MHonArc v2.5.0b2 -->
emergency message
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
| [List Home]
Subject: Re: [emergency] Draft TC Letter on HazCollect
While I understand the point with regard to the word "compliant" and
the implications of a testing regimen to back it, I think we could
use the words "conformant" or "non-conformant" and qualify that use
by saying that it indicates that all mandatory elements indicated by
"MUST" in the specification are implemented or not when "conformance"
or "non-conformance" is used in regard to a given implementation. We
should probably also acknowledge the current (temporary?) lack of a
certification program backed by a rigorous "compliance testing"
regimen.
Cheers,
Rex
At 6:33 PM -0600 6/13/06, Carl Reed OGC Account wrote:
>My one concern is the use of the phrase, "non-compliant
>implementation". What is a compliant implementation? In the
>standards world, compliance suggests a rigorous testing framework in
>which assertions (rules) as specified in the standard are
>specifically tested against a given implementation. The OGC has
>almost 4 years experience in building and providing compliance
>testing for a number of our core standards. This framework
>determines that a product implementation of a particular
>Specification fulfills all mandatory elements as specified and that
>these elements are operable. There is significant value to an
>organization or vendor in having their implementation of a given
>standard pass a compliance test.
>
>I am not trying to be overly sticky on the semantics, but if I were
>on the NOAA side, I would ask the question "and what exactly do you
>mean by compliance?"
>
>Also, just to stir the pot a bit, even compliance does not insure
>interoperability. Interoperability testing is a whole other topic.
>
>OGC is not the only organization that provide compliance testing. So
>does ISO, IEEE, the ITU, and many other standards organizations.
>
>Regards
>
>Carl
>
>
>
>
>
>