[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [provision] Overlap between spec and profile docs.
At the high level, we should not allow inconsistencies to exist between MUST/SHOULD in core and anything defined in the profiles. If this potential exists, we should consider leaving out of core and making explicit in profile. Right? Darran Gary P Cole wrote: > I'm thinking about XPath at the moment, but I think the issue is > broader. We've agreed that an SPMLv2 provider MUST support XPath 2.0 > abbreviated location paths (and MAY support arbitrary XPath > expressions), but where does this language belong? > > I want to describe XPath in the specification (because XPath and the > requirement to support abbreviated location paths seems general), but > one might argue that the issue of XPath support is specific to the XSD > profile. I'm not trying to argue that specific point right now--I'm > after something else. If we *assume* for the moment that XPath is > irrelevant to the DSML profile, then should the specification (or > should the XSD profile) describe Xpath support? > > Suppose that we decide that a certain aspect of behavior should be > described in the specification--because we believe that the > requirement is general (or is general enough to apply to more than one > profile). Should a specific profile be allowed to override the > specification in any way it sees fit, or should a profile only be > allowed to *add to* the specification? > > Other standards groups have probably already tackled questions like > these. PSTC members (with more experience than I have) may already > know the convention. If so, please share. > > --------------------------------------------------------------------- > To unsubscribe, e-mail: provision-unsubscribe@lists.oasis-open.org > For additional commands, e-mail: provision-help@lists.oasis-open.org >
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]