amadeus84

If you're compiling with g++, compile with address sanitizer ( -fsanitize=address and link against libasan), e.g. for examples/surface.cpp g++ -g -Wall -fsanitize=address -o surface $(python-config --includes) surface.cpp $(python-config --ldflags --embed)...

Currently there's 791 forks, I guess everyone interested is already maintaining their own repository. The original repository doesn't seem to be maintained anymore. It's a pity.

I'd happily contribute, time permitting, to an actively maintained plotting library. The only knowledge I have about the Python C-API, though, is very limited, just what I could gather this...

So if you guys are interested, check out my repository https://github.com/amadeus84/matplotlib-cpp I fixed the segfault on exit bug and added support for contiguous ranges, without internally copying the data. Also,...

@pascal-mueller It is related, it's explained in the stackoverflow link. Py_Finalize() is called from the _interpreter destructor, which is being called too late, because the interpreter is static. calling kill()...

Following up on unddoch's analysis on stackoverflow https://stackoverflow.com/questions/67533541/py-finalize-resulting-in-segmentation-fault-for-python-3-9-but-not-for-python/67577360#67577360 . In basic.cpp, adding plt::detail::_interpreter::kill(); just before return from main, fixes the segfault. Compiled the code like this: g++ -g -Wall -o...

See issue #268 and the proposed fix. Add plt::detail::_interpreter::kill(); in main, before return. Alternatively, you may be able to fix it by changing the backend to TkAgg, although I haven't...

I ran into the same problem some months ago, unfortunately looks like nobody maintains this anymore. I have searched into the equivalent code to initialize the interpreter for the new...