From perl-cleaner script: SONAME="$(scanelf -qBS "$(realpath /usr/lib/libperl.so 2>/dev/null )" | awk '{ print $1 }')" As a result it fails to determine perl soname on systems with SYMLINK_LIB=no profile and after each run it complains: * It seems like perl-cleaner had to rebuild some packages. * * If you have just updated your major Perl version (e.g. from 5.20.2 to 5.22.0), * and have run perl-cleaner _after_ that update, then this means most likely * that these packages are buggy. Please file a bug on http://bugs.gentoo.org/ and * report that perl-cleaner needed to reinstall the following list: * app-editors/vim:0 app-editors/gvim:0 mail-client/claws-mail:0 mail-mta/exim:0 media-gfx/imagemagick:0 net-im/pidgin:0 net-irc/hexchat:0 net-nds/openldap:0 sys-apps/texinfo:0 x11-terms/rxvt-unicode:0 *
I think the best solution of having many reports is to make only one bug with the whole list of applications involved in the case. Don't you think so, devs? Should we merge this with the earliest non-resolved report about such type of case into bug 589874 ?
I guess this tool shouldn't be really necessary, when the common case is covered by := deps and the uncommon case is broken.
(In reply to Michał Górny from comment #2) > I guess this tool shouldn't be really necessary, when the common case is > covered by := deps and the uncommon case is broken. I will rename my bug accordingly, see https://bugs.gentoo.org/show_bug.cgi?id=686168
*** Bug 686168 has been marked as a duplicate of this bug. ***
@perl team, is perl-cleaner still useful or can be treecleaned finally? If still needed, we would need to get this fixed as new profiles went to stable and people will start to use them
(In reply to Pacho Ramos from comment #5) > @perl team, is perl-cleaner still useful or can be treecleaned finally? If > still needed, we would need to get this fixed as new profiles went to stable > and people will start to use them Sigh, "needed" is a complicated word. Mostly due to portage bugs, yes, somewhat. I'll have a look over the next days.
Sigh, the patch in my "duplicate" bug https://bugs.gentoo.org/686168 is forgotten and not applied : https://bugs.gentoo.org/attachment.cgi?id=577078&action=diff
Should be fixed in 2.30. There were several nameless horrors in that code line.
