Summary: | x11-libs/lesstif removal | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Arvid Norlander <anmaster> |
Component: | [OLD] Unspecified | Assignee: | Gentoo TreeCleaner Project <treecleaner> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | achurch+gentoo, anmaster, coldwind, gentoo, jer, maintainer-needed, ulm, webmaster |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | AMD64 | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Bug Depends on: | 204265 | ||
Bug Blocks: | 204249 |
Description
Arvid Norlander
2007-09-23 11:20:28 UTC
I'm also having the same problem. Here's the (relevant) error output I got: In file included from /usr/include/Xm/XmP.h:1647, from ScrolleGEP.h:38, from ScrolledGE.C:34: /usr/include/X11/VendorP.h:87: error: previous declaration of 'VendorShellClassRec vendorShellClassRec' with 'C++' linkage /usr/include/Xm/VendorSP.h:58: error: conflicts with new declaration with 'C' linkage make[2]: *** [ScrolledGE.o] Error 1 BTW, while I'm no authority on this matter, I think that the gentoo bug killers only care about the errors. However, definitely don't go quoting me on that. Below is my output for 'emerge --info': Portage 2.1.3.9 (default-linux/x86/2007.0/desktop, gcc-4.2.0, glibc-2.6.1-r0, 2.6.22-gentoo-r5 i686) ================================================================= System uname: 2.6.22-gentoo-r5 i686 Genuine Intel(R) CPU T2050 @ 1.60GHz Timestamp of tree: Mon, 24 Sep 2007 04:50:01 +0000 distcc 2.18.3 i686-pc-linux-gnu (protocols 1 and 2) (default port 3632) [enabled] ccache version 2.4 [enabled] app-shells/bash: 3.2_p17-r1 dev-java/java-config: 1.3.7, 2.0.33-r1 dev-lang/python: 2.4.4-r5, 2.5.1-r2 dev-python/pycrypto: 2.0.1-r6 dev-util/ccache: 2.4-r7 sys-apps/baselayout: 1.12.10-r4 sys-apps/sandbox: 1.2.17 sys-devel/autoconf: 2.13, 2.61-r1 sys-devel/automake: 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2, 1.10 sys-devel/binutils: 2.18 sys-devel/gcc-config: 1.4.0-r2 sys-devel/libtool: 1.5.24 virtual/os-headers: 2.6.22-r2 ACCEPT_KEYWORDS="x86 ~x86" CBUILD="i686-pc-linux-gnu" CFLAGS="-O2 -mtune=i686 -pipe" 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/fonts/fonts.conf /etc/gconf /etc/php/apache2-php5/ext-active/ /etc/php/cgi-php5/ext-active/ /etc/php/cli-php5/ext-active/ /etc/revdep-rebuild /etc/terminfo /etc/texmf/web2c /etc/udev/rules.d" CXXFLAGS="-O2 -mtune=i686 -pipe" DISTDIR="/usr/portage/distfiles" FEATURES="ccache distcc distlocks metadata-transfer parallel-fetch sandbox sfperms strict unmerge-orphans userfetch" GENTOO_MIRRORS="http://gentoo.osuosl.org/ ftp://distro.ibiblio.org/pub/linux/distributions/gentoo/" MAKEOPTS="-j5" 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 --filter=H_**/files/digest-*" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" SYNC="rsync://rsync.namerica.gentoo.org/gentoo-portage" USE="X a52 aac aalib acl acpi adns alsa apache2 arts bash-completion bcmath berkdb bindist bitmap-fonts bzip2 cairo caps cddp cdr cjk clamav cli cracklib crypt cscope ctype cups curl curlwrappers dbase dbm dbus dga directfb dri dts dvd dvdr dvdread eds emacs emboss encode evo examples exif expat fam fastcgi fbcron ffmpeg fftw firefox flac foomaticdb fortran freetds ftp gd gdbm geoip ggi gif glut gmp gnome gnustep gnutls gpm graphviz gstreamer gtk gtkhtml hal howl icc iconv icq idn imagemagick imap imlib informix innodb ipod ipv6 isdnlog jabber jack java javascript jbig joystick jpeg jpeg2k junit kde kdexdeltas kerberos krb4 ladspa lapack lcms ldap leim lesstif libcaca libg++ libnotify libsamplerate libwww lirc lm_sensors mad maildir matroska matrox mbox mcal midi mikmod mime mmx mono mp3 mpeg mpi mplayer mpm-threadpool msn mudflap mule musepack musicbrainz mysql mysqli nas ncurses nls nntp nocd nptl nptlonly offensive ofx ogg openexr opengl openmp osc oscar oss pam pcntl pcre pdf perl php plotutils png postfix ppds pppd profile python qt3 qt3support qt4 quicktime rdesktop readline reflection sdl session sharedext shorten smp snmp soap sockets source speex spell spl sse sse2 ssl startup-notification svg svga symlink sysvipc tcpd tetex theora threads tidy tiff tokenizer truetype truetype-fonts type1-fonts unicode usb vcd vhosts videos vorbis win32codecs x86 xcomposite xface xine xml xorg xosd xpm xprint xscreensaver xsl xulrunner xv xvid yahoo zlib" ALSA_CARDS="ali5451 als4000 atiixp atiixp-modem bt87x ca0106 cmipci emu10k1 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="keyboard mouse evdev" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" USERLAND="GNU" VIDEO_CARDS="apm ark chips cirrus cyrix dummy fbdev glint i128 i740 i810 imstt mach64 mga neomagic nsc nv r128 radeon rendition s3 s3virge savage siliconmotion sis sisusb tdfx tga trident tseng v4l vesa vga via vmware voodoo" Unset: CTARGET, EMERGE_DEFAULT_OPTS, INSTALL_MASK, LANG, LC_ALL, LDFLAGS, LINGUAS, PORTAGE_COMPRESS, PORTAGE_COMPRESS_FLAGS, PORTAGE_RSYNC_EXTRA_OPTS, PORTDIR_OVERLAY This looks like a problem with lesstif-0.95.0; I get the same VendorShellClassRec error when compiling xpdf. While I don't know whether it's the correct fix, moving the line #include <X11/ShellP.h> to below the #ifdef __cplusplus extern "C" { #endif in /usr/include/Xm/VendorSP.h solved the problem for me. *** Bug 204246 has been marked as a duplicate of this bug. *** (In reply to comment #2) > This looks like a problem with lesstif-0.95.0 Does 0.94.4 work, or is it broken as well? (In reply to comment #4) > (In reply to comment #2) > > This looks like a problem with lesstif-0.95.0 > > Does 0.94.4 work, or is it broken as well? > i use lesstif-0.94.4 and xpdf compilation also fail. Maybe try w/ openmotif instead, then... (In reply to comment #6) > Maybe try w/ openmotif instead, then... > this works ! Thanks. (In reply to comment #7) > (In reply to comment #6) > > Maybe try w/ openmotif instead, then... > this works ! Thanks. Well, one more reason to get rid of lesstif, the broken motif-config thingy & co. *** Bug 199554 has been marked as a duplicate of this bug. *** Going away, nothing to fix here... Reopening for removal; will be done as soon as Bug 204265 is finished. This would be a disaster for me, I need some software that doesn't compile with openmotif, only with lesstif. Therefore I hope you do not remove the package. (In reply to comment #12) > This would be a disaster for me, I need some software that doesn't compile with > openmotif, only with lesstif. Therefore I hope you do not remove the package. You are welcome to maintain it in your overlay. The reasons for removal have been states numerous times, see Bug 204249 and the dependent bugs. Reassigning to treecleaners. Package.masked, since lesstif is not installable anyway. Here are some thoughts: In case that this package should be revived, it IMHO makes only sense with mutual blockers between openmotif and lesstif. See bug 29388, bug 117458, and the bugs depending on bug 204249 for the reasons why motif-config cannot work; no need to repeat it once again. However, we already know that many packages in the tree do not work properly with lesstif. It is not-quite Motif 2.1, incomplete, orders of magnitude more buggy, and even uglier than the original (one wouldn't think that this is possible ;-) ). So, about the only reason to prefer lesstif could be the licence issue. But in this case, it would be paradoxical to have openmotif installed alongside lesstif, and we could have the mutual blockers as well. In short: I am for removing it (but I don't have a vote). I don't see why slotted versions does not work. After all, don't they have different SONAMEs? Shouldn't that be enough? (In reply to comment #16) > I don't see why slotted versions does not work. After all, don't they have > different SONAMEs? Shouldn't that be enough? Maybe you should re-read the previous comment and act accordingly before commenting here. (In reply to comment #16) > I don't see why slotted versions does not work. After all, don't they > have different SONAMEs? Shouldn't that be enough? No, since the libraries are not everything, there are also header files and binaries (like mwm). See again the references that I have cited in comment #15. And it doesn't work to have both packages installed at the same time; the following line from openmotif-2.2.3-r9.ebuild pretty much summarizes it: emake -j1 || die "make failed, if you have lesstif installed removed it, compile openmotif and recompile lesstif" (In reply to comment #17) > Maybe you should re-read the previous comment and act accordingly before > commenting here. > I did read those bugs you mentioned but I'm afraid it seems to mostly be different developers fighting each other, rather than trying to fix issues. As far as I know different versions of a library can co-exist. ABI is incompatible sure, but portage should record the needed information for that IMO. Lesstif and openmotif got partly incompatible APIs. That is much worse and means having to provide alternatives. After all, GCC broke the C++ ABI a few times (I think). No one objected to that. Both gcc 3.x and gcc 4.x are still provided, slotted. (In reply to comment #18) > No, since the libraries are not everything, there are also header files and > binaries (like mwm) Uhh, IMO a window manager does not belong in same package as a programming library for general use. Should be a separate package. For header files you got a point, but other packages manages, for example, wx-config --version=2.6 --unicode=yes and so on. > And it doesn't work to have both packages installed at the same time; the > following line from openmotif-2.2.3-r9.ebuild pretty much summarizes it: > emake -j1 || die "make failed, if you have lesstif installed removed it, > compile openmotif and recompile lesstif That looks broken. OpenMotif is clearly broken if it can't handle lesstif being installed. It should be reported upstream IMO. (In reply to comment #19) > That looks broken. OpenMotif is clearly broken if it can't handle lesstif being > installed. It should be reported upstream IMO. Please, stop producing more noise on this bug before doing a basic research at least. Seriously - once and for all. I'm going to restrict the bug to prevent totally unproductive noise otherwise. These two packages are basically two implementations of the exact same thing. Complaining that they can't be installed at the same time is like complaining that cdrkit and cdrtools can't be installed at the same time. They are *not* binary compatible -> they cannot co-exist on the system in a sane way. Tested and it does NOT work. Please, save us from more completely uninformed comments. the only way wxGTK installing two different versions side by side works is it was _designed that way_. The headers are installed in version separated directories and written in a way that the correct version will be used. slotted gcc versions work because they are _designed to_. and in both cases, you're talking about different versions of the same project. but here we're talking about two completely different projects. they simply cannot be installed together. this isn't an uncommon thing. most packages providing a virtual can't be installed together. What is the proper approach for contesting something like this because you consider the non-libre alternative morally wrong? (Besides switching to Debian) (In reply to comment #22) > What is the proper approach for contesting something like this because > you consider the non-libre alternative morally wrong? (Besides switching > to Debian) Please read comment #13 and comment #15 again. Or find a developer for us who is going to maintain the lesstif package. Preferably, it should be someone who has never looked at the openmotif source code (because of the licence issue: he might be in the situation that he has to submit patches upstream). (In reply to comment #22) > What is the proper approach for contesting something like this because you > consider the non-libre alternative morally wrong? (Besides switching to Debian) I think switching to Debian is an excellent solution... seriously. You've chosen a a wrong distribution if you consider 'non-libre' software to be immoral. so...? Remove it or what? About time; noone can install it. Let's vote then, treecleaners, +1 from me +2 - voting doesn't make much sense here, it is not installable at all because it depends on its own blocker (motif-config blocks lesstif now). I was able to install x11-libs/lesstif-0.95.0 with paludis on an amd64 machine. openmotif is no Free Software and that's why no option for me. (In reply to comment #29) > I was able to install x11-libs/lesstif-0.95.0 with paludis on an amd64 machine. If paludis ignores the blockers in place, you should file a bug about it. > openmotif is no Free Software and that's why no option for me. So maintain it in your own overlay. Been explained over and over again, this doesn't bring anything new here. Since nothing is happening here, I have now committed the ebuild that I had in my private overlay since three months: *lesstif-0.95.0-r1 (10 May 2008) 10 May 2008; Ulrich Mueller <ulm@gentoo.org> +files/lesstif-0.95.0-vendorsp-cxx.patch, +lesstif-0.95.0-r1.ebuild: Ebuild completely rewritten: SLOTting is finally gone. Install libraries, include files, and man pages in their standard locations. No more need for motif-config and profile manipulation. Libtool tweaks to fix linkage against already installed version are no longer needed. Fix issue with inclusion of VendorSP.h in C++ code, original issue of bug 193505, patch extracted from upstream CVS. Update LICENSE: add libXpm; add GPL-2 and FVWM for mwm. Dependencies scrutinised and fixed: libXft/freetype is needed also at run time, remove unnecessary libXaw and perl, remove redundant printproto. Block against openmotif and motif-config. Of course this fixes none of the issues that packages are broken with LessTif. Therefore I suggest that we leave it in package.mask indefinitely, to make clear that it is an unsupported package and that users are on their own if they use it. Reassigning to maintainer-needed again. (Treecleaners, take it back if you disagree.) I shall remove the old revisions 0.94.4 and 0.95.0 together with removal of x11-libs/motif-config (bug 117057) in a few days. This implies dropping of all keywords to unstable, but the stable version isn't installable anyway. Another year has passed, so probably it's time to reopen this bug. Last upstream release was three years ago. Reassigning to treecleaners, since I'd like to know your opinion on this. Vote for removal unless someone is willing to keep it maintained. I will update the mask when the treecleaner queue gets processed in a few days. 30 days left for this one. (In reply to comment #2) > This looks like a problem with lesstif-0.95.0; I get the same > VendorShellClassRec error when compiling xpdf. While I don't know whether it's > the correct fix, moving the line > #include <X11/ShellP.h> > to below the > #ifdef __cplusplus > extern "C" { > #endif > in /usr/include/Xm/VendorSP.h solved the problem for me. This fix works perfectly, thanks for the help > # Ulrich Mueller <ulm@gentoo.org> (10 May 2008) # LessTif is unsupported since it has open issues with several packages. # See bug 193505 and bug 204249. You are on your own here. # Jeremy Olexa <darkside@gentoo.org> (30 May 2009) # Treecleaners will remove in 30 days. bug 193505 x11-libs/lesstif removed from tree, fixed |