On Thu, 2012-11-29 at 10:04 +0100, Rob Godfrey wrote: > All, > > I know this has been an issue for a while, but I'm struggling to understand > why we have not yet fixed the issue whereby the C++ automake build fails on > recent Ubuntu releases. > > From a clean checkout of the 0.20 branch, a build on my x86_64 Ubuntu > 12.04.1 machine, a build will end with the following error: > > libtool: link: g++ -Werror -pedantic -Wall -Wextra -Wno-shadow > -Wpointer-arith -Wcast-qual -Wcast-align -Wno-long-long > -Wvolatile-register-var -Winvalid-pch -Wno-system-headers > -Woverloaded-virtual -DBOOST_FILESYSTEM_VERSION=2 > -DQPIDD_MODULE_DIR=\"/usr/local/lib/qpid/daemon\" > -DQPIDD_CONF_FILE=\"/usr/local/etc/qpidd.conf\" > -DQPIDC_CONF_FILE=\"/usr/local/etc/qpid/qpidc.conf\" -g -O2 -o .libs/qpidd > qpidd-qpidd.o posix/qpidd-QpiddBroker.o ./.libs/libqpidbroker.so > ./.libs/libqpidcommon.so > /usr/bin/ld: posix/qpidd-QpiddBroker.o: undefined reference to symbol > 'boost::program_options::validate(boost::any&, > std::vector<std::basic_string<char, std::char_traits<char>, > std::allocator<char> >, std::allocator<std::basic_string<char, > std::char_traits<char>, std::allocator<char> > > > const&, > std::basic_string<char, std::char_traits<char>, std::allocator<char> >*, > int)' > /usr/bin/ld: note: 'boost::program_options::validate(boost::any&, > std::vector<std::basic_string<char, std::char_traits<char>, > std::allocator<char> >, std::allocator<std::basic_string<char, > std::char_traits<char>, std::allocator<char> > > > const&, > std::basic_string<char, std::char_traits<char>, std::allocator<char> >*, > int)' is defined in DSO /usr/lib/libboost_program_options.so.1.48.0 so try > adding it to the linker command line > /usr/lib/libboost_program_options.so.1.48.0: could not read symbols: > Invalid operation > > Fraser Adams raised issue QPID-3723[1] back in January about this issue, > and has a supplied patch[2] (attached to the JIRA) which solves the issue. > > Since the Apache Jenkins instances are running on Ubuntu. we are actually > applying that patch to our own source code before running the CI build. > > Can someone from the C++ side of things explain why we can't fix this / > haven't responded to this issue in nearly a year? > > Thanks, > Rob > > [1] https://issues.apache.org/jira/browse/QPID-3723 > [2] > https://issues.apache.org/jira/secure/attachment/12509530/fix-explicit-link-problem.patch
Gordon fixed this yesterday, good timing :) You're right the delay is unreasonable. I think we should set a default assignee for all the components so at least we'll have someone to yell at. Anyone know how to do that? How to get volunteers is another question, maybe pick the person who has been assigned the most JIRAs for the component in the past? Cheers, Alan. --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
