Attempt to cleanup the multiple min/max definitions at the cost of some
macro redefining. The issue has been reported upstream, hopefully they
will provide a better cleanup in the future.
Either some versions of clang/libc++ are more strict or this may have been
causing issues from a while back.
Approved by: gerald (implicit)
CalculiX has alwasy been comoposed of two independent packages: ccx and
cgx. Ccx, the main tool for doing FEA, depends on fortran and currently
has to be built with GCC. Cgx, on the other hand, uses OpenGL and the C++
libraries it depends on are built with our native clang so buuilding it
with GCC causes C++ library conflicts.
By making ccx an independent port we can better focus on the specific
requirements of each tool and avoid compiler-related conflicts. This also
helps the port work again with newer versions GCC.
PR: 219290. 22043, 220434
Approved by: gerald
GCC 6 changed its default from -std=gnu++03 to -std=gnu++14, and this port's
code is not compatible with C++11. Since the project looks abandoned upstream,
it makes more sense to just continue building with an older C++ standard.
PR: 219297
Approved by: portmgr (blanket approval)
This fixes the build with GCC 6, which switched its default from -std=gnu++98
to -std=gnu++14. With this switch, it added a `operator delete(void*, size_t)'
overload and uses it for all delete calls. This does not play well with
dependencies built with other compilers (such as base clang), which use the old
operator delete overload and cause linking errors.
PR: 219484
Submitted by: fernando.apesteguia@gmail.com (maintainer)
MFH: 2017Q3
New features listed here:
http://www.dhondt.de/new_calc.htm
Rename a patch to make portlint/makepatch less noisy.
Reset the maintainer at request.
PR: 218635
Approved by: maho
Using ninja instead of make (1) can lead to significant speed ups while building.
Therefore switch from having the ninja generator opt-in to having it opt-out.
Previously cmake-ports that wanted to use ninja could set
CMAKE_NINJA=yes
now, ports that do not work with ninja can set
cmake:<existing args>,noninja
Note, that needing this should be an exception and most often points to a broken
cmake of the port.
The ports using cmake were modified
* removed USES=gmake, if ninja is used
* removed MAKE_ARGS, if ninja is used
* added the cmake-argument noninja if necessary
PR: 219629
PR: 213331
Exp-run by: antoine
Reviewed by: rakuco
Differential Revision: https://reviews.freebsd.org/D10748
Update cad/openvsp to 3.12.1 _and_ unbreak cad/openvsp. The recent
update to eigen3 broke cad/openvsp (since it is used directly) and
graphics/code-eli that is a dependency of cad/openvsp.
Until there is a fix for code-eli change the port to use the internal
eigen3 and code-eli packages bundled in the distribution.
In addition, one of the local patches was accepted upstream and is no
longer necessary.
From the announcement:
Features:
- Parasite drag buildup tool
- Full control of VSPAERO from OpenVSP GUI
- DegenGeom preview visualization
- Advanced control surface modeling
Bug Fixes:
- Advanced links not loading from file with 'Loop' policy geoms
- Crash conformal components with some file airfoils
- CompGeom when subsurfaces fall on tessellation line
- DegenGeom handling of tip caps with blended wings
- CustomGeom name matching GUI mismatch
PR: 220212
Submitted by: fernando.apesteguia@gmail.com (maintainer)
tochnog hasn't maintained the superlu_mt interface (or much else) in a long
time.
Let's just drop the option as it won't work when we bring in the modern
superlu_mt 3.1 version.
PR: 219932
Submitted by: pfg
