Re: XSLT WG Telcon Agenda 10 September 2015

[Notes from the scribe:

1 as you know, I lost network contact with the call just before the end.
Please check the record and correct the account of our action on bug
29081 and anything discussed after that.

2 following the example of the minutes from this week's joint QT call, 
I am sending these minutes to the public list, not the member-only list.
The telcon connection details have accordingly been excised.

-CMSMcQ]



The XSL WG met on Thursday, 10 September 2015 at the usual 90-minute
start time.

>   09:00 a.m. Pacific Standard US/Canada,
>   10:00 a.m. Mountain Standard US/Canada,
>   11:00 a.m. Central Standard US/Canada,
>   12:00 Eastern Standard US/Canada,
>   17:00 UK, and
>   18:00 for France, Germany, Sweden and Belgium


> AGENDA:

Sharon chaired this week.

> 1 Roll call 

Present:

  Sharon Adler
  Anders Berglund
  Carine Bournez
  Abel Braaksma
  Florent Georges
  Michael Kay
  Michael Sperberg-McQueen


> 2 Assign minute-taker - SCA requested MSMQ act as scribe again.

MSM took these minutes.


> 3 Approval of minutes:from XSLT Telcon 27 August 2015
>  https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0006.html

The minutes were approved as submitted.


> 4 Action Items

> ACTION-2015-04-23-001: MK to add a section to the back matter suitable
> for pointing to the RelaxNG schema.
>    status: still pending

> ACTION 2015-08-27-004: ABr to integrate the revised override example
> (to be produced in response to action 2015-08-27-001 and -002) into
> the test suite.
>    status: still pending

> ACTION 2015-08-27-006: ABr to raise an issue against the XPath 3.1
> sentence "A variable binding with a type declaration xs:error always
> raises a type error" with explanation / justification as in comment 5
> of 28955.
>    status: still pending

> ACTION 2015-09-03-001: MKay to add fn:available-system-properties() to the spec
>   Completed, see 20.3.5


> ACTION 2015-09-03-002: Abel to add tests for fn:available-system-properties() to the
> test suite
>   status: pending

> ACTION 2015-09-03-003: MKay to review the impact of supporting XP3.1 with arrays in
>  XSLT3.0 (bug 29079)
>    completed - see email

https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0007.html


> ACTION 2015-09-03-004: MK to propose wording for bug 29081 prohibiting use
>  reserved namespaces in extensions functions/attributes

>  completed - see Bugzilla

https://www.w3.org/Bugs/Public/show_bug.cgi?id=29081#c1


> ACTION 2015-09-03-005: MK to fix 29084 according to the proposed resolution
>  completed

> ACTION 2015-09-03-006: MK to fix 29086 by removing the sentence as proposed.
>  completed

> ACTION 2015-09-03-007: MK to send email summarizing the proposed changes for bug 29088.
>    completed - see Bugzilla

https://www.w3.org/Bugs/Public/show_bug.cgi?id=29088#c4
https://www.w3.org/Bugs/Public/show_bug.cgi?id=29088#c5


> 5 Other administrative business

> 5.1 Review of agenda - items to be added to Agenda?

From here we jumped briefly to item 7.1 on the agenda (recently added)
to discuss the status of tests.

Some topics to be added to the agenda were inserted here:

5.1.1 Collations (and merge):
https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0033.html

We discussed but no action seems necessary.  


5.1.2 Public work in the WG

See https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0031.html
and following.

SCA noted that all purely technical work should take place on the
public mailing list. Administrative work can and should be done on the
member-only list. Topics that for some reason involve
member-confidential information will also be discussed on the
member-only list.

SCA reminded the WG that members who speak on the work on the group in
public should let her know.

In response to a query from ABr, CB noted that while email addresses
are humanly visible, they are obfuscated by default in a way that
makes the addresses harder for robots to harvest. It does not
completely guarantee unharvestability, she said, but it is
surprisingly effective.


5.1.3 JSON data in the wild

We discussed the thread beginning at
https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0028.html

SCA pointed out that normally, what happens at this stage in the
development of a spec is that the WG starts to gather potential
requirements for the next version. But we are not committed to a new
version, and starting a requirements list might convey the idea that
we were committed to one.

If we do start a collection of potential requirements for a future
version, we should NOT regard that as a sign that we will be doing an
XSLT 3.1 (and should take care that it does not convey that impression
to others).

From here we moved to section 6.1 of the agenda.


> 5.2 Telcon schedule - September/October

>     September - MK will be at XML Summer School on Sept 17th and on vacation Sept 24th.
> No telcons on those dates.

>     Current Planned CR date - 13 October 2015.

> 6 XSLT Draft

> W3C Internal Working Draft 5 September 2015

> https://www.w3.org/XML/Group/qtspecs/specifications/xslt-30/html/Overview-diff.html

> Last public draft:
> W3C Last Call Working Draft 2 October 2014

> http://www.w3.org/TR/2014/WD-xslt-30-20141002/

> 6.1 Example illustrating overriding and interaction with a library
> package - MKay changes
>    https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0008.html
>    https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0011.html - Abel response
>    https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0022.html - Abel points out a new error
>    https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0016.html - SperberMcQueen

RESOLVED: No further changes are needed to the spec in light of the
email thread at
https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0008.html


> 6.2 "Available" variables
>    https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0010.html
>    https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0012.html

We reviewed this topic; further discussion will be tied to the bug
https://www.w3.org/Bugs/Public/show_bug.cgi?id=29113


> 6.3 xsl:attribute-set sequence constructor
>    https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0013.html
>    No more "nice to haves"

No action needed.


> 6.4 XSLT's Relationship to XPath 3.1
>    Continue discussion of XPath 3.0 and 3.1 in relation to XSLT 3.0

>   6.4.1 XSLT 3.0, XPath 3.1, including arrays
>     https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0007.html - MK
>     https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0014.html - Abel

>     Abel wants to consider xsl:array / xsl:array-item, similar to
> xsl:map/xsl:map-entry.

We reviewed the background of this topic.

Part (A) on streamability:  we just need to draft the rule changes.

ACTION 2015-09-10-001: MK to draft the streamability rule changes
needed to address the use of XSLT 3.0 with XPath 3.1 (as outlined in
https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0007.html part
(A))

Part (B) on arrays.

MK summarized the relevant part of the email:

  - B.1 templates handle maps the way they handle any function; we can
    say the same w.r.t. arrays.

  - B.2 we can follow XQuery in prescribing atomization, which is well
    defined.

ABr suggested that since arrays can be atomized, we could make default
template rules atomize them, rather than treating them like other
functions. MSM argued against this on the grounds that it's easier to
deal with if arrays are treated the same as other functions. ABr
argued that it's easier to deal with if all atomizable items are
treated the same way. SCA concluded that we need a proposal on the
basis of which to make a decision.

ACTION 2015-09-10-002: MK to make a proposal on how to handle arrays
in the template-match rules (as outlined in
https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0007.html part
(B) and discussed on 10 September [arrays as functions, or atomized?])

Part (C) on serialization.

MK noted on reflection that he may already have added all the relevant
properties to serialization, because of the way the schema has
developed.



>   6.4.2 ACTION 2015-06-18-002: Abel will look at the XPath 3.1 and 3.0 bug
>    lists and report back the bugs applicable to XPath 3.0..
>    https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0005.html

This agenda item is a placeholder.


> 6.5 Validating the examples
>     https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0025.html

On (A) (children of override), MK has made a fix.

On (B), we discussed whether we *should* allow xsl:use-package as a
child of xsl:stylesheet.

RESOLVED: to allow xsl:use-package as a child of xsl:stylesheet.

ACTION 2015-09-10-003: MK to revised the spec text to allow
xsl:use-package as a child of xsl:stylesheet (and clearly so).

On (C) (validation), we wish we could improve this situation, but we
cannot at this stage. This is a possible area of activity for EXPath.


> 6.6 Default mode through xsl:default-mode, if it is not declared - Abel
>     https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0015.html

We discussed this, without all WG members achieving a full
understanding of the issues.

ACTION 2015-09-10-004: MK to open a Bugzilla entry for the issue
identified in
https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0015.html and
explain the issues.


> 6.7 Review Updated Appendix F

>     6.7.1 "compare" MKay and ABraaksma's implementations

> 7.0 Spec bugs

> 28937  [xslt 3.0] Keeping XSLT 3.0 maps in sync with XPath 3.1 maps
>     No issue to discuss.  Will be kept open to use as placeholder for
>     changes to be added in future.

>     See MKs note in bug report on To Do work in spec.
>      "Note that in the Change Log, against bug 24266, there is a TODO 
>      reminding us that we have not defined streamability rules for new 
>      expressions and functions in XPath 3.1."


> 28955  [XSLT30] xsl:try should probably include xs:error in its enumeration
>        https://www.w3.org/Bugs/Public/show_bug.cgi?id=28955
>         Resolved at last meeting. To be closed when MK has implemented
> in spec. and ABr has completed action item to submit bug for XP 3.1

The resolution is now in the spec; the bug can be closed.

ACTION 2015-09-10-005: SCA to close bug 18955.


> 29060  [xslt3.0] Functions: determinism and memoization
>        Discussed 08-27-2015: Left open for primarily editorial work.
>        https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Aug/0047.html et seq.

> 29079  [xslt 3.0] xsl:where-populated and empty arrays - MK submitter
>   https://www.w3.org/Bugs/Public/show_bug.cgi?id=29079

Discussed earlier in this meeting.


> 29081  [xslt 3.0] Use of reserved namespaces in extensions - MK submitter
>   https://www.w3.org/Bugs/Public/show_bug.cgi?id=29081
>   see proposal in bugzilla  

We reviewed the proposal and approved it.

[Scribe's note: at least, I believe we approved it. I lost voice
contact with the call just as the discussion here was concluding.]


> 29084  [xslt 3.0]default-mode - ancestor, or ancestor-or-self? - MK submitter
>   https://www.w3.org/Bugs/Public/show_bug.cgi?id=29084

> 29086  [xslt 3.0] Conformance section: recoverable errors - MK submitter
>   https://www.w3.org/Bugs/Public/show_bug.cgi?id=29086

> *29088  [xslt 3.0] visibility="abstract" on xsl:expose - MK submitter
>   https://www.w3.org/Bugs/Public/show_bug.cgi?id=29088

>   Review summarization of changes as result of MK action item 2015-09-03-007.



> 29092  [XT30] (editorial) Reference error in the spec - ABr submitter
>    (no discussion necessary)
>   https://www.w3.org/Bugs/Public/show_bug.cgi?id=29092

> New Bugs to be discussed  

> 29094  [xslt 3.0] conflict between xsl:accept and xsl:override - MK submitter
>   https://www.w3.org/Bugs/Public/show_bug.cgi?id=29094

> 29095  [xslt 3.0] Edge case inconsistency between xsl:expose and xsl:accept - MK submitter
>   https://www.w3.org/Bugs/Public/show_bug.cgi?id=29095

> 29109  [xslt 3.0] New function available-system-properties() - MK
>   https://www.w3.org/Bugs/Public/show_bug.cgi?id=29109

> 29113  [XSLT30] fn:system-property does not a allow an EQName - Abel
>   https://www.w3.org/Bugs/Public/show_bug.cgi?id=29113

29117 default-collation and xsl:merge-key
https://www.w3.org/Bugs/Public/show_bug.cgi?id=29117

RESOLVED: to adopt the proposal in this bug report.



> 7.1 Test suite bugs

> 20672 type-functions-0304

> 20843 function-1902
>   Relates to 29060

> 28926 Merge with accumulators - no tests

RESOLVED: Abel Braaksma to take ownership of making tests for merge
with accumulators.


> 28927 Streamable stylesheet functions - no tests

RESOLVED: Abel Braaksma to take ownership of making tests for
streamable stylesheet functions.


> 29052 error-0640o-2


> 7.2 CR Progression Planning and Discussion

>     7.2.1  Testing - test case development, test case coverage and test
> reports - status report

>     Test coverage -MK
> https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0023.html

> https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Sep/0024.html

>     7.2.2 CR Exit Criteria: OK for single implementation of optional
> features only; otherwise require at least two implementations.
> Exception to this is for the Streaming Feature; we require 2
> implementations of it.

>     7.2.3 Any features "at risk" for final recommendation

-- 
****************************************************************
* C. M. Sperberg-McQueen, Black Mesa Technologies LLC
* http://www.blackmesatech.com 
* http://cmsmcq.com/mib                 
* http://balisage.net
****************************************************************

Received on Thursday, 10 September 2015 19:12:58 UTC