In D22378#491515, @james.wright_digital-chaos.com wrote:

Hi again,

   Now that the port is avialable via the pkg repo, I noticed that
trying to install the new "geany-themes" package ignores the
"geany-gtk2" package if already installed and tries to install "geany"
(the gtk3 version) instead?

   Is there anything I can change in the port to depend on either
"geany" or "geany-gtk2" packages?

Hi again,

In D22378#489434, @james.wright_digital-chaos.com wrote:

In D22378#489403, @madpilot wrote:

Thanks for this port, it looks like a good addition to the ports tree.

The port looks fine by reading it.

I'll test it and commit it once done with tests.

Thanks, that would be great!

I'm not precious over the maintainership of this port, would it make sense for you to take this one alongside all the other geany related ports which you maintain?

In D22378#489403, @madpilot wrote:

Thanks for this port, it looks like a good addition to the ports tree.

The port looks fine by reading it.

I'll test it and commit it once done with tests.

Thanks for this port, it looks like a good addition to the ports tree.

Changed RUN_DEPENDS from geany>=1.24:devel/geany to geany:devel/geany.

• poudriere builds look good
• I used the new version for an entire day without any issues

I would say just add IGNORE_FreeBSD_11= does not build between LICENSE* and *DEPENDS

I'm fine with both, IGNORE would tell people don't really need to try it, while BROKEN will still be built by the package builders.

Added IGNORE_FreeBSD_11 due to breaking build.

In D20251#484315, @james.wright_digital-chaos.com wrote:

In D20251#484295, @lwhsu wrote:

Still fails: https://gist.github.com/987197a48472bfc4790495dfdebab8de
I suggest mark IGNORE on 11 first.

What is best way to do this, I tried adding following block near the bottom of the Makfile but portlint complains about it's position no matter where I put it;

.include <bsd.port.pre.mk>
.if (${OSVERSION} < 1200000)
IGNORE= build fails on FreeBSD 11
.endif

Also should we use BROKEN= rather than IGNORE= for this case?

In D20251#484295, @lwhsu wrote:

Still fails: https://gist.github.com/987197a48472bfc4790495dfdebab8de
I suggest mark IGNORE on 11 first.

Still fails: https://gist.github.com/987197a48472bfc4790495dfdebab8de
I suggest mark IGNORE on 11 first.

I've purged my poudriere's pkg repository of 11 and kicked out a new test.

BTW, it fails to build on 11: https://gist.github.com/3d72c1ad6338657c385a86ec8bbb2526

If this is hard to fix, we could consider mark this IGNORE on 11.

In D20251#483238, @james.wright_digital-chaos.com wrote:

In D20251#483236, @lwhsu wrote:

This patch build failed in poudriere: https://gist.github.com/5919089c2e8d247342cf9a8749df7ca7

I added these files in include/ to pkg-plist and it pass the poudriere test. Is that the right fix? (these files need to be installed and installed to the right place)

The orphaned files mentioned in that poudriere log were header files I deliberately excluded from the pkg-plist, as I think they are for internal libs and headers only required during the build process, so didn't need to be included in the final package.

Updated pkg-plist to include missing header files.

In D20251#483236, @lwhsu wrote:

This patch build failed in poudriere: https://gist.github.com/5919089c2e8d247342cf9a8749df7ca7

I added these files in include/ to pkg-plist and it pass the poudriere test. Is that the right fix? (these files need to be installed and installed to the right place)

Updated pkg-plist: https://gist.github.com/cc17d08532e48039d9c5c30fc2a145f9

This patch build failed in poudriere: https://gist.github.com/5919089c2e8d247342cf9a8749df7ca7

• Updated port to 1.9.0
• Added databases/leveldb dependency
• Added devel/elfio dependency
• Added www/http-parser dependency
• Additional patches to fix compile errors
• Removed CLI option (always installed)

In D21770#475048, @lwhsu wrote:

I'll commit with minor fixes.

I'll commit with minor fixes.

Still:

====> Running Q/A tests (stage-qa)
Error: '/usr/bin/env python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-ps'
Error: '/usr/bin/env python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-tail'
Error: '/usr/bin/env python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-scp'
Error: '/usr/bin/env python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-cat'
*** Error code 1

The problem is those files under bin/ need Python, are installed even when CLI is not defined, while python:run and shebangfix are only defined when CLI is on.
So, the possible solution is either only install these files when CLI is defined, or having python:run and shebangfix defined globally.

In D20251#473528, @james.wright_digital-chaos.com wrote:

In D20251#472897, @lwhsu wrote:

Because python is used for those bin/* files, the quickest way is moving python:run to USE=, however, I tried this and poudriere gives many Error: Orphaned: include/*.
Can you check this part?

I think the previous errors (regarding python and shebangfix) were caused by a typo in CLI_USES=, I have since updated the diff and they seem to have been resolved.

Please could you run a build through poudriere and let me know if there are any outstanding issues?

In D20251#472897, @lwhsu wrote:

Tried moving shebangfix from USES= to CLI_USES= but still get the same error when running make stage-qa.

Please could you provide a suggestion how to fix this issue as I'm stumped as to what is the actual problem here.

Because python is used for those bin/* files, the quickest way is moving python:run to USE=, however, I tried this and poudriere gives many Error: Orphaned: include/*.
Can you check this part?

Makefile changes

• Fixed typo; CLI_USE= should have been CLI_USES=
• Added 'shebangfix' to CLI_USES=
• Removed extraneous whitespace (to satisfy portlint)

In D20251#472520, @james.wright_digital-chaos.com wrote:

====> Running Q/A tests (stage-qa)
Error: '/usr/local/bin/python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-ps'
Error: '/usr/local/bin/python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-tail'
Error: '/usr/local/bin/python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-cat'
Error: '/usr/local/bin/python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-scp'
*** Error code 1

I think this is because USE=python:run is only defined when CLI option enabled.

Tried moving shebangfix from USES= to CLI_USES= but still get the same error when running make stage-qa.

Please could you provide a suggestion how to fix this issue as I'm stumped as to what is the actual problem here.

====> Running Q/A tests (stage-qa)
Error: '/usr/local/bin/python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-ps'
Error: '/usr/local/bin/python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-tail'
Error: '/usr/local/bin/python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-cat'
Error: '/usr/local/bin/python' is an invalid shebang you need USES=shebangfix for 'bin/mesos-scp'
*** Error code 1

I think this is because USE=python:run is only defined when CLI option enabled.

In D20251#472087, @lwhsu wrote:

@james.wright_digital-chaos.com: do you want to also submit the patch to the upstream?

I can build this port in poudriere fine. I think previous error was because of patching issue.

In D20251#472124, @arrowd wrote:

Sorry for forgetting about this. I didn't added it to my TODO list.

This doesn't build on FreeBSD 12 due to following error:

libtool: compile:  c++ -DPACKAGE_NAME=\"mesos\" -DPACKAGE_TARNAME=\"mesos\" -DPACKAGE_VERSION=\"1.8.1\" "-DPACKAGE_STRING=\"mesos 1.8.1\"" -DPACKAGE_BUGREPORT=\"\" -DPACKAGE_URL=\"\" -DPACKAGE=\"mesos\" -DVERSION=\"1.8.1\" -DSTDC_HEADERS=1 -DHAVE_SYS_TYPES_H=1 -DHAVE_SYS_STAT_H=1 -DHAVE_STDLIB_H=1 -DHAVE_STRING_H=1 -DHAVE_MEMORY_H=1 -DHAVE_STRINGS_H=1 -DHAVE_INTTYPES_H=1 -DHAVE_STDINT_H=1 -DHAVE_UNISTD_H=1 -DHAVE_DLFCN_H=1 -DLT_OBJDIR=\".libs/\" -DHAVE_CXX11=1 -DHAVE_PTHREAD_PRIO_INHERIT=1 -DHAVE_PTHREAD=1 -DHAVE_FTS_H=1 -DHAVE_APR_POOLS_H=1 -DHAVE_LIBAPR_1=1 -DHAVE_BOOST_VERSION_HPP=1 -DHAVE_LIBCURL=1 -DHAVE_GLOG_LOGGING_H=1 -DHAVE_ARCHIVE_H=1 -DHAVE_GOOGLE_PROTOBUF_MAP_H=1 -DHAVE_EV_H=1 -DHAVE_PICOJSON_H=1 -DHAVE_RAPIDJSON_RAPIDJSON_H=1 -DHAVE_LIBSASL2=1 -DHAVE_OPENSSL_SSL_H=1 -DHAVE_SVN_VERSION_H=1 -DHAVE_LIBSVN_SUBR_1=1 -DHAVE_SVN_DELTA_H=1 -DHAVE_LIBSVN_DELTA_1=1 -DHAVE_ZLIB_H=1 -DHAVE_LIBZ=1 -DHAVE_GRPCPP_GRPCPP_H=1 -DHAVE_ZOOKEEPER_H=1 -I. -DLIBDIR=\"/usr/local/lib\" -DPKGLIBEXECDIR=\"/usr/local/libexec/mesos\" -DPKGDATADIR=\"/usr/local/share/mesos\" -DPKGMODULEDIR=\"/usr/local/lib/mesos/modules\" -I../include -I../include -I../include/mesos -D__STDC_FORMAT_MACROS -I../3rdparty/concurrentqueue-7b69a8f -I../3rdparty/elfio-3.2 -I../3rdparty/leveldb-1.19/include -I../3rdparty/libprocess/include -I../3rdparty/nvml-352.79 -I../3rdparty/stout/include -I/usr/local/include/subversion-1 -I/usr/local/include -I/usr/local/include -I/usr/local/include/apr-1 -I/usr/local/include/apr-1.0 -isystem /usr/local/include -I/usr/local/include -I/usr/local/include -I/usr/local/include -I/usr/local/include -I/usr/local/include -I/usr/local/include -I/usr/local/include -I/usr/local/include/zookeeper -D_THREAD_SAFE -pthread -Wall -Wsign-compare -Wformat-security -fstack-protector-strong -fPIC -O2 -pipe -fstack-protector-strong -isystem /usr/local/include -fno-strict-aliasing -isystem /usr/local/include -Wno-unused-local-typedef -Wno-inconsistent-missing-override -std=c++11 -stdlib=libc++ -DGTEST_HAS_PTHREAD -D_WITH_GETLINE -Wno-sentinel -MT master/libmesos_no_3rdparty_la-constants.lo -MD -MP -MF master/.deps/libmesos_no_3rdparty_la-constants.Tpo -c master/constants.cpp  -fPIC -DPIC -o master/.libs/libmesos_no_3rdparty_la-constants.o
master/allocator/mesos/hierarchical.cpp:1649:5: error: use of undeclared identifier 'random_shuffle'; did you mean 'std::random_shuffle'?
    random_shuffle(resourceVector.begin(), resourceVector.end());
    ^~~~~~~~~~~~~~
    std::random_shuffle
/usr/include/c++/v1/algorithm:3168:1: note: 'std::random_shuffle' declared here
random_shuffle(_RandomAccessIterator __first, _RandomAccessIterator __last)
^

Sorry for forgetting about this. I didn't added it to my TODO list.

@arrowd: any update on this?
@james.wright_digital-chaos.com: do you want to also submit the patch to the upstream?

Looks good to me. If no one objects, I'll commit this after giving it a run in poudriere.

Updated to 1.8.1
Added patch to fix compilation error in src/master/allocator/mesos/hierarchical.cpp

Please could take a look over the updated diff, as I appreciate the constructive feedback! :-)

Makefile changes;

• Fixed incorrect braces for PORTVERSION
• Added USES= libtool
• Added USES= shebangfix and SHEBANG_FILES= (although make stage-qa still seems unhappy with these files?)
• Added USE_LDCONFIG=yes
• Added option for installing CLI tools (which require the python runtime)
• Added option for installing Deploy scripts (which should now install the *.template files as samples)
• Added --disable-werror flag as suggested

Added SUBDIR += apache-mesos to sysutils/Makefile, as requested.

This needs to be hooked into sysutils/Makefile

Are there any further changes that this port needs, or can it be merged?

Updated Makefile based on suggestions;

Fixed typo in pkg-descr.

Updated existing devel/elfio port rather than creating a new one (devel/elfio3).

In D19879#427686, @tobik wrote:

Added a new port for ELFIO v3.x as it is incompatible with the existing v1.x.x

devel/elfio is unmaintained and no ports depend on it. Why not update it to 3.x instead of adding a new port? Is there a reason devel/elfio needs to stay at 1.x?

Updated Makefile as suggested by tobik;
Removed ALL_TARGET=
Added NO_BUILD= yes
Added TEST_TARGET= check
Moved 'OPTIONS' variables down for correct ordering

Added a new port for ELFIO v3.x as it is incompatible with the existing v1.x.x

Added elfio3 directory to devel/Makefile
Added myself as maintainer

Added picojson subdirectory to devel/Makefile
Added myself as maintainer of this port

The port needs to be hooked into devel/Makefile

Sorry, I had changes in progress when this came in and then life got busy. This one slipped out of my buffer. I'll commit shortly.

ping @brooks

Committed already