[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [office] One strictly conforming document?
On 02/03/09 16:03, Patrick Durusau wrote: > Rob, > > firstname.lastname@example.org wrote: >> Patrick Durusau <email@example.com> wrote on 02/03/2009 08:42:50 AM: >> >> > <snip> >> I'd agree that if we required that an implementation report pricing >> models, sales targets, market share analysis, etc., then that would be >> disallowed. But it is quite the norm these days for an application of >> the type we are defining conformance for to give the user a choice of >> format targets and to report warnings or errors when a document is >> loaded that does not conform to that format. So it is entirely >> reasonable and proper for a standard that defines some behaviors of >> such an application with respect to file formats to also specify how >> and when it informs the user with regards to extensions. >> >> > Sure. Perhaps I was mis-reading Michael's proposal. > > Let me say what I was reading and you or others can say if that was what > you saw as well: > > I was reading the conformance clause to exclude from ODF 1.2 conformance > an application that produced only a non-strictly conforming ODF document. > > In other words, an application that produced only ODF plus some unknown > extension could not claim to be ODF 1.2 conforming. This is what I intended to say here. > > Was that what you read as well? > > The reason I find that problematic is that I may want to have an > application that never produces ODF 1.2 strictly conforming documents. This may be okay. What is important to me however is that one can differ an application of this kind from one that only produces strictly conforming documents. Assume we have only one conformance level for producers, and this allows to produce conforming and strictly conforming documents. In this case, there is no way to figure out whether an applications that claims to be conforming is actually able to produce strictly conforming documents. We can only solve this issue by either requiring that an application must be able to produce only strictly conforming documents on demand. Or we must have two conformance levels for producers. In this case, an application that claims to be strictly conforming must be able to produce only strictly conforming documents on demand. An application that does not have this capability may call itself conforming, but it must not call itself strictly conforming. > > For example, what if I write an XSLT stylesheet that adds custom work > flow management information to an ODF file and then stores that in a > package. When I sent that file to an ODF application, it simply ignores > the work flow management information that is solely managed by another > application. Since my application is only meaningful in terms of the ODF > 1.2 format, I would like to say that it is an ODF 1.2 conforming > application but obviously I am going to tell users that it uses > extensions for the CMS capabilities. As a matter of fact, it could be > the case that the consuming ODF application doesn't even preserve the > work flow information but that is added on every commit to the CMS. I'm not sure if the the work flow management information would be stored as RDF/XML, or as extension to the schema. I therefore would like to clarify that the extensions we are currently talking about are extensions to the schema we use in content.xml, styles.xml, meta.xml or settings.xml. Whether an application stores information in a RDF/XML stream has no influence on whether it may call itself conforming. > > What value do you get by excluding my XSLT stylesheet from saying that > it is an ODF 1.2 conforming application, assuming that it accurately > reports that it uses extensions for the CMS capabilities? > > I mean for it to consume and produce ODF 1.2 documents for consumption > by other ODF 1.2 applications. Why deny me the label of ODF 1.2 > conformance? > > Note that I freely grant that the conformance requirements can say: > Conforming applications must/should say whether they do a, b, c, or a + > b, etc. Just so I can compare applications in a meaningful way. I think this is essentially what I mean by having two conformance levels. An application must provide the information whether it stores extensions, or does not do so. But we must provide the appropriate conformance levels, so that implementers can do so, and actually have to do so. Best regards Michael > > I would assume that creators of non-strictly conforming producers aren't > going to conceal those capabilities. They are no doubt the rationale for > some feature, etc. > > Hope you are having a great day! > > Patrick > > > -- Michael Brauer, Technical Architect Software Engineering StarOffice/OpenOffice.org Sun Microsystems GmbH Nagelsweg 55 D-20097 Hamburg, Germany firstname.lastname@example.org http://sun.com/staroffice +49 40 23646 500 http://blogs.sun.com/GullFOSS Sitz der Gesellschaft: Sun Microsystems GmbH, Sonnenallee 1, D-85551 Kirchheim-Heimstetten Amtsgericht Muenchen: HRB 161028 Geschaeftsfuehrer: Thomas Schroeder, Wolfgang Engels, Dr. Roland Boemer Vorsitzender des Aufsichtsrates: Martin Haering