[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Profiling for Interoperability
NOTE I notice that it is possible to consider profiles as a mechanized feature for processing of documents in a given interoperability setting. That is, it involves agreement within a community of practice around the behavioral interoperability they will count on in the products they adopt and configure for their situation. Although this may be far beyond what the OIC TC could take on, we might want to keep in mind that this prospect could matter in the future. I don't think there is anything needed in the charter beyond what is already proposed, however. - Dennis Dennis E. Hamilton ------------------ NuovoDoc: Design for Document System Interoperability mailto:Dennis.Hamilton@acm.org | gsm:+1-206.779.9430 http://NuovoDoc.com http://ODMA.info/dev/ http://nfoWorks.org EXTENDED DISCUSSION I note that there are elements of this TC-formation discussion that are relevant to profiling for particular kinds of software interoperability and interchange scenarios concerning ODF documents. In general, there's a very big difference between determining that expression of a document using ODF is verifiable and valid against the specification and what software and its users rely upon and do in various circumstances. The determination that a processor acts in a conformant way cannot be done without knowing what happens when given a conformant document, not just whether or not the ODF document is "accepted" by the processor and a conformant ODF document is produced (directly or via transformation of a conformant input document). This second kind of behavioral conformance assessment may depend on a declaration by a software producer (on being given a checklist or guide on how to prepare the declaration) so that adopters of the product can assess its suitability in a particular context of use. Such a declaration would include deviations and extensions used/accepted as well as limitations on the degree of support for the ODF format (unimplemented features, how occurrence of such features are handled, etc.) Another kind of assessment has to do with interoperability in a particular context of use or interoperability scenario. This leads, in my understanding, to the need for some sort of profiling, where a community of practice could agree on the details of behavioral conformance required for successful interchange in an interoperability situation. It then becomes an interesting matter to consider whether such a community has technical means to ensure that the boundaries of the profile are not exceeded when using any qualified products and when creating documents or receiving them. One could view schema/feature restriction as a simple case of agreement enforcement, but that's certainly not the only case. I suspect that this kind of interoperability profile is far beyond conformance assessment and declaration, yet it is relevant to providing assurance among product adopters that the documents they produce are interchangeable in the expected context and the requisite interoperability scenario is satisfied. Since this is generally unexplored territory, it is interesting to know how much its consideration would be in scope for OIIC at some level. I looked through the discussion archives and I notice that this sense of profile is somewhat different than the use that has been given with regard to leveling of conformance, even for vertical situations. I am thinking of provisions that would have adopters of the format be in control of their profile agreement and have technical means to make it real. I am thinking that it takes advances in the ODF specification to provide for this. A format for specifying a profile, for interchange with or as part of an ODF document, would more likely be the responsibility of the ODF TC. In contrast, one could have guidelines on how a product could facilitate its mechanized conformance testing, including how deviations, extensions, and limitations are handled and accounted for. This could remain separate from the definition of the OpenDcument Format. Were there such arrangements, they might provide a different avenue for profile-honoring as well. BACKGROUND FROM THE LIST ARCHIVES 2008-06-08 David R.R. Webber, What is interoperability anyway / really? http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00030.html 2008-06-09 Dave Pawson, Re: Interoperability versus Conformity http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00044.html 2008-06-11 Rob Weir, Re: Interoperability versus Conformity http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00092.html 2008-06-11 Matthew Reingold, Re: Interoperability versus Conformity [not linking, has an idiotic confidentiality-warning boilerplate] 2008-06-11 Rob Weir, ODF Interoperability Presentation http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00100.html 2008-06-11 Rob Weir, Acid Tests http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00123.html 2008-06-12 Dave Pawson, Re: Acid Tests http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00156.html 2008-06-12 Luc Bollen, Re: Acid Tests http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00168.html and interesting issue concerning what can be automated (an interesting -ility with regard to processors) 2008-06-16 Rob Weir, Interop with other standards http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00357.html hmm, as a profile case - seems way far out there for OIIC 2008-06-17 Shawn Grover, Profiles http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00389.html thread for teasing out what "profile" is to mean for OIIC 2008-06-17 Dave Pawson, Profiles. Attempting a definition http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00414.html looks for an use of the term that is applicable here. Needs to go back to ODA "Document Architecture" and "Document Profile" 2008-06-17 Garry Hurley, Re: Profiles ... http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00416.html Gets specific about profiles as levels of conformance and as descriptive 2008-06-18 Dave Pawson, Re: Profiles ... http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00420.html gets into problem about profiles for what; collides with what different ideas each of us might have over what an "application" is 2008-06-18 Rob Weir, Re: Profiles ... http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00426.html adds an ISO definition that is also applicable (though Dawson misses how it aligns with his use) 2008-06-19 Robert Weir, Re: The importance to users of documents looking the same http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00489.html Suggests possible "ODF Profile Requirements" as a TC output following survey of what is already done in this area, etc. 2008-06-20 Robert Weir, Profiles -- How much detail in the charter? http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00574.html 2008-06-21 Sander Marechal, Profiles: suggested use-case http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00609.html Interesting ... thought not my kind of profile, but now not so sure ... 2008-06-23 Jose Lorenzo, Summary of some "profiles" usage and ideas http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00705.html Identifying how broadly this is being used 2008-06-23 Robert Weir, Use Cases http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00653.html 2008-06-30 Robert Weir, Re: My perspective http://lists.oasis-open.org/archives/oiic-formation-discuss/200806/msg00824.html nice discussion between Weir and Thomas Zander that also touches on profiling along with conformance and interoperability 2008-07-29 Dave Pawson, Profiles http://lists.oasis-open.org/archives/oiic-formation-discuss/200807/msg00225.html referencing http://www.w3.org/TR/2005/NOTE-spec-variability-20050831/ [end]
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]