OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

sca-assembly message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: NEW ISSUE: Conformance to OPTIONAL Normative Statements


All,
    prompted by ASSEMBLY-101 I've reviewed the OASIS Guidelines for writing
Conformance Clauses at:

	
http://docs.oasis-open.org/templates/TCHandbook/ConformanceGuidelines.html

I think these say that we need to make a statement in section 13
"Conformance" that states how optional normative statements are to be
handled. The specific wording is:

"Within the Conformance section, a clear statement MUST be made as to how
optional Normative Statements (i.e. those using the MAY keywords) must be
handled."

Obviously all the mandatory normative statements MUST be satisfied, but how
many, and which, optional statements can a vendor implement? If two vendors
implement subsets of the optional statements how can they know if the two
subsets are compatible or interoperable?

The example they give (ebXML Registry Services Specification v2.0) doesn't
map well onto the way the SCA Assembly specification is organized. We may
need to collect the optional statements into groups such that if one
statement in a group is implemented then all the others in that group MUST
also be implemented.

Of course we could always say that vendors are free to implement any
combination of the optional statements they desire?!

The guidelines also seem to imply that there should be no normative
statements in the conformance section, but this isn't so clear. (The stated
purpose of the conformance section is to list "Conformance Clauses that need
to be observed by implementers or users of the specification in order to
claim successful use of a specification. The Conformance Clauses then refer
to the normative statements).

PROPOSAL:
I do not have any specific proposal as to which combination(s) of optional
statements should be implemented. I do however think we should, for the sake
of interoperability, do more than just allow any combination.

Best Regards,
             Eric.

Eric Wells.
Consulting Engineer.
Hitachi Computer Products (America) Inc.
San Francisco, CA. USA.
+1 (415) 656-4346
eric.wells@hitachisoftware.com



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]