Open MPI logo

Open MPI Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Development mailing list

From: Jeff Squyres (jsquyres_at_[hidden])
Date: 2007-01-05 20:25:01


Whoops. that was *supposed* to be off list. :-)

I meant to take this off list to discuss with Ralph further and come
up with the final answer because I didn't want to clutter up everyone
INBOXes any further...

On Jan 5, 2007, at 8:22 PM, Jeff Squyres wrote:

> On Jan 5, 2007, at 8:16 PM, Ralph Castain wrote:
>
>>> I actually don't quite understand -- in this case, we *did* just
>>> take
>>> a big drop from ORTE. If the problem truly is in the MPI side of
>>> the
>>> house, then the "big drop" philosophy doesn't help here.
>>
>> No, we took a patch - one among many. That means I can't be certain
>> that the
>> RTE is fully consistent. While we did our best, that doesn't mean
>> it's
>> consistent.
>
> Off list.
>
> Now wait a minute -- I'm now very confused. You told me on the phone
> the other day that for #700, you made a patch that made the RTE be
> the same between the trunk and v1.2. Granted, a few things happened
> on the trunk *after* you made that patch, and a few non-ORTE patches
> were applied to 1.2 before yours was applied. But I'm going on what
> I understood you said in that we effectively *did* take a drop
> because your patch make ORTE in 1.2 become exactly like the ORTE in
> the trunk...
>
> --
> Jeff Squyres
> Server Virtualization Business Unit
> Cisco Systems
>
> _______________________________________________
> devel mailing list
> devel_at_[hidden]
> http://www.open-mpi.org/mailman/listinfo.cgi/devel

-- 
Jeff Squyres
Server Virtualization Business Unit
Cisco Systems