libboost-static-devel rpm only contains two files: $ urpmq -l libboost-static-devel /usr/lib/libboost_exception.a /usr/lib/libboost_test_exec_monitor.a Whereas it should contain all the modules. EG with a debian: ls libboost_*a libboost_atomic.a libboost_locale.a libboost_mpi_python.a@ libboost_serialization.a libboost_chrono.a libboost_log.a libboost_mpi_python-py27.a libboost_signals.a libboost_context.a libboost_log_setup.a libboost_mpi_python-py34.a libboost_system.a libboost_coroutine.a libboost_math_c99.a libboost_prg_exec_monitor.a libboost_test_exec_monitor.a libboost_date_time.a libboost_math_c99f.a libboost_program_options.a libboost_thread.a libboost_exception.a libboost_math_c99l.a libboost_python.a@ libboost_timer.a libboost_filesystem.a libboost_math_tr1.a libboost_python-py27.a libboost_unit_test_framework.a libboost_graph.a libboost_math_tr1f.a libboost_python-py34.a libboost_wave.a libboost_graph_parallel.a libboost_math_tr1l.a libboost_random.a libboost_wserialization.a libboost_iostreams.a libboost_mpi.a libboost_regex.a
Assignee: bugsquad => zen25000
CC: (none) => zen25000Assignee: zen25000 => shlomifSource RPM: (none) => boost-1.55.0-7.mga5.src.rpm
Created attachment 5675 [details] Proposed patch to fix the problem against the svn. This proposed patch fixes the problem, but will have to wait until Mageia 5 is released.
Nice, but why not push it now?
(In reply to Barry Jackson from comment #2) > Nice, but why not push it now? Quite a few packages depend on boost and this may break them now in the last minute - close to the release.
(In reply to Shlomi Fish from comment #3) > (In reply to Barry Jackson from comment #2) > > Nice, but why not push it now? > > Quite a few packages depend on boost and this may break them now in the last > minute - close to the release. I very much doubt it as there is no version change and only missing libs are added. Obviously nothing depends on those. I did test a rebuild in iurt with your changes and it does build OK. To push later as an update to Mga5 will involve qa who are already overstretched, but it's your call :)
Thanks Barjac for the insight. This bug is fixed in boost-1.55.0-8.mga5 that was just uploaded to Cauldron.
Status: NEW => RESOLVEDResolution: (none) => FIXED