Changes in 'mode'; Manual single thread in examples for docs
Try minor optimization OMP
Minor improve in 'mode'
Compatibility changes with Qt6
Improve in 'mode' functions
Add 'mode' to find the mode of 1D or 2D waveguide
FindPythonInterp and FindPythonLibs modules are removed
Fixes for recent changes in Qt5; Start to add 'mode' for determioning 1d or 2d wave-guide mode (not finished)
I have the same error. What needs to be done for the compilation to complete successfully?
Bugfix for wrong position of left edge of 'bbox'
Bugfix for using dashing in hex color with 'd'
Add HistL() for histogram by linear interpolation; Add operators for mglDataC; Start adding multi-fontface features (not finished)
Add support for Gtk 4 widget class
Update fonts and set STIX2 (STIXTwo) font as default. Add CJK fonts. Note, changes are not well tested -- you can use old STIX in the case of any problems
Bugfix in make_font for correct numbers in wire font
Changes in config files
MathGL: some optimizations to Qt rework and fixes to JSON compile.
MathGL: some fixes to Qt rework.
MathGl: selecting, using and searching for various Qt versions reworked.
Fix and speed up sinfft and cosfft for complex numbers
Build breaks on armv7 with sse-related errors
Fix sinfft to be in the standard form (i.e. without zero values at edges)
Add jump functionality to ODE solver (not tested)
Add some utility/interface functions in mglGraph
Minor fixes for reading complex data and NAN handling
With libharu of 2.4.3 #include <hpdf_version.h> is not required.
With libharu of 2.4.3 #include <hpdf_version.h> is not required.</hpdf_version.h>
FYI - I've filed https://github.com/libharu/libharu/pull/247 to request that hpdf_version.h be re-added to hpdf.h. No idea if it will be accepted or not. Directly including it though would still be required to support 2.4.0-.2.
Following Stefan's advice, here is what I ended up doing, so that mathgl builds against libharu older as well as newer than or at version 2.4.0: Index: mathgl-8.0.1/src/prc.cpp =================================================================== --- mathgl-8.0.1.orig/src/prc.cpp +++ mathgl-8.0.1/src/prc.cpp @@ -36,6 +36,7 @@ #include <hpdf.h> #include <hpdf_u3d.h> #include <hpdf_annotation.h> +#include <hpdf_version.h> #endif // MGL_HAVE_PDF @@ -959,7 +960,12 @@ void MGL_EXPORT mgl_write_prc(HMGL...
This should be made conditional, based on HPDF_VERSION_MAJOR/MINOR, available from hpdf_version.h.
Compilation failure against libharu 2.4.x: too few arguments to function HPDF_Page_Create3DAnnot
Add power deposition for QO2d if k0<0; Minor improvements
Minor fix in PDE solver
Bugfixes in quasioptical PDE solver
Some speed improve in formula parsing
For anyone facing similar issue, to use static build of mathGL in your application, EXPORT can be disabled from your CMake target as target_compile_definitions(${CMAKE_PROJECT_NAME} PRIVATE mgl_EXPORTS=0)
Were you able to solve it. I'm too facing similar issue, bin contains libmgl-wx.dll and libmgl.dll, even though static flag is set
Were you able to solve it. I'm too facing similar issue, bin contains libmgl-wx.dll and libmgl.dll, even though static flag is set
Optimization of textual formula evaluation
Extend preparsed formula evaluation to the case of data arrays (NOT TESTED)
Add duplicate initialization, which later one somehow become missed in current gcc with -O3 optimization
Move definitions to header; Minor fix; Spelling
Add robust approximation for 'fit' and 'fits'
There is minor update v.8.0.1 which contain increase accuracy at line segment skipping, changes in SOVERSION numbering, minor spelling fix.
Revision v.8.0.1
Make MathGL relocatable
Changes in SOVERSION usage; Minor spelling fix (thanks to Rafael Laboissière)
Increase accuracy at line segment skipping
MathGL: version 8.0 (Stable) released
Fix for cmake 3.22 or newer, and .pyc filename version numbering.
Update translation and website
Update version information
Minor warnings remove
Disable simd for openmp in current version
Update HDF5 interface
Final changes for quad and trig cutting at axis borders
Finish with quad and trig cutting at axis boundaries; Fix minor TriCont bug
Removed Qt4 compilation
Continue with quad and trig cutting (not well tested use rev.1693)
Add advanced quad and trig cutting at axis boundary (quad is OK, but something wrong with trig)
Compatibility changes for MSVC; Add option argument for colorbar C-functions; Add variant of 'clf' and 'background' functions which handle alpha value for colors
Replace obsolete functions for Qt5.15
Add accurate line segment cutting at axis border crossing
I/O error : Permission denied
Just follow instruction "try rerunning as root" if you install mathgl globally.
There are a number of deprecated functions in new Qt5 version. To use new version of Qt5 it I need to exclude Qt4 interface effectively. This will be done in v.2.5.1 by somehow.
warning: 'delta' is deprecated: Use angleDelta()
warning: result of comparison of constant 128 with expression of type 'char' is always false
This is done on purpose (not a bug) to bypass the situation using an unsigned char instead of signed one.
I/O error : Permission denied
warning: 'delta' is deprecated: Use angleDelta()
warning: result of comparison of constant 128 with expression of type 'char' is always false
MacPorts MathGL maintainer here. This morning, I noticed that MathGL v2.5 supports some Armadillo functionallity. However when building under macOS Big Sur I get In file included from /opt/local/include/armadillo:87: /opt/local/include/armadillo_bits/arma_forward.hpp:152:12: error: unknown type name 'constexpr' static constexpr bool is_row = false; ... with the suggestion /opt/local/include/armadillo_bits/compiler_check.hpp:72:4: error: " C++11 compiler required; enable C++11 mode in your compiler,...
MathGL: version 2.5 (Stable) released
Version 2.5