Marc Glisse

I'd say that the base C++ code is more important than the wrapper, and for someone using anything based on the cubical complex, no matter if it is through a...

Before I remove it from my clone, here is the patch I used to test that this was indeed the problem. ```diff --- a/src/Alpha_complex/include/gudhi/Alpha_complex.h +++ b/src/Alpha_complex/include/gudhi/Alpha_complex.h @@ -384,6 +384,22 @@...

> With MSVC, a warning message to explain the feature is not available is displayed. https://docs.microsoft.com/en-us/cpp/sanitizers/asan ?

> I see that the C++ library supports 3D periodic alpha complexes (I think) Yes. > but the Python interface doesn't have this option. See #378, CGAL is in the...

We don't know how big the changes in CGAL are going to be in terms of API, maybe small, maybe large. We should have a look at that PR to...

> As @hschreiber needs this feature for the persistence matrix PR #669 , I would like to merge it now. Is it ok for you ? The cleaner way would...

I assume it is the same issue, you don't need any code to reproduce, just starting a new notebook. The browser console shows: ``` Failed to load extension: Array [...

I don't remember, sorry. This platform also seems to fail somewhere in Alpha_shape_3, maybe debugging that one first would be easier, in case it is related?

Your explanation is a bit confusing. You seem to rely on the fact that the handles all point to objects of size at least `sizeof(void*)`. I think this deserves a...

This still looks confusing. A `void**` (after casting to an integer type) will be a multiple of `sizeof(void*)` on most platforms, but there is no reason for this to apply...