David Bold
David Bold
@bmribler: As I tried to say above: I did not manage to come up with a more minimal producer -just running the segfaulting tests does not seem to reproduce it,...
Or we might use a BOUT++ container - that should also make it easy to reproduce issues locally. And we do not even have to build it the first time...
I am not sure it is the scipy update, or the zoidberg update. This seems to be solveable by specifying `kx` - which is probably not needed for real grids,...
I am working on fixes in the scipy-1.14.1 branch - some of the changes probably should go to zoidberg ...
Unfortunately the whole code would need to be cleaned up, as right now there is no easy fix. See also https://github.com/boutproject/hermes-3/issues/137
Would it not be fine to just remove "EXCLUDE_FROM_ALL"? If you explicitly enable it, you should be fine with building it? If not, disable it?
It does do more then just call sphinx, it also sets up the environment variables, so that python can pick up the python interface for BOUT++. I think that will...
Ah, I remember. For RTD we do in-source-builds. There is even an extra option to allow that, but I do not want to recommend that to anybody that wants to...
I see, so I do not add an extra global option like this: https://github.com/boutproject/BOUT-dev/compare/build-docs-all-option?expand=1 Yes, if you have the time to do that, that would be great. Currently it is...
That sounds interesting, but is it possible to map the different parts of a field3d, such that it appears as a contiguous array in memory, when it is not? Otherwise...