15 Jul
2026
15 Jul
'26
3:40 p.m.
Current agenda: * Form working group consensus on reasons for and against extending the fault model beyond hard process failure * Agree on a name for a simplified `MPI_Comm_agree(comm, &agreed_failed_group)` function and decide on who wants to write the first draft * Begin edits to the `MPI_Comm_shrink` text based on comments Best, Matthew Whitlock