Sounds like another case of "No Bucks - No Buck Rogers" story. As much as
we want Open Source and Open Standards to solve problems, it still comes
down to a matter of funding. At least, with Open Standards and Open Source
combined, if one company goes belly up the product can be continued. The
challenge with Open Standards is to make sure that standards board can be
Daniel Beatty, Ph.D.
Computer Scientist, Detonation Sciences Branch
1 Administration Circle M/S 1109
China Lake, CA 93555
On 2/22/13 3:02 PM, "Jeff Squyres (jsquyres)" <jsquyres_at_[hidden]> wrote:
> Meh. UDAPL never really caught on.
> It was the low latency stack in Solaris for a while. Then they shifted to
> verbs. Then Sun got bought by Oracle, and they didn't really care about UDAPL
> anymore (at least from an HPC perspective).
> Intel used to use UDAPL as their main communication layer for Intel MPI (I
> don't know if they still do or not).
> But outside of those use cases, I'm not aware of much UDAPL usage. UDAPL Was
> only in OMPI for Solaris. And now that they're backing away from HPC, I can
> foresee us removing the UDAPL BTL someday.
> On Feb 22, 2013, at 5:49 PM, "Shamis, Pavel" <shamisp_at_[hidden]>
>> You may find more updated version on ofed website:
>> Pavel (Pasha) Shamis
>> Computer Science Research Group
>> Computer Science and Math Division
>> Oak Ridge National Laboratory
>> On Feb 22, 2013, at 2:18 PM, Orion Poplawski <orion_at_[hidden]> wrote:
>>> Is the uDAPL project still active? The SourceForge page says it isn't:
>>> As of 2009-03-13, this project is no longer under active development.
>>> openmpi's configure emits;
>>> checking --with-udapl value... simple ok (unspecified)
>>> checking --with-udapl-libdir value... simple ok (unspecified)
>>> configure: WARNING: On Linux and --with-udapl was not specified
>>> configure: WARNING: Not building the udapl BTL
>>> Should it still emit this warning?
>>> Orion Poplawski
>>> Technical Manager 303-415-9701 x222
>>> NWRA, Boulder Office FAX: 303-415-9702
>>> 3380 Mitchell Lane orion_at_[hidden]
>>> Boulder, CO 80301 http://www.nwra.com
>>> users mailing list
>> users mailing list