RE: Recommending a Testbed Vendor
Jack,
I was just suggesting that Jim Schoening has already found some alternatives
for which he has identified costs that he can pay within his budget.
Presumably, it would take additional time and incur additional costs to find
options that would support more advanced kinds of hardware.
I'm interested in exploring use of alternative computing paradigms to
support ontology processing, but I'm not convinced that worthwhile work
cannot be done in ontology processing using conventional computers. So, I'd
be interested in reading any papers you would recommend on this topic, and
also would be interested in the thoughts of others in this forum on this
topic.
Cheers,
Phil Jackson
> -----Original Message-----
> From: standard-upper-ontology@LISTSERV.IEEE.ORG [mailto:standard-upper-
> ontology@LISTSERV.IEEE.ORG] On Behalf Of Jack Ring
> Sent: Wednesday, August 20, 2008 11:10 AM
> To: standard-upper-ontology@LISTSERV.IEEE.ORG
> Subject: Re: Recommending a Testbed Vendor
>
> Phil,
> I appreciate the benefit of starting small and inexpensive. However, I am
> wondering how you got the idea that 'initial costs and delays' would be
> encountered and needs to be avoided? Have vendors with mixed hardware
> been
> considered?
> Why not be sensitive to the dual -- that the conventional test bed will
> support only testing of conventional ontologies --- which are already
> known
> to be 'the enemy of knowledge exchange?'
> cheers,
> Jack Ring
>
> ----- Original Message -----
> From: "Philip Jackson" <phil.jackson@computer.org>
> To: "'Schoening, James R Mr CIV USA AMC'" <James.Schoening@us.Army.Mil>;
> <standard-upper-ontology@LISTSERV.IEEE.ORG>
> Cc: "'Jack Ring'" <jring@AMUG.ORG>
> Sent: Wednesday, August 20, 2008 3:03 AM
> Subject: RE: Recommending a Testbed Vendor
>
>
> >I agree with the idea of getting a testbed going with conventional
> > computers, and exploring options for other kinds of hardware as well,
> > perhaps in another testbed and with another vendor. This would minimize
> > initial costs and delays, and avoid a situation where "the perfect is
> the
> > enemy of the good".
> >
> > Phil Jackson
> >
> >> -----Original Message-----
> >> From: standard-upper-ontology@LISTSERV.IEEE.ORG [mailto:standard-upper-
> >> ontology@LISTSERV.IEEE.ORG] On Behalf Of Schoening, James R Mr CIV USA
> >> AMC
> >> Sent: Tuesday, August 19, 2008 6:15 PM
> >> To: standard-upper-ontology@LISTSERV.IEEE.ORG
> >> Cc: Jack Ring
> >> Subject: RE: Recommending a Testbed Vendor
> >>
> >> Jack,
> >>
> >> Interesting. One approach might be to set up another testbed for
> >> your ideas and link it to this testbed.
> >>
> >> If this testbed succeeds and grows, I envision other testbeds
> >> being set up for related purposes by other groups. If we break up the
> >> different functions into components, a demonstration could access
> >> different components in different testbeds.
> >>
> >> Jim Schoening
> >>
> >> -----Original Message-----
> >> From: Jack Ring [mailto:jring@amug.org]
> >> Sent: Tuesday, August 19, 2008 12:34 PM
> >> To: Schoening, James R Mr CIV USA AMC;
> >> standard-upper-ontology@LISTSERV.IEEE.ORG
> >> Subject: Re: Recommending a Testbed Vendor
> >>
> >> Why limit our thinking to solutions on stored program computers? Isn't
> >> it clear by now that computers as we know them are the primary
> >> obstruction to making sense out of data within the context of
> >> multifaceted ontologies? How shall we experiment with FPGA and
> >> transputer kinds of hardware? My lag in prompting topic P4 may be a
> >> large part of this issue but I must speak up now even though my message
> >> may not be meaningful to all.
> >> We need a vendor who will be willing to install some non-cpu hardware
> in
> >> the test bed.
> >> Jack Ring
> >