I recommend against using the version-independent variant, because,
while unlikely, a version might be updated in a way that makes
implementations using an older version incorrect. ITU-T may also
require that references be to a fixed, standard version.
I would vote for either retaining the new URNs used in the errata and
the ITU-T version, with a reference in the specification itself to the
corresponding W3C Recommendation, or else just use the URLs for the
current W3C Recommendation.
Regards,
Anne
Erik Rissanen wrote:
> What should the identiers for the two relevant data types be in XACML? I
> have so many choices here. :-)
>
> In XACML 2.0 OS they are:
>
> http://www.w3.org/TR/2002/WD-xquery-operators-20020816#dayTimeDuration
> http://www.w3.org/TR/2002/WD-xquery-operators-20020816#yearMonthDuration
>
> They were changed in the errata (and the ITU-T version):
>
> urn:oasis:names:tc:xacml:2.0:data-type:dayTimeDuration
> urn:oasis:names:tc:xacml:2.0:data-type:yearMonthDuration
>
> Based on the current W3C recommendation we could have:
>
> http://www.w3.org/TR/2007/REC-xpath-functions-20070123/#dt-yearMonthDuration
> http://www.w3.org/TR/2007/REC-xpath-functions-20070123/#dt-dayTimeDuration
>
> These are resolvable URLs pointing to type definitions.
>
> We could also use a version independent variant:
>
> http://www.w3.org/TR/xquery-operators/#dt-yearMonthDuration
> http://www.w3.org/TR/xquery-operators/#dt-dayTimeDuration
>
> These are also resolvable and will point to the latest version of the
> document.
>
> (Note the "#dt-" and how that wasn't there in the XACML 2.0 version. We
> could create yet another variant by removing it.)
>
> Another interesting twist to it is that the xquery document calls them
> xs:yearMonthDuration, so they treat them like they are in the XML Schema
> namespace. So, we could perhaps do this as well:
>
> |http://www.w3.org/2001/XMLSchema|#yearMonthDuration
> |http://www.w3.org/2001/XMLSchema|#dayTimeDuration
>
> What do people think?
>
> My preference are these:
>
> http://www.w3.org/TR/xquery-operators/#dt-yearMonthDuration
> http://www.w3.org/TR/xquery-operators/#dt-dayTimeDuration
>
> Note that this will break the names compared to the 2.0 OS, but I
> presume it is ok since the errata was also breaking them.
>
> Regards,
> Erik
>
>
> Anne Anderson - Sun Microsystems wrote:
>
>>I recommend referring to the now-approved XPath and XQuery standards.
>>Referencing the standard means people can re-use standard
>>implementations of the XPath and XQuery functions with confidence. It
>>also makes our own specification simpler and less prone to introduced
>>errors.
>>
>>Regards,
>>Anne
>>
>>Erik Rissanen wrote:
>>
>>>All,
>>>
>>>For the next 3.0 core draft I was planning to incorporate all the 2.0
>>>errata. Most of it concerns the definition of some data types and
>>>functions copied from xpath/xquery.
>>>
>>>The copying was made to accommodate ITU since xpath/xquery was not
>>>approved at the time. Anne raised the issue earlier about going back to
>>>referring to the xpath/xquery, rather than copying text, now that they
>>>have been approved.
>>>
>>>What should I do it for the next 3.0 draft? Should I copy text like the
>>>current 2.0 errata, or should I refer to the approved version of
>>>xpath/xquery?
>>>
>>>What should I do for future 2.0 errata? Should I keep the copied text,
>>>or should we revert back to referring to xpath/xquery?
>>>
>>>Regards,
>>>Erik
>>>
>>
>
--
Anne H. Anderson Email: Anne.Anderson@Sun.COM
Sun Microsystems Laboratories
1 Network Drive,UBUR02-311 Tel: 781/442-0928
Burlington, MA 01803-0902 USA Fax: 781/442-1692