- From: Michael Kay <mike@saxonica.com>
- Date: Fri, 30 Oct 2015 15:49:32 +0000
- To: Public XSLWG <public-xsl-wg@w3.org>
1 Roll call - Need manual roll call
Sharon, Anders, Michael Kay, Abel. Plus Luis in listening-only mode because of technology problems.
2 Assign minute-taker - MKay
3 Approval of minutes:from XSLT Telcon 22 Oct 2015
https://lists.w3.org/Archives/Member/w3c-xsl-wg/2015Oct/0018.html
§§ approved
4 Action Items
ACTION-2015-04-23-001: MK to add a section to the back matter suitable
for pointing to the RelaxNG schema.
pending
ACTION 2015-10-01-001: Mike: turn proposal on streamability
(see https://lists.w3.org/Archives/Public/public-xsl-wg/2015Sep/0006.html) into a bug
entry and create a more detailed wording to be included in the spec
status: completed see bug 29229
ACTION 2015-10-15-005: Abel to produce a test coverage report for next
telcon.
§§ done just before meeting: see
https://lists.w3.org/Archives/Public/public-xsl-wg/2015Oct/0044.html
https://lists.w3.org/Archives/Public/public-xsl-wg/2015Oct/0045.html
ACTION 2015-10-22-001: Mike Kay to respond to
https://lists.w3.org/Archives/Public/public-xsl-wg/2015Oct/0030.html
giving his rationale for his view that Abel's email does not raise any
new issues.
§§ done after this telcon: see
https://lists.w3.org/Archives/Public/public-xsl-wg/2015Oct/0047.html
ACTION 2015-10-22-002: Mike Kay to make a formal response to the
I18N comment [XSLT30] collection lang processing should refer to BCP47 re
BCP47 raised on the public-qt-comments list.
§§ Pending. Needs priority.
§§ Done after this meeting: see
https://lists.w3.org/Archives/Public/public-xsl-wg/2015Oct/0048.html
ACTION 2015-10-22-003: Mike Kay and Abel to update and resend their
documents related to Appendix F on their implementations both their
progress on implementing and how they are handling
implementation-defined features.
§§ Pending. Not absolutely essential for entry to CR, though desirable.
ACTION 2015-10-22-004: Mike Kay to propose text for bug 29210.
completed - review text in Bug Report - see comments 3,4,5
§§ DONE. Bug is now marked resolved.
ACTION 2015-10-22-005: Mike Kay and Abel to discuss optional features
and features at risk and make a proposal for next telcon.
Status?
§§ DONE during the telcon.
ACTION 2015-10-22-006: Mike Kay to close all closable "resolved" bugs.
Status?
§§ DONE, for some definition of “closable”. Since the meeting I have marked all resolved bugs as closed.
ACTION 2015-10-22-007: Sharon to make the appropriate contact with
XML Prague for some XSLT 3.0 interoperability "session"/"event".
§§ DONE.
completed after telcon. Need to OK Jirka's plan and decide if we want a meeting in Prague around the conference.
§§ Suggestion is to do this on the Thursday before the conference. We'll be competing with other streams so it needs to be made interesting. Abel suggests it would be more appealing if the focus was wider than just interoperability. Anders suggests one hour on interoperability and one hour on hands-on new features. Mike thinks the two objectives can be achieved in combination. Sharon would prefer not to be competing with other streams.
5 Other administrative business
• Review of agenda - items to be added to Agenda?
§§ Added a couple of items
5.2 Telcon schedule
New CR date MUST be determined at this call.
§§ See “publication schedule” below.
5.3 Interoperability workshop for XML Prague
5.4 Jirka invited us for F2F meeting in Prague before/after the
conference. Shall we accept this opportunity?
§§ Mike: yes. Thinks we will have a range of issues arising from interoperability testing. Hard to predict how much time we will need. Perhaps do Mon-Weds XSL WG meeting, then Thurs interop workshop, then conference.
§§ Sharon to write to Jirka accepting the invitation.
5.5 Do we need to discuss anything regarding EXI and the JSON Schema
questions?
§§ MK has responded on behalf of both WGs. This may be the end of the matter, or they may push back.
6 XSLT Draft
W3C Internal Working Draft 12 October 2015
§§ Actually the latest is 23 Oct 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/
• Abstract
§§ There's little in the process document that constrains this.
§§ We reviewed the abstract and decided it was fine. Sharon is concerned that we present the story as regards XPath 3.0/3.1 accurately and in a way that answers any potential objections.
6.2 Status Section
a) Process document is 2015 - http://www.w3.org/2015/Process-20150901/
b) Call for implementations
§§ Need to add words to this effect.
§§ Need to change "Publication as a last call WD".
c) How Long do we want CR period? Minimum is 4 weeks. Query/XPath
has been in CR for 1 year (not quite)
§§ Suggest we go for 4 months. Suggest 31 March 2016.
d) Need to list features at Risk
§§ Streamable stylesheet function
§§ Streamed grouping
§§ Requirement to report all instances of non-guaranteed-streamability
§§ Suggest crafting some language that points out that the report can come from a third-party tool, so it's not necessarily something that has to be in every processor.
ACTION-2015-10-28-001 MHK, in regards to stating that the requirement to report violations of guaranteed streamability is “at risk”, to add some language that points out that the report can come from a third-party tool, so it's not necessarily something that has to be in every processor..
e) Do we want to request anything else from the public during this
review?
§§ MK: prefer not to have specific requests for feedback on particular decisions, it makes it seem we couldn't decide for ourselves.
f) Anything else to say about xpath 3.1?
§§ Noted also that our spec in its current form cannot "overtake" XP31 (or Serialization 3.1) in the process. But we don't think this is something we need to highlight as a risk or dependency.
g) Patent policy
§§ Not planning any exclusions.
h) Expectations for Next Steps
§§ We expect the next step to be PR.
i) Provide link to sections detailing changes since 2.0 and also
since the Last Call Draft.
§§ The links are already there
j) Do we highlight bug 27615 in Status section (we have to highlight
for director but I am not certain it needs to be in the Status section).
§§ (This is the I18N one. Should add a Note on this because they asked us to.)
§§ - POST MEETING: I have put a para in the Status section on this; Sharon please review.
§§ Also need exit criteria from CR. "Two implementations of a "basic XSLT processor" and of the streaming feature, and a single implementation of all other optional features".
• Conformance - see section 27 and review.
§§ Noted that XQuery invocation is very vaguely defined. Perhaps we should make this a feature at risk.
§§ DECIDED to make XQuery invocation a feature at risk. (If we retain this feature we may need to say something about load-xquery-module().
§§ Later in the meeting we changed this decision and removed the feature entirely.
§§ Noted also that higher-order-functions should be an optional feature.
§§ ACTION-2015-10-28-002 MHK to raise a bug to say HOF should be an optional feature. Could bundle in removal of XQuery invocation (on the grounds that it's superseded by load-xquery-module()).
§§ DONE after the meeting. See bug 29251
6.4 Incompatibilities
6.5 Features at Risk - definitive List
MK and ABr Action item - action not yet completed.
§§ DONE during the meeting
6.6 Implementation Experience (Questions copied from Process Doc)
a)The group must document how adequate implementation experience
will be demonstrated.
b)Are there tests or test suites available that will allow the WG to
demonstrate/evaluate that features have been
implemented (e.g., a matrix showing how different pieces or classes of
software implement different features)?
Yes, should have link to preliminary report for Directors' Call
c)Is the expectation to show two complete
implementations?
Yes, with exception of optional features that must be
discussed at this telcon.
d)What are the Group's plans for showing implementation of optional
features? In general, the Director expects mandatory features and optional features that affect
interoperability to be handled similarly. Optional features that are truly
optional (i.e., that do not affect interoperability) may require less
implementability testing.
e)Does the WG have additional implementation experience that will
help demonstrate interoperability (e.g., has there been an interoperability day or workshop? Is one planned?)?
Yes (planned) - Jirka has offered time on the Thursday prior to XML
Prague 2016.
6.7 Requirements
Requirements Document
JSON support
Maps
Arrays
Do we wish to highlight these (we started some and put into XPath
3.1) or other things.
Need to review requirements document.
:
http://www.w3.org/TR/xslt-21-requirements/
§§ We know the requirements document is incomplete. We need to show in the transition document that in such areas, we know where the requirements came from.
§§ ACTION-2015-10-28-003 on MHK to review Requirements document and point out any areas where we do not meet the requirement.
§§ We did this partly during the meeting and I did more after the meeting: see
https://lists.w3.org/Archives/Public/public-xsl-wg/2015Oct/0046.html
§§ Haven't done 2.9, 2.10
§§ 2.13: is done in XPath 3.1
§§ 2.14: partially done
§§ 2.15: not done as described, but the underlying requirements were met by other facilities such as static parameters and higher order functions.
§§ Not included in the requirements document: maps, JSON, etc. (MK include this in the action above).
6.8 Review Updated Appendix F
6.8.1 "compare" MKay and ABraaksma's implementations
Is this information in a state where I can use some of it
in the Transition Request or at the Directors Call to show
interoperability?
6.9 Decision to go forward
• Normative References – dependencies??
• Publications schedule.
§§ Deferred until next week. Pub date is 17 or 19 November, everything works back from there. Vote on 5th; There should be time for this, but worst case is slip to the 24th.
7.0 Spec bugs
28937 [xslt 3.0] Keeping XSLT 3.0 maps in sync with XPath 3.1 maps
No issue to discuss. Needs to be closed now.
29153 [XSLT30] have we created a loop-hole with windowed streaming
and copy-of or snapshot?
https://www.w3.org/Bugs/Public/show_bug.cgi?id=29153
See https://lists.w3.org/Archives/Public/public-xsl-wg/2015Oct/0038.html
§§ DECISION: close with the action that has already been taken.
29210 [xslt30] xsl:apply-imports and xsl:next-match from within xsl:override
https://www.w3.org/Bugs/Public/show_bug.cgi?id=29210
Resolved fixed does not show up in Bugzilla Bug List. Text in bug to
be reviewed.
§§ DECISION: close with the action that has already been taken.
29226 [xt30] Empty enclosed expressions
https://www.w3.org/Bugs/Public/show_bug.cgi?id=29226
§§ DECISION: agreed to make this change
29229 [xslt30] Streamability of XPath 3.1 constructs
https://www.w3.org/Bugs/Public/show_bug.cgi?id=29229
§§ DECISION: Accepted with a change noted in Bugzilla.
29237 [xslt 3.0] Bug in CSV package example
https://www.w3.org/Bugs/Public/show_bug.cgi?id=29237
§§ Didn't have time for this:
§§ ACTION-2015-10-28-004 Abel to consider proposing an alternative solution to bug 29237 using inline functions.
29239 Bad example in 3.5.3.6: use of focus in library package
https://www.w3.org/Bugs/Public/show_bug.cgi?id=29239
§§ ACTION Editor to treat as editorial.
29242 Overriding a required parameter with an optional parameter
https://www.w3.org/Bugs/Public/show_bug.cgi?id=29242
§§ DECISION: Accepted (but Abel might want to push back after thinking about it further).
Received on Friday, 30 October 2015 15:50:13 UTC