Joseph Ciurej
Joseph Ciurej
It should be possible to transform between Silo meshes on disk and Blueprint meshes in memory. For full support, this will require all of the following features: - [ ]...
In order to support AMR and other problems with nontrivial rank-domain distributions, the Blueprint index needs to be extended to include sparser and more descriptive domain metadata. More specifically, the...
There are a number of places where the `conduit::blueprint::mesh::topology::unstructured::generate_offsets` function is used to generate offsets in their default path (i.e. for input topology `n`, at path `n["elements/offsets"]`). In order to...
For the sake of convenience, many codes choose to represent their 1D meshes as "fake 1D meshes," which are meshes with multiple zones in the *real* direction (e.g. `r`) and...
In order to support backwards compatibility as Blueprint evolves, the schema should be augmented to include version information.
In order to enable more coherent communication of relationship data, this data should be codified as a top-level entity type with a dedicated Blueprint section. Initially, this top-level entity type...
The design philosophy behind all of the curent `conduit::blueprint::verify` functions has been to approach the problem of schema validation from a data-agnostic perspective. While this makes validation straightforward and efficient,...
The recently merged pull request #285 includes validation support and test cases for AMR meshes, but these updates aren't reflected in the official documentation anywhere. In order to rectify this,...