-
Hi,
the code at [Cavity 2D](https://steriva.github.io/FEniCSx-tutorials/chapter3/Cavity2D.html) doesn't run anymore as is. Adding the data type to the domain fixes the first error:
`self.domain = cr…
-
### Summarize the issue
Possibly related to #1465.
Creating a mesh of a manifold always assumes that interior (owned) facets will share more than one cell. In the case of manifolds this may not …
-
### Summarize the issue
Using VTXWriter to output a DG function and one of its collapsed sub functions yields incorrect data in parallel.
### How to reproduce the bug
The following MWE and at…
-
> Note: we need to think of a way for users to provide their own fenics mesh with their own meshtags.
>
> Here we have only two options: tags from XDMF, or 1D mesh. What about users providing a cub…
-
### Summarize the issue
Attempting to interpolate the following UFL form yields unexpected data.
### How to reproduce the bug
Consider the following example. I'd expect the output to yield
In the …
-
If PETSc was an optional dependency we should, in principle, be able to build under Windows and in particular via conda.
We have native distributed vectors and a beta-version native parallel CSR ma…
-
Just store the adjacencylist generated by forward scatter of cells.
This can be directly used as a mesh partitioner.
Ref discussions with @IgorBaratta
-
The PlazaRefinementND has considerable dynamic memory allocation inside loops over all cells. This should be eliminated for performance.
A data structure that may help is an `Eigen::Matrix` with an…
-
Hi,
while working on a time-dependent problem with @AnnaSanfilippo we encountered some segmentation faults that might be related to duplication of MPI communicators.
I have slightly changed the po…
-
### Summarize the issue
Adios2 has a difficult relationship with big-endian systems (such as s390x, ppc64). It builds fine, but upstream notes that it is [not really tested](https://adios2.readthedo…