Peter Hill
Peter Hill
As long as readthedocs still builds it fine, and it's documented how to build the docs locally, I think it's fine
Some very brief thoughts that may or may not be relevant: - it's possible to create custom datatypes in MPI that can deal with the non-contiguous guard cells - MPI...
Maybe `is_y_periodic`?
I'm not sure we really want to be uploading a new version on every pull into `next` either!
Please can you split this into one PR per change? It's 100 files and 5k lines of changes, it's going to be very difficult to review. GitHub will only show...
Ah, clang-tidy has spotted a bunch of virtual functions that will need qualifying with `BoutMesh::` Yeah, and some tests fail, so this trivial change is definitely not enough. Shame
Ok the more fundamental problem is that we may need to read some quite important variables (`nx`, for instance) from `Options`, which involves parsing a string, which involves creating the...
> Perhaps we could remove the `Mesh*` argument to the `FieldFactory` constructor. It sets a default `Mesh*`, overridden by optional `Mesh*` arguments to `create2D` and `create3D` methods. Those arguments would...
Looks like you just need to merge recent `next` in to fix the test failures
Also something like `.update(const Options&)`