Guillaume Giudicelli
Guillaume Giudicelli
> Getting rid of the Order argument I don't understand. it s not gone gone it would just become a template argument instead of a parameter But you're right that...
> How does this get rid of any switch statements, though? With the current code in fe_lagrange_shape_2D.C it would just take away one switch statement in each of those functions...
> ElemType same problem. Maybe we should have the "1 elem type per subdomain" limitation to help us loop over elements with a fixed element type
Where is the element numbering code in libMesh? I was thinking already sorting by subdomain would achieve something signficant in MOOSE, skipping the "reinitSubdomain" calls. I did not even know...
Seems like if we renumber it during the simulation we need the numbering schemes to be implemented in libmesh (not in a moose mesh generator). > Sort by subdomain first,...
@vincentlaboure do you have any nodal variables? For nodal variables it does look a little tricky to add this. But for element variable there is already all the code in...
> Does that simplify the request? Unfortunately not, the code for nodal variables is as @jwpeterson making large use of the match between solution vector and mesh ordering. We could...
Code changes are fine here but the case might fit better in OpenPronghorn? or under VTB
Linked to #21288
> which the "coarse quad" was turned into a degenerate pentagon? yup