I have completed the removal of the cellid from the orte_process_name_t
struct on a tmp branch. Tim Prins has successfully tested it on odin, thor,
and bigred at IU - I have checked it on coyote and yellowrail at LANL, as
well as on a standalone Mac. It seems to work just fine in all those
Accordingly, I plan to check it into the trunk late Monday afternoon
(probably around 4pm Eastern). As part of the change, I'll be getting rid of
some obsolete test code, plus the setup_hnp and orte-console code (the
latter depends upon the former, which is so obsolete it won't run anyway -
and our revised launch procedure won't support).
Just wanted to give you a "heads-up" in case you have any tmp branches out
there that might be impacted. I rolled the Voltaire tmp branch into the
trunk in preparation for this change so I could deal with the conflicts
(rather than asking them to do so).