Phil Miller - NOAA
Phil Miller - NOAA
Discussed this with Nels, and made some small revisions accordingly.
@RolfHut Eric mentioned you may not have seen this suggestion at a previous meeting. It had pretty general support. It replaces `input_vars = get_input_var_names` with `input_vars = get_varset_members("input")`, so minimal...
Austin suggested an object-free extension, tentatively named `bmi:introspection`, that indicates availability of variable sets `bmi:introspection:all_sets` and/or `bmi:introspection:all_variables`. Those sets would be usable for a model-description utility to just enumerate everything,...
Possible variation on the introspection idea: an extension that asks for a static version of everything that might be available, independent of any specific config file. In that case, the...
Rolf suggests this hurts usability and code readability for simpler use cases Bert noted concerns around dynamic or parseable variables names. It could be a time savings, but shifts burdens...
Fred: suggests an alternative approach of adding a `foo_by_index` set of advanced/fast versions, to keep the simple readable/teachable versions in place, while allowing optimization.
Agreement on "can implement a friendly interface on an efficient interface, but can't implement an efficient interface on a cumbersome interface"
Potential to implement desired functions in a "high performance methods" extension
https://github.com/trilinos/Trilinos/pull/10984 @bartlettroscoe There's the PR
Not resolved. Don't autoclose