Nice features put into this version (including some bugs that were patched in the 0.18 ebuild) Reproducible: Always Steps to Reproduce: 1. 2. 3.
Using the ebuild for 11.8-r1 seems to work (after commenting out all the patches), but I get compilation errors if the java, python or tcltk use flags are on.
0.12 is a major rewrite of the backend and lots of depending apps break because of this, most notably gnome-pilot.
So if an application that depends on it breaks with the new version is should block the upgrade, right?
don't understand what you're asking. but if the underlying library breaks the applications that depend on it, i don't see any reason why we should add that library.
Creating symlink seems to work for me. Ugly though. Wish I could find where upstream this was hard coded as i386
To #4, what I'm saying is, if an updated version of a library breaks a dependency it doesn't mean that you don't want it. Some people don't have that dependency installed and may still enjoy the library. If someone does have that dependency installed, portage should prevent him from upgrading (blocking).
#6, I totally agree with you. 0.12.0-pre3 is out now and the 0.12 version adds a whole new bunch of features and fixes a lot of bugs. After all, kpilot aleady has support for 0.12.0 so it should really be added into portage.
*** This bug has been marked as a duplicate of 87774 ***
oops, wrong bug
Could someone please create an ebuild of the newest version and put it here? I tried to to it but did not succeed. It was my first one ;-) Thx.
Created attachment 62585 [details] pilot-link-0.12.0-pre4 Ebuild This one is my first ebuild and its not yet perfect. But ist works here.
I'd also like to see this get in, also, could you please add an option, or dependency for libusb and the --enable-libusb configure option. For more information, please see http://pilot-link.org/README.usb
(In reply to comment #11) > This one is my first ebuild and its not yet perfect. But ist works here. Thank you, I was looking for this ebuild for some time, and I wasn't able to create one by myself. Unfortunately, the emerge failed, with the error message in the bottom. Emerging without the java USE flag resolved it. "./configure --help" gave me "--with-perl=perlexec"; does that mean that "yes" or "no" isn't enough? By the way, why are perl and python commented out? Output from emerge: creating reminders make[3]: Leaving directory `/var/tmp/portage/pilot-link-0.12.0_pre4/work/pilot-link-0.12.0-pre4/src' make[2]: Leaving directory `/var/tmp/portage/pilot-link-0.12.0_pre4/work/pilot-link-0.12.0-pre4/src' Making all in bindings make[2]: Entering directory `/var/tmp/portage/pilot-link-0.12.0_pre4/work/pilot-link-0.12.0-pre4/bindings' Making all in Java make[3]: Entering directory `/var/tmp/portage/pilot-link-0.12.0_pre4/work/pilot-link-0.12.0-pre4/bindings/Java' /opt/blackdown-jdk-1.4.2.02/bin/javac -d . ./MidiMgr.java ./TestTodos.java ./org/gnu/pilotlink/AddressAppInfo.java ./org/gnu/pilotlink/AppInfo.java ./org/gnu/pilotlink/DBInfo.java ./org/gnu/pilotlink/DatabaseExistsException.java ./org/gnu/pilotlink/DatabaseNotFoundException.java ./org/gnu/pilotlink/DatebookAppInfo.java ./org/gnu/pilotlink/FTB3CostCatRecord.java ./org/gnu/pilotlink/MemoRecord.java ./org/gnu/pilotlink/PilotLink.java ./org/gnu/pilotlink/PilotLinkException.java ./org/gnu/pilotlink/RawAppInfo.java ./org/gnu/pilotlink/RawRecord.java ./org/gnu/pilotlink/SysInfo.java ./org/gnu/pilotlink/User.java ./org/gnu/pilotlink/Version.java ./org/gnu/pilotlink/AddressRecord.java ./org/gnu/pilotlink/FTB3TripRecord.java ./org/gnu/pilotlink/FTB3VehicleRecord.java ./org/gnu/pilotlink/DatebookRecord.java ./org/gnu/pilotlink/MidiRecord.java ./org/gnu/pilotlink/ToDoRecord.java ./org/gnu/pilotlink/Record.java ./test.java ./datebooktst.java ./adrtest.java i686-pc-linux-gnu-g++ -I../../include -I ../../include -I ../../include -I /opt/blackdown-jdk-1.4.2.02/include -I /opt/blackdown-jdk-1.4.2.02/include/linux -fPIC -DHAVE_CONFIG_H -L ../../libpisock/.libs -lpisock -shared -o libjpisock.so ./libjpisock.c ./libjpisock.c: In function `_jobject* Java_org_gnu_pilotlink_PilotLink_readAppInfo(JNIEnv*, _jobject*, jint, jint)': ./libjpisock.c:107: error: invalid conversion from `jbyte*' to `int' ./libjpisock.c:107: error: initializing argument 4 of `PI_ERR dlp_ReadAppBlock(int, int, int, int, pi_buffer_t*)' ./libjpisock.c:107: error: invalid conversion from `int' to `pi_buffer_t*' ./libjpisock.c:107: error: initializing argument 5 of `PI_ERR dlp_ReadAppBlock(int, int, int, int, pi_buffer_t*)' ./libjpisock.c: In function `_jobject* Java_org_gnu_pilotlink_PilotLink_getAppInfoBlock(JNIEnv*, _jobject*, jint, _jstring*)': ./libjpisock.c:568: error: invalid conversion from `jbyte (*)[65535]' to `int' ./libjpisock.c:568: error: initializing argument 4 of `PI_ERR dlp_ReadAppBlock(int, int, int, int, pi_buffer_t*)' ./libjpisock.c:568: error: invalid conversion from `int' to `pi_buffer_t*' ./libjpisock.c:568: error: initializing argument 5 of `PI_ERR dlp_ReadAppBlock(int, int, int, int, pi_buffer_t*)' ./libjpisock.c: In function `_jobject* Java_org_gnu_pilotlink_PilotLink_getRecordByIndex(JNIEnv*, _jobject*, jint, jint, jint)': ./libjpisock.c:666: error: cannot convert `jbyte*' to `pi_buffer_t*' for argument `4' to `PI_ERR dlp_ReadRecordByIndex(int, int, int, pi_buffer_t*, recordid_t*, int*, int*)' ./libjpisock.c: In function `int pilot_connect(JNIEnv*, const char*)': ./libjpisock.c:881: error: cannot convert `sockaddr*' to `const char*' for argument `2' to `PI_ERR pi_bind(int, const char*)'./libjpisock.c:1005: error: invalid conversion from `int*' to `size_t*' ./libjpisock.c:1005: error: initializing argument 5 of `int pi_setsockopt(int, int, int, const void*, size_t*)' ./libjpisock.c: In function `_jobject* Java_org_gnu_pilotlink_PilotLink_getResourceByIndex(JNIEnv*, _jobject*, jint, jint, jint)': ./libjpisock.c:1058: error: cannot convert `jbyte*' to `pi_buffer_t*' for argument `4' to `PI_ERR dlp_ReadResourceByIndex(int, int, int, pi_buffer_t*, long unsigned int*, int*)' ./libjpisock.c: In function `_jobject* Java_org_gnu_pilotlink_PilotLink_readDBList(JNIEnv*, _jobject*, jint, jint, jint, jint)': ./libjpisock.c:1181: error: cannot convert `DBInfo*' to `pi_buffer_t*' for argument `5' to `PI_ERR dlp_ReadDBList(int, int, int, int, pi_buffer_t*)' make[3]: ** [java_lib] Erro 1 make[3]: Leaving directory `/var/tmp/portage/pilot-link-0.12.0_pre4/work/pilot-link-0.12.0-pre4/bindings/Java' make[2]: ** [all-recursive] Erro 1 make[2]: Leaving directory `/var/tmp/portage/pilot-link-0.12.0_pre4/work/pilot-link-0.12.0-pre4/bindings' make[1]: ** [all-recursive] Erro 1 make[1]: Leaving directory `/var/tmp/portage/pilot-link-0.12.0_pre4/work/pilot-link-0.12.0-pre4' make: ** [all] Erro 2
(In reply to comment #11) > Created an attachment (id=62585) [edit] Please add support support for libusb in the ebuild. 1. Add libusb to IUSE 2. Bellow "readline? ( sys-libs/readline )" add: libusb? ( dev-libs/libusb )" (watch the quote!) 3. Before "econf ${myconf} || die" add: use libusb \ && myconf="${myconf} --enable-libusb"
Error in compiling on ~x86. Additional info on request. Happens with and without the additional changes in comment 14. if /bin/sh ../libtool --tag=CC --mode=compile i686-pc-linux-gnu-gcc -DHAVE_CONFIG_H -I. -I. -I.. -I../include -I../include -march=athlon-xp -mtune=athlon-xp -O2 -fomit-frame-pointer -frename-registers -ftracer -finline-limit=600 -finline-all-stringops -pipe -fno-ident -Werror -MT libpisock_la-libusb.lo -MD -MP -MF ".deps/libpisock_la-libusb.Tpo" -c -o libpisock_la-libusb.lo `test -f 'libusb.c' || echo './'`libusb.c; \ then mv -f ".deps/libpisock_la-libusb.Tpo" ".deps/libpisock_la-libusb.Plo"; else rm -f ".deps/libpisock_la-libusb.Tpo"; exit 1; fi i686-pc-linux-gnu-gcc -DHAVE_CONFIG_H -I. -I. -I.. -I../include -I../include -march=athlon-xp -mtune=athlon-xp -O2 -fomit-frame-pointer -frename-registers -ftracer -finline-limit=600 -finline-all-stringops -pipe -fno-ident -Werror -MT libpisock_la-libusb.lo -MD -MP -MF .deps/libpisock_la-libusb.Tpo -c libusb.c -fPIC -DPIC -o .libs/libpisock_la-libusb.o libusb.c: In function `u_write': libusb.c:378: warning: passing arg 3 of `usb_bulk_write' discards qualifiers from pointer target type make[2]: *** [libpisock_la-libusb.lo] Error 1 make[2]: Leaving directory `/var/tmp/portage/pilot-link-0.12.0_pre4/work/pilot-link-0.12.0-pre4/libpisock' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/var/tmp/portage/pilot-link-0.12.0_pre4/work/pilot-link-0.12.0-pre4' make: *** [all] Error 2
Created attachment 76960 [details] pilot-link-0.12.0_pre4.ebuild added the libusb USE-flag and disabled python, since it does not compile properly
just for info: kpilot 3.5.0 does NOT compile against pilot-link-0.12pre4
Hi, trying to emerge pilot-link-0.12.0_pre4 and this is what I get: # emerge pilot-link ... if i686-pc-linux-gnu-gcc -DHAVE_CONFIG_H -I. -I. -I.. -I/usr/include -I../include -I/usr/include -O2 -march=pentium3 -pipe -fomit-frame-pointer -MT dlpsh.o -MD -MP -MF ".deps/dlpsh.Tpo" -c -o dlpsh.o dlpsh.c; \ then mv -f ".deps/dlpsh.Tpo" ".deps/dlpsh.Po"; else rm -f ".deps/dlpsh.Tpo"; exit 1; fi /bin/sh ../libtool --tag=CC --mode=link i686-pc-linux-gnu-gcc -O2 -march=pentium3 -pipe -fomit-frame-pointer -o dlpsh dlpsh.o libpiuserland.la -lpopt ../libpisock/libpisock.la i686-pc-linux-gnu-gcc -O2 -march=pentium3 -pipe -fomit-frame-pointer -o .libs/dlpsh dlpsh.o ./.libs/libpiuserland.a /usr/lib/libpopt.so ../libpisock/.libs/libpisock.so dlpsh.o: In function `handle_user_commands': dlpsh.c:(.text+0xec1): undefined reference to `readline' dlpsh.c:(.text+0xed4): undefined reference to `add_history' collect2: ld returned 1 exit status # emerge info Portage 2.0.54 (default-linux/x86/2005.0, gcc-3.4.4, glibc-2.3.5-r2, 2.6.15-gentoo-r1 i686) ================================================================= System uname: 2.6.15-gentoo-r1 i686 Intel(R) Pentium(R) M processor 1400MHz Gentoo Base System version 1.6.14 dev-lang/python: 2.3.5-r2, 2.4.2-r1 sys-apps/sandbox: 1.2.12 sys-devel/autoconf: 2.13, 2.59-r6 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r1 sys-devel/binutils: 2.16.1 sys-devel/libtool: 1.5.22 virtual/os-headers: 2.6.11-r2 ACCEPT_KEYWORDS="x86" AUTOCLEAN="yes" CBUILD="i686-pc-linux-gnu" CFLAGS="-O2 -march=pentium3 -pipe -fomit-frame-pointer" CHOST="i686-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/2/share/config /usr/kde/3.4/env /usr/kde/3.4/share/config /usr/kde/3.4/shutdown /usr/kde/3/share/config /usr/lib/X11/xkb /usr/lib/mozilla/defaults/pref /usr/share/config /usr/share/xsessions /var/qmail/control"CONFIG_PROTECT_MASK="/etc/gconf /etc/splash /etc/terminfo /etc/env.d" CXXFLAGS="-O2 -march=pentium3 -pipe -fomit-frame-pointer" DISTDIR="/usr/portage/distfiles" FEATURES="autoconfig distlocks sandbox sfperms strict" GENTOO_MIRRORS="http://trumpetti.atm.tut.fi/gentoo/ http://ftp-stud.fht-esslingen.de/pub/Mirrors/gentoo/ ftp://ftp-stud.fht-esslingen.de/pub/Mirrors/gentoo/ http://ftp.snt.utwente.nl/pub/os/linux/gentoo ftp://ftp.snt.utwente.nl/pub/os/linux/gentoo" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage" SYNC="rsync://rsync.nl.gentoo.org/gentoo-portage" USE="X acpi alsa avi cdr cups dvd foomaticdb gif gnome gtk gtk2 gtkhtml icq jpeg mmx mozilla mozsvg mpeg ncurses ogg opengl pam pda perl png pnp ppds quicktime sse ssl truetype trusted unicode usb vorbis win32codecs x86 xprint xv xvid userland_GNU kernel_linux elibc_glibc" Unset: ASFLAGS, CTARGET, LANG, LC_ALL, LDFLAGS, LINGUAS
*** Bug 127233 has been marked as a duplicate of this bug. ***
In bug 127233 I made a patch to fix the java bindings compile error seen in comment #13, and also ported the existing patches to the newer ebuild as needed. Feel free to use those files, and I apologize for not adding the libusb USE flag, I'm not booted into gentoo right now, and I'm a bit busy.
Concerning post #18 dlpsh.c:(.text+0xec1): undefined reference to `readline' dlpsh.c:(.text+0xed4): undefined reference to `add_history' Add readline as a useflag.. The configure script (as supplied by pilot-link) is broken.. If you do ./configure it will compiles, if you do it with --with-readline=yes it will compiles, but if you do it with --with-readline=no it won't(if you check the configure output it still says "yes" for the readline checks, i think because of the no it excludes some includes needed to compile it). The ebuild adds a --with-readline=no unless you add readline as a useflag.
Just a note, right on pilot-link.org's homepage is a message from the maintainers asking people not to package 0.12 for their distros. Until they say it is ok, I'm not going to look into adding this into portage. But I'm happy for this to live in an overlay for people to test so we can be ready when the ban on packaging is lifted. Also note that gnome-pilot will require 2.0.14 to be released proper before pilot-link-0.12 goes in.
Anyone got libusb mode to actually work? Seems to hang every time for me. Instructions at http://code.pilot-link.org/README.libusb
0.12.1 has been released, it works with the 0.12_pre4 ebuild as long as the java use flag has been disabled
The userland tools don't seem to build with 0.12.1 ebuild, investigating...
needs --enable-conduits to build the userland tools
anyone want to volunteer to modify the submitted ebuild that works with the new pilot link? thanks!
Created attachment 97608 [details] 0.12.1 ebuild Managed to get my system to have 0.12.1 pilot-link with this ebuild, but having java use flag causes a protection alarm on install: ACCESS DENIED open_wr: /etc/java-config-2/current-system-vm/jre/lib/amd64/libjpisock.so Works for me without having USE flags java nor tcltk turned on. tcltk USE flag causes the compile to fail when compiling tcl bits. I will be happy to improve the ebuild if so desired as I really could use a working sync for my Palm TX. :)
(In reply to comment #28) > Created an attachment (id=97608) [edit] > 0.12.1 ebuild > The ebuild works here on ~amd64 with the following flags: USE="perl png python readline -java -tcltk" (tested with jpilot-0.99.9) Sincerely, Gour
(In reply to comment #23) > Anyone got libusb mode to actually work? Seems to hang every time for me. Yes. Libusb works well for me on amd64. I modified the 0.12.1 ebuild attachment (dated 2006-09-21) as follows: local myconf="--enable-libusb --enable-conduits --includedir=/usr/include/libpisock" I had to copy /usr/share/pilot-link/udev/60-libpisock.rules to /etc/udev/rules.d. Perhaps this should be done by the ebuild. Finally since this udev rules file created the devices files with group id dialout, I added my user to the "dialout" group. Should the udev rules file use another group? After logging out and back in again I could access the Tungsten E using the port "usb:".
I've documented how to backup a Palm Tungsten T3: http://forums.gentoo.org/viewtopic-p-3488734.html#3488734
*** Bug 160451 has been marked as a duplicate of this bug. ***
pilot-link 0.12.0 or later is required for kde-base/kpilot-3.5.6; see bug #163853.
I am getting a sandbox violation when using this ebuild if [ -e /etc/java-config-2/current-system-vm/jre/lib/i386 ]; then cp libjpisock.so /etc/java-config-2/current-system-vm/jre/lib/i386/; elif [ -e /etc/java-config-2/current-system-vm/jre/lib/amd64 ]; then cp libjpisock.so /etc/java-config-2/current-system-vm/jre/lib/amd64/; else echo "Unknown architecture. Copy lipijsock.so to /etc/java-config-2/current-system-vm/jre/lib/ARCH/"; fi ACCESS DENIED open_wr: /etc/java-config-2/current-system-vm/jre/lib/i386/libjpisock.so cp: cannot create regular file `/etc/java-config-2/current-system-vm/jre/lib/i386/libjpisock.so': Permission denied make[2]: *** [install] Error 1 make[2]: Leaving directory `/var/tmp/portage/app-pda/pilot-link-0.12.1/work/pilot-link-0.12.1/bindings/Java' make[1]: *** [install-recursive] Error 1 make[1]: Leaving directory `/var/tmp/portage/app-pda/pilot-link-0.12.1/work/pilot-link-0.12.1/bindings' make: *** [install-recursive] Error 1 !!! ERROR: app-pda/pilot-link-0.12.1 failed. Call stack: ebuild.sh, line 1611: Called dyn_install ebuild.sh, line 1057: Called qa_call 'src_install' environment, line 3209: Called src_install pilot-link-0.12.1.ebuild, line 83: Called die !!! (no error message) !!! If you need support, post the topmost build error, and the call stack if relevant. !!! A complete build log is located at '/var/log/portage/app-pda:pilot-link-0.12.1:20070130-075444.log'. --------------------------- ACCESS VIOLATION SUMMARY --------------------------- LOG FILE = "/var/log/sandbox/sandbox-app-pda_-_pilot-link-0.12.1-26933.log" open_wr: /etc/java-config-2/current-system-vm/jre/lib/i386/libjpisock.so (symlink to /opt/sun-jdk-1.5.0.10/jre/lib/i386/libjpisock.so) -------------------------------------------------------------------------------- !!! This ebuild is from an overlay: '/usr/local/portage' emerge -iPortage 2.1.2-r4 (default-linux/x86/2006.1/desktop, gcc-4.1.1, glibc-2.5-r0, 2.6.19-suspend2-r1 i686) ================================================================= System uname: 2.6.19-suspend2-r1 i686 Genuine Intel(R) CPU T2300 @ 1.66GHz Gentoo Base System release 1.12.6 Timestamp of tree: Mon, 29 Jan 2007 18:30:01 +0000 distcc 2.18.3 i686-pc-linux-gnu (protocols 1 and 2) (default port 3632) [disabled] ccache version 2.4 [enabled] dev-java/java-config: 1.3.7, 2.0.31-r3 dev-lang/python: 2.4.3-r4 dev-python/pycrypto: 2.0.1-r5 dev-util/ccache: 2.4-r6 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: 2.16.1-r3 sys-devel/gcc-config: 1.3.14 sys-devel/libtool: 1.5.22 virtual/os-headers: 2.6.17-r2 ACCEPT_KEYWORDS="x86" AUTOCLEAN="yes" CBUILD="i686-pc-linux-gnu" CFLAGS="-march=prescott -O2 -pipe -fomit-frame-pointer" CHOST="i686-pc-linux-gnu" 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/terminfo /etc/texmf/web2c" CXXFLAGS="-march=prescott -O2 -pipe -fomit-frame-pointer -fvisibility-inlines-hidden" DISTDIR="/usr/portage/distfiles" FEATURES="autoconfig ccache distlocks metadata-transfer parallel-fetch prelink sandbox sfperms strict userpriv usersandbox" GENTOO_MIRRORS="http://gentoo.osuosl.org http://www.ibiblio.org/pub/Linux/distributions/gentoo" LDFLAGS="-Wl,-O1 -Wl,--sort-common -s" LINGUAS="en_US" MAKEOPTS="-j3" PKGDIR="/usr/portage/packages" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --delete-after --stats --timeout=180 --exclude=/distfiles --exclude=/local --exclude=/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/local/portage /usr/portage/local/layman/xeffects" SYNC="rsync://rsync.gentoo.org/gentoo-portage" USE="X a52 aac acpi alsa avi bash-completion berkdb bitmap-fonts cdr cdrw cli cracklib crypt css cups dbus divx4linux dlloader dri dv dvd dvdr dvdread dvdrw emboss encode fam fame fbcon ffmpeg flac gdbm gif gnutls gpm gstreamer gtk hal iconv isdnlog jabber java jpeg kde kdeenablefinal kdehiddenvisibility kdexdeltas kerberos lame libg++ logrotate mikmod mmx mp3 mpeg msn musicbrainz ncurses nls nptl nptlonly nsplugin nvidia ogg opengl pam pcre pda perl pertty pic png pnp ppds pppd python qt qt3 qt4 quicktime readline reflection rss samba sdl seamonkey session spell spl sse sse2 sse3 ssl subtitles tcltk tcpd tetex tiff truetype truetype-fonts type1-fonts udev unicode usb userlocales vcd vorbis wifi win32codecs x86 xcomposite xinerama xml xorg xscreensaver xv xvid zeroconf zlib" ALSA_CARDS="ali5451 als4000 atiixp atiixp-modem bt87x ca0106 cmipci emu10k1x ens1370 ens1371 es1938 es1968 fm801 hda-intel intel8x0 intel8x0m maestro3 trident usb-audio via82xx via82xx-modem ymfpci" 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" ELIBC="glibc" INPUT_DEVICES="evdev keyboard mouse synaptics" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LINGUAS="en_US" USERLAND="GNU" VIDEO_CARDS="nv nvidia" Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LANG, LC_ALL, PORTAGE_RSYNC_EXTRA_OPTS
(In reply to comment #34) > I am getting a sandbox violation when using this ebuild I fixed that, and did some more changes (compared to the 0.11.8 ebuild, which is more or less equal to the one attached here so far): * version bump * support separate tcl/tk use flags instead of the deprecated tcltk one (TODO: not tested yet, as I do not like/use TCL) * repair python support (patch sent to upstream authors) * fix sandboxed Java install * fix sandboxed Python install * remove netsync patch (came from upstream, assume it was included long ago) * remove M4 patch (included upstream, checked) * added --enable-conduits for building userland tools * added "usb" use flag for --enable-libusb * TODO: install udev rules (with proper gid, see comment #30)?
Created attachment 109051 [details] updated 0.12.1 ebuild w/ python/java/tcltk/USB/userland support this is annoying: [no attachments can be made obsolete], actually this *does* obsolete the "0.12.1 ebuild"
Created attachment 109053 [details, diff] arch-aware sandboxed install of java libpisock.so sent to upstream ML
Created attachment 109054 [details, diff] fixes --with-python support in configure (equivalent patch) sent to upstream ML
Created attachment 109055 [details, diff] fix sandboxed python install equivalent to http://bugs.gentoo.org/attachment.cgi?id=82887 (which I found later)
Why not yet in portage, even hard masked? still missing the kde 3.5.6 pim application.
Can someone enlighten me on the situation with pilot-link? Why isn't it in portage already? Any chance this going to happen soon? Debian stable has the same version for crying out loud.I really would like to use the new kpilot (I have serious troubles with this one). Thanks in advance! Debian: http://packages.debian.org/cgi-bin/search_packages.pl?keywords=pilot-link+&searchon=names&subword=1&version=all&release=all
0.12.2 is out (with great features). ebuild above seems to work without the patches (commented out the epatch in src_unpack() maybe some bluetooth use flag will be needed
Just to give some feedback: 0.12.1 works fine for me (but i needed to disable the tests) 0.12.2 compile but doesn't work zith kpilot-3.5.6.
Evolution doesn't compile against pilot-link-0.12.* without a patch. I found a patch in the gnome bugzilla to update the evolution conduits. http://bugzilla.gnome.org/attachment.cgi?id=78892&action=view This applies cleanly against evolution-2.8.3, and appears to work fine, for those who might be interested.
Created attachment 113813 [details] pilot-link-0.12.2.ebuild Appears that pilot-link-0.12.2.ebuild does not need the patches, but it does need a newer libmal and some backport fixes to kpilot.
Created attachment 113815 [details] dev-libs/libmal-0.44.ebuild To support >=pilot-link-0.12.0.
Created attachment 113816 [details] kde-base/kpilot-3.5.6.ebuild This ebuild applies the backported patches for >=pilot-link-0.12.2.
Created attachment 113818 [details, diff] patchset for kpilot place in kpilot $FILESDIR
Just a note that pilot-link-0.12.2 + libmal-0.44 appear to work just fine with jpilot-0.99.9 (also not in the portage tree although released on 8/27/06). WIth the newer pilot-link a friend can now sync his Palm m515, previously he couldn't. Chris
Looks like the 0.12.2 ebuild causes some problems with evince update: >>> Emerging (1 of 6) app-text/evince-0.6.1-r2 to / * evince-0.6.1.tar.bz2 RMD160 ;-) ... [ ok ] * evince-0.6.1.tar.bz2 SHA1 ;-) ... [ ok ] * evince-0.6.1.tar.bz2 SHA256 ;-) ... [ ok ] * evince-0.6.1.tar.bz2 size ;-) ... [ ok ] * checking ebuild checksums ;-) ... [ ok ] * checking auxfile checksums ;-) ... [ ok ] * checking miscfile checksums ;-) ... [ ok ] * checking evince-0.6.1.tar.bz2 ;-) ... [ ok ] >>> Unpacking source... >>> Unpacking evince-0.6.1.tar.bz2 to /var/tmp/portage/app-text/evince-0.6.1-r2/work * Applying evince-0.5.3-display-menu.patch ... [ ok ] * Applying evince-0.6.1-dbus-switch.patch ... [ ok ] * Running eautoreconf in '/var/tmp/portage/app-text/evince-0.6.1-r2/work/evince-0.6.1' ... * Running aclocal ... [ !! ] * Failed Running aclocal ! * * Include in your bugreport the contents of: * * /var/tmp/portage/app-text/evince-0.6.1-r2/temp/aclocal-17440.out !!! ERROR: app-text/evince-0.6.1-r2 failed. Call stack: ebuild.sh, line 1614: Called dyn_unpack ebuild.sh, line 751: Called qa_call 'src_unpack' environment, line 3654: Called src_unpack evince-0.6.1-r2.ebuild, line 79: Called eautoreconf autotools.eclass, line 93: Called eaclocal autotools.eclass, line 130: Called autotools_run_tool 'aclocal' autotools.eclass, line 218: Called die !!! Failed Running aclocal ! !!! If you need support, post the topmost build error, and the call stack if relevant. !!! A complete build log is located at '/var/tmp/portage/app-text/evince-0.6.1-r2/temp/build.log'. # cat /var/tmp/portage/app-text/evince-0.6.1-r2/temp/aclocal-17440.out ***** aclocal ***** /usr/bin/m4:/usr/share/aclocal/pilot-link.m4:1: ERROR: end of file in string autom4te-2.61: /usr/bin/m4 failed with exit status: 1 aclocal-1.10: autom4te failed with exit status: 1
(In reply to comment #50) > Looks like the 0.12.2 ebuild causes some problems with evince update: Strange, since pilot-link isn't a depend or an rdepend for evince. As far as I can tell the two packages should have nothing to do with each other. But I'm not a developer, was just trying to help out a friend who needed to sync his palm. Most of the real work here was done by Hans Meine, finding/creating/etc. the patches and getting them to upstream. The ebuilds work for me and others but it's one of those YMMV stories. BTW evince fails here as well - it's not just a one off, but I don't use it so as a non-dev there is little I can do to help. And since the devs haven't put this stuff in portage yet, there isn't much avenue for a real bug report. My feeling is that it's an ebuild bug, but also the evince version in the portage tree is old as 0.8.0 is out and only 0.6.1 is offered in portage. File a version bump bug, and maybe it will all work. Chris
(In reply to comment #50) > Looks like the 0.12.2 ebuild causes some problems with evince update: Found this: http://www.mail-archive.com/autoconf%40gnu.org/msg15786.html Now all we need is someone who knows how to fix it....
Created attachment 114174 [details, diff] pilot-link-0.12.2.ebuild applies patch to fix m4 file
Created attachment 114176 [details, diff] pilot-link-0.12.2-m4.patch patch for pilot-link-0.12.2, place in $FILESDIR
Created attachment 114179 [details] pilot-link-0.12.2.ebuild oops..had this listed as patch previously This fixes the evince hand on pilot-link.m4, although evince still doesn't build for me - I get other errors: ***** automake ***** gnome-doc-utils.make:63: HAVE_GNOME_DOC_UTILS does not appear in AM_CONDITIONAL help/Makefile.am:3: `gnome-doc-utils.make' included from here gnome-doc-utils.make:74: if $(DOC_H_FILE: non-POSIX variable name gnome-doc-utils.make:74: (probably a GNU make extension) help/Makefile.am:3: `gnome-doc-utils.make' included from here gnome-doc-utils.make:77: if $(DOC_H_FILE: non-POSIX variable name gnome-doc-utils.make:77: (probably a GNU make extension)
(In reply to comment #55) > This fixes the evince hand on pilot-link.m4, although evince still doesn't > build for me - I get other errors: > ***** automake ***** > > gnome-doc-utils.make:63: HAVE_GNOME_DOC_UTILS does not appear in AM_CONDITIONAL Actually, app-text/evince-0.6.1-r2 installs just fine now. I forgot I had been testing it renamed to app-text/evince-0.8.0 which apparently wont work without some patches.
Shouldn't it be epatch ${FILESDIR}/$(P)-m4.patch in the ebuild?
(In reply to comment #57) > Shouldn't it be > epatch ${FILESDIR}/$(P)-m4.patch > in the ebuild? Works the same. It's a flexible naming convention that's probably more useful for gentoo specific patches that might change slightly with each package version. This patch is not gentoo specific, it's from upstream, and we probably wont need it again for the next release. It's GPL, write it however you like. Chris
Okay, I'm just curious of gentoo ebuild writing "standards" and "policies". Best distro around and I'd like to be able to contribute into it in "the proper way". It's described in a way on gentoo developer info pages, but real life ebuilds teach much more than those definitions that are clear to people already involved in contributiog.
(In reply to comment #59) > Okay, I'm just curious of gentoo ebuild writing "standards" and "policies". so submit a patched ebuild and recommend it be included in the portage tree, no need to start a pissing contest over the semantics of an unofficial ebuild provided by someone who was just trying to help out guess it's time for me to focus on writing apt packages instead
No, you got me wrong, I didn't mean to start a pissing contest. I'm sorry if I explained myself badly. I'm just interested in what gentoo developers consider to be the right way to do things.
(In reply to comment #60) > so submit a patched ebuild and recommend it be included in the portage tree, no > need to start a pissing contest over the semantics of an unofficial ebuild > provided by someone who was just trying to help out It's nice that you're helping out and all, but don't be rude, all he did was ask if a change made sense, I wouldn't call that starting a pissing contest. With regards to the question, think about what would happen if you renamed the ebuild file to build a newer version of pilot-link. If the name and version are hard coded, then that line will still attempt to apply the patch to the newer version of the source code, whereas if you have $(P) in there, you'll have to copy the patch file, or delete that epatch line. It's not a big difference, but it doesn't make sense to put $(P) in on that line.
Anyway, the 0.12.2 + the m4 patch confirmed to work nicely on my machine as well.
Alastair, it's time to do something about this bug. The KPilot version which is blocked atm. fixes quite some bugs. And if I read the upstream information correctly, gnome-pilot 2.0.15 should work with 0.12.x. Please add it at least masked to the tree, so people can actually test if their app works with it.
I am sorry for intervening at such a late stage of development, but I managed to come up with an ebuild of my own before I looked into the bug. And despite my modesty (which is my only fault, so it's not all that bad), I have to state that my ebuild is better in a number of ways. It 1. adds a bluetooth USE to control --with-bluez and also contains a patch that makes it work. 2. contains a fix for java bindings parallel build. It contains another java install fix too, but java binding won't work, anyway. It should pretty easy to fix, but since I have absolutely no expierence with it I'd leave it for those who do. 3. it adds threads USE-flag to control thread-safety. 4. it removes tk USE-flags which has no effect on the package. 5. it looks better (to me, don't take it to heart) ;-) It has shortcomings too (not that other ebuilds don't suffer these). As I mentioned java bindings don't work and tcl as well, because it seems tcl-bindings don't go in sync with library. Well, I hope java and tcl guys can take care of it. Thanks for your time, files follow...
Created attachment 116442 [details] pilot-link-0.12.2.ebuild
Created attachment 116443 [details, diff] pilot-link-0.12.2-build.patch
Meanwhile, I've talked to upstream and they seem to accept most if not all patches for 0.12.3. And they're going to remove tcl bindings too, which sounds like a great relief ;-)
Michael: Input is always appreciated. :) pda team: If I do not get a reply, I'll go ahead.
(In reply to comment #69) > Michael: Input is always appreciated. :) Good to hear ;-) I think we can safely drop tcl and java should be relatively easy to install provided that someone has any idea about gentoo's java unlike me ;-) And if, by chance, you will want more input feel free to find me @freenode as polyonymous (mostly on #gentoo-db) almost anytime.
Is there a reason, why the ebuild is not even included in portage? e.g. Kpilot 3.5.6 depends on pilot-link >= 0.12
(In reply to comment #66) > Created an attachment (id=116442) [edit] > pilot-link-0.12.2.ebuild I had to add the m4 patch to this, otherwise I was getting aclocal failures when emerging hal and xscreensaver.
> > I had to add the m4 patch to this, otherwise I was getting aclocal failures > when emerging hal and xscreensaver. > Hmm... Strange, for what I see in the patch (from comment 67) the m4 patch is there.
It's not the same patch. The one in pilot-link-0.12.2-m4.patch is more extensive. I tried without and with pilot-link-0.12.2-m4.patch and the problem only went away after applying the separate patch, which I applied after pilot-link-0.12.2-build.patch src_unpack() { unpack ${A} cd "${S}" epatch "${FILESDIR}/${P}-build.patch" epatch "${FILESDIR}/${P}-m4.patch" eautoconf eautomake }
(In reply to comment #74) > It's not the same patch. The one in pilot-link-0.12.2-m4.patch is more > extensive. I tried without and with pilot-link-0.12.2-m4.patch and the problem Yes, but the m4 problem should be solved by the shorter one... Besides, I'm not sure if the extensive version contains anything but this short one and whitespace changes (haven't checked through, though). > only went away after applying the separate patch, which I applied after > pilot-link-0.12.2-build.patch > > src_unpack() { > unpack ${A} > cd "${S}" > epatch "${FILESDIR}/${P}-build.patch" > epatch "${FILESDIR}/${P}-m4.patch" > eautoconf > eautomake > } This is strange, since I have my version installed here and experience no problem... Should, probably recheck it... The funny thing is that the m4 patch should not apply _after_ build one, it should fail... Well, again, need to check it once more.
(In reply to comment #75) > Yes, but the m4 problem should be solved by the shorter one... Since when did computers do what they should. I know I tried this on two systems, one ~amd64 and one ~ppc and the longer patch was needed to get rid of the errors. If it will help, I can reinstall with the shorter patch and let you have the error messages.
Don't know if it helps, but I just installed pilot-link-0.12.2, with only the pilot-link-0.12.2-build.patch, as well ass kpilot-3.5.6 with the attached patchset and the attached libmal-0.44. After unmasking oodles of kde 3.5.6 packages (split-ebuilds,) I had no compile problems, for good measure I remerged hal. This was done on one amd64 system and 2 x86 systems. kpilot seems to be working properly.
(In reply to comment #76) > Since when did computers do what they should. That depends on how convincing you sound when you try to persuade them ;-) > I know I tried this on two systems, one ~amd64 and one ~ppc and the longer > patch was needed to get rid of the errors. If it will help, I can reinstall > with the shorter patch and let you have the error messages. I think I'll try it myself if I get time. So, you install pilot-link and then install of xscreensaver fails on aclocal, right? (In reply to comment #77) > Don't know if it helps, but I just installed pilot-link-0.12.2, with only the > pilot-link-0.12.2-build.patch, as well ass kpilot-3.5.6 with the attached > patchset and the attached libmal-0.44. After unmasking oodles of kde 3.5.6 > packages (split-ebuilds,) I had no compile problems, for good measure I > remerged hal. This was done on one amd64 system and 2 x86 systems. kpilot > seems to be working properly. The problem is not installing pilot-link, but doing auto* stuff afterwords. Not sure if kde is affected, although, most likely...
I can confirm the xscreensaver-5.02 aclocal bug, on both amd64 & x86. The output of aclocal is: ***** aclocal ***** configure.in:307: warning: underquoted definition of AC_PROG_CC_ANSI configure.in:307: run info '(automake)Extending aclocal' configure.in:307: or see http://sources.redhat.com/automake/automake.html#Extending-aclocal /usr/bin/m4:/usr/share/aclocal/pilot-link.m4:1: ERROR: end of file in string autom4te-2.61: /usr/bin/m4 failed with exit status: 1 aclocal-1.10: autom4te failed with exit status: 1
(In reply to comment #78) > So, you install pilot-link and then > install of xscreensaver fails on aclocal, right? Right.
Created attachment 118752 [details, diff] fixed pilot-link-0.12.2-build.patch Thanks and sorry, here's an updated build patch.
(In reply to comment #80) > (In reply to comment #78) > > So, you install pilot-link and then > > install of xscreensaver fails on aclocal, right? > > Right. Once again, thanks and, as usually, computer have done exactly what it should have done, but I haven't :)
Sorry, it still doesn't work. If I emerge pilot-link with the new build.patch and then emerge hal, I get this * Running eautoreconf in '/var/tmp/portage/sys-apps/hal-0.5.9-r1/work/hal-0.5.9' ... * Running aclocal ... [ !! ] * Failed Running aclocal ! * * Include in your bugreport the contents of: * * /var/tmp/portage/sys-apps/hal-0.5.9-r1/temp/aclocal-1698.out !!! ERROR: sys-apps/hal-0.5.9-r1 failed. And % cat /var/tmp/portage/sys-apps/hal-0.5.9-r1/temp/aclocal-1698.out ***** aclocal ***** /usr/share/aclocal/wxwin.m4:36: warning: underquoted definition of AM_OPTIONS_WXCONFIG /usr/share/aclocal/wxwin.m4:36: run info '(automake)Extending aclocal' /usr/share/aclocal/wxwin.m4:36: or see http://sources.redhat.com/automake/automake.html#Extending-aclocal /usr/bin/m4:/usr/share/aclocal/pilot-link.m4:1: ERROR: end of file in string autom4te-2.61: /usr/bin/m4 failed with exit status: 1 aclocal-1.10: autom4te failed with exit status: 1 When I added the separate m4 patch to the ebuild, it worked again.
Created attachment 118753 [details, diff] really fixed pilot-link-0.12.2-build.patch oops ;-)
(In reply to comment #83) > Sorry, it still doesn't work. If I emerge pilot-link with the new build.patch Yes, I know :-( attached too soon.
xscreensaver-5.02 now gets past aclocal fine with 'really fixed pilot-link-0.12.2-build.patch' on amd64 & x86 for me.
> xscreensaver-5.02 now gets past aclocal fine with 'really fixed > pilot-link-0.12.2-build.patch' on amd64 & x86 for me. Thanks for report, Sal!
It's working now, on ~amd64 and ~ppc. Thanks for the fix.
(In reply to comment #88) > It's working now, on ~amd64 and ~ppc. Thanks for the fix. Thanks for the report, Neil.
Had to allow kpilot-3.5.5 with kde 3.5.7. This is far from ideal but at least works for now with the split ebuilds. It would be good to get this bumped for those kpilot users. Personally I do not use this stuff and so cannot test.
0.12.2 is now out for over 4 months and the latest ebuild in portage is for 0.11.8 (released 4 years ago). There are significant improvements in 0.12.x and it would be helpful for many users to be able to use 0.12. I urge the maintainers to please look into this (again) and make an ebuild available via portage.
xscreensaver and hal emerge without problems on my system after installing with the really fixed patch. I'd be happy if this ebuild would make it soon into the portage tree.
I would also like to request the inclusion of pilot-link-0.12.2 into the portage tree. Among other things, the new branch makes it possible to transfer Palm standard-violating program files with resources exceeding 64 kB (two such files are included in the IBM WebSphere Java environment for PalmOS), which with older versions would make it impossible to back one's Palm up without creating exclusion lists for pilot-xfer.
A working ebuild for the new version is provided here, and this is a blocking bug for kde-3.5.7 stabilization. What are we waiting for? KDE-5.0?
Hello, pda people? Is someone alive? There is an ebuild, you only have to put it on portage tree.
(In reply to comment #95) > Hello, pda people? Is someone alive? There is an ebuild, you only have to put > it on portage tree. Yes, and this ebuild unfortunately has numerous issues: - automagic deps on readline, libpng - the threads USE flag doesn't work, i. e. another automagic dep - debugging is unconditionally switched on (i. e. debug USE flag is missing) - java handling is wrong (cf. java-pkg-opt-2.eclass) - the tcl bindings don't compile here and upstream says they should be disabled. - virtual/libc shouldn't be in RDEPEND. - a DEPEND on dev-libs/popt is missing - DEPEND contains virtual/jre which is wrong - doesn't compile with -readline - instead of eautoconf and eautomake, eautoreconf should be used. - etc. etc. I'm working on an ebuild but am battling myself with several automagic problems, the worst of which being bluez.
I can fix some of these issues with my ebuild, but not all. In particular, I don't want to mess with java part. Anyway, I'm willing to help and of the gentoo channels you can find me on #gentoo-db as a 'polyonymous'. Feel free to /msg too. I didn't want to disturb you with /msg-ing and don't see an appropriate channel in your /whois. > - automagic deps on readline, libpng > - the threads USE flag doesn't work, i. e. another automagic dep > - debugging is unconditionally switched on (i. e. debug USE flag is missing) > - java handling is wrong (cf. java-pkg-opt-2.eclass) > - the tcl bindings don't compile here and upstream says they should be > disabled. > - virtual/libc shouldn't be in RDEPEND. > - a DEPEND on dev-libs/popt is missing > - DEPEND contains virtual/jre which is wrong > - doesn't compile with -readline > - instead of eautoconf and eautomake, eautoreconf should be used. > - etc. etc. > > I'm working on an ebuild but am battling myself with several automagic > problems, the worst of which being bluez.
(In reply to comment #97) > I can fix some of these issues with my ebuild, but not all. In particular, I > don't want to mess with java part. Anyway, I'm willing to help and of the gentoo > channels you can find me on #gentoo-db as a 'polyonymous'. Feel free to /msg > too. I didn't want to disturb you with /msg-ing and don't see an appropriate > channel in your /whois. Don't worry about disturbing me. You (and anyone else) can /query or /msg me at any time - just an answer is not guaranteed at all times. ;-) Anyway, I was more stubborn than pilot-link ;) and, thus, 0.12.2 is now in the tree. At the moment, it's package.mask'ed because it would cause other packages to break and so I've filed bug 190218 to the maintainers/herds of the affected packages. Please test it in any way you can and gives us feedback here. (Leaving this bug open for now.) Thanks, Michael, for your efforts here. They are appreciated - this is just one of those package versions from hell. ;-)
> Don't worry about disturbing me. You (and anyone else) can /query or /msg me at > any time - just an answer is not guaranteed at all times. ;-) I will keep that in mind ;-) > Anyway, I was more stubborn than pilot-link ;) and, thus, 0.12.2 is now in the > tree. At the moment, it's package.mask'ed because it would cause other Built fine. I use it once in a rare while, so I haven't tested it extensively, but reporting as much as I can ;-) > Thanks, Michael, for your efforts here. They are appreciated - this is just one > of those package versions from hell. ;-) You're welcome, I was just scratching my own itch and letting others benefit from itch ;-)
I've unmasked 0.12.2 earlier today. Thus: Fixed in CVS.
