Hi All, For those who were not able to attend the Forum meeting in Chicago last week, this is an update to inform you about the status of the nonblocking collective I/O ticket. We decided to add nonblocking version of the collective I/O routines to access data with individual file pointers and explicit offsets (and deprecate their split collective counterpart) but not for shared file pointers. The semantics for nonblocking collective shared file pointers were just too weird to have a majority of the forum agreeing upon. Since we don't really have a use case for them and in order stay in the MPI 3.0 bracket we decided to leave them out of this ticket. The current pdf attached to the ticket is the draft that was read at the last meeting and passed its formal reading. We will vote on this ticket at the next meeting in Japan. Thanks, Mohamad