www.cl.cam.ac.uk/~mgk25/iso-time.html). IMO 2005.07.08 is more readable than 20050708. I didn't know about this standard before. You can assume that 2005.07.08 is only version number no date ;) -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
Re: SPECS: makewhatis.spec (NEW) - added - linux port of FreeBSD makew...
On Fri 8. of July 2005 18:34, Elan Ruusamäe wrote: On Friday 08 July 2005 19:17, twittner wrote: +Version: 2005.07.08 +Release: 0.1 +License: BSD uh! such versioning just begs for epoch bump afterwards! I can't imagine such situation (but I'm not brilliant ;) - version is taken from date '+%Y.%m.%d' . Please, provide an example when bumping epoch is necessary - show me output from: rpmvercmp .MM.DD-relX .mm.dd-relY -- Tomasz Wittner ___ pld-devel-en mailing list pld-devel-en@lists.pld-linux.org http://lists.pld-linux.org/mailman/listinfo/pld-devel-en