add fcl/0.6.1

[SpaceIm]

Specify library name and version: fcl/0.6.1

• I've read the guidelines for contributing.
• I've followed the PEP8 style guides for Python code in the recipes.
• I've used the latest Conan client version.
• I've tried at least one configuration locally with the
  conan-center hook activated.

#621

[SpaceIm]

Just to mention two things;

• MinGW builds might fail, even with -Wa,-mbig-obj (in Debug mode there is little hope).
• Does self.deps_cpp_info["octomap"].version return the true version of octomap in the dependency tree (eventually overriden by consumer?)

[conan-center-bot]

Some configurations of 'fcl/0.6.1' failed in build 1 (998d20d53e39150756f1bc0e5534fa32aaa4b6d0):

• Windows x86_64, Debug, Visual Studio 14, MTd. Options: fcl:shared-False
  • Log download
  • Log error Download
  • Profile
• Access to all the logs

[conan-center-bot]

All green in build 3 (c3875eb257e069d95a933748982034b6ef207ffe)! 😊

• fcl/0.6.1: Generated 120 packages. All logs here

[conan-center-bot]

All green in build 4 (e08bd64652372fcdcd6bae3d536cbce4cef2f2e5)! 😊

• fcl/0.6.1: Generated 124 packages. All logs here

[SSE4]

1. MinGW is okay to be added at later point. for now, you may just throw ConanInvalidConfiguration, or leave as is
2. yes, I think so, e.g. the same approach worked well in OpenCV recipe

[SpaceIm]

It works for MinGW as static lib, so point 1 is not longer relevant.

I've also a huge patch to fix shared lib on Windows (MinGW and Visual Studio), but honestly, I can't put that in the recipe (it exceeds size limit for one recipe). I'll submit a PR to upstream. It was hard and tedious to fix.

[madebr]

I've also a huge patch to fix shared lib on Windows (MinGW and Visual Studio), but honestly, I can't put that in the recipe (it exceeds size limit for one recipe). I'll submit a PR to upstream. It was hard and tedious to fix.

That would be nice!
In my experience, adding a shared build to the test matrix (.appveyor.yml) will avoid breaking this in the future and makes upstream more willing to accept these patches.

[SpaceIm]

I've submitted a PR: flexible-collision-library/fcl#461
As is, it will be rejected, because it bumps CMake minimum version to 3.7. Could be modified a little bit with a less elegant solution in order to preserve CMake version.

[madebr]

I've submitted a PR: flexible-collision-library/fcl#461
As is, it will be rejected, because it bumps CMake minimum version to 3.7. Could be modified a little bit with a less elegant solution in order to preserve CMake version.

Great job!
Why does it need 3.7? The reviewers are talking about supporting 3.5, which is not much less.

In a similar situation, I had to avoid using link_options/target_link_options, and use CMAKE_EXE_LINKER_FLAGS/CMAKE_SHARED_LINKER_FLAGS and similar.
I always added a FIXME about minimum required cmake version.

[SpaceIm]

Because I use an option of generate_export_header introduced in 3.7 (CUSTOM_CONTENT_FROM_VARIABLE ) in order to inject a bunch of macros (those related to explicit instantiation declaration and definition) at the end of export file:
https://cmake.org/cmake/help/v3.7/module/GenerateExportHeader.html

https://github.com/flexible-collision-library/fcl/blob/3b3a1245d9a4fd50f663e144a5909000db84a2c3/include/fcl/CMakeLists.txt#L39

fmt has similars macros, but fmt doesn't use at all generate_export_header, while fcl does.

[conan-center-bot]

All green in build 5 (ea8c956bab1a0c84536a2b906f3830ee6fc232d0)! 😊

• fcl/0.6.1: Generated 124 packages. All logs here