AW: AW: Any interest in participating Canonical EXI interoperability test? canonical time representations

All,

Please find below the changes [1, 2] as discussed during yesterdays telecon.

* Revised addition in 4.5.5. to  "Doing so requires applying the algorithm defined in <adding durations to dateTimes> [XML Schema Datatypes] without modifying the value of seconds"

* Revised para in D.4 to "The subsequently following examples depict possible date-Time values opposed to their canonical form. The canonical EXI form always retains the original seconds part."

* UTC fix in example indicated by Carine

Please control the changes/additions one more time.

If there are no further issues I believe we have a stable document and we are ready to move on.

Moreover, I encourage any implementer of Canonical EXI to test against the testsuite [3]. OpenEXI and EXIficient produce exactly the same results and hence have a clean run.

Thanks a lot to all pushing this forward!

-- Daniel

[1] https://services.w3.org/htmldiff?doc1=http://www.w3.org/TR/exi-c14n/&doc2=https://www.w3.org/XML/EXI/docs/canonical/canonical-exi.html#dt-dateTime
[2] https://services.w3.org/htmldiff?doc1=http://www.w3.org/TR/exi-c14n/&doc2=https://www.w3.org/XML/EXI/docs/canonical/canonical-exi.html#exampleDateTime
[3] https://github.com/w3c/exi-testsuite



________________________________
Von: Carine Bournez [carine@w3.org]
Gesendet: Montag, 22. Januar 2018 13:17
An: Peintner, Daniel (ext) (CT RDA NEC EMB-DE)
Cc: Takuki Kamiya; Rich Rollman; 'Don Brutzman'; public-exi@w3.org
Betreff: Re: AW: Any interest in participating Canonical EXI interoperability test? canonical time representations


Hi,
Thanks Daniel.

On Mon, Jan 22, 2018 at 09:21:33AM +0000, Peintner, Daniel wrote:
>
> I looked into the Cannical EXI spec and added information w.r.t. how to normalize EXI date-times by adding XSD references (see [1]).
>
> Moreover, I added some examples Date-Time examples (see [2]).
>
> Note: The diff shows some more differences that are not any different.
>
> Please let me know whether the new text makes clear what needs to be done.


For the 1st change, in 4.5.5, "the normalization steps" seems a strange
wording, because from the schema spec POV, it's not normalization.
(no strong opinion against this wording, though). "algorithm" seems
sufficient.

In D.4, the first example for utcTime set to true, the result
should be normalized to UTC and it is not.