W3C

– DRAFT –
ixml Group Teleconference

13 December 2022

Attendees

Present
Bethan, John, MSM, Norm, Steven
Regrets
Joel
Chair
Steven
Scribe
John

Meeting minutes

<Steven> https://lists.w3.org/Archives/Public/public-ixml/2022Dec/0000.html

<john> s/Person_Person//

<john> s/Person+Person//

previous actions

Joel to send details of his implementation to Norm, Norm to put it on the web site's splash page. - still outstanding

MSM to comment on the QT4 issue suggestiong an XPath function supporting ixml and possibly suggest further function signatures. DONE

Norm to update free-standing document on EBNF to BNF translation (PR #146) - Outstanding

<Github> https://github.com/invisibleXML/ixml/pull/146 : First draft of EBNF to BNF conversion document

Norm to prototype an alphabetic reference list of elements in the vxml vocabulary, in XHTML, to serve as a basis for discussion. Outstanding

SP to edit group page to say we meet at 15:00 UTC (during the winter, at least). DONE

Implementations

Steven: his implementation is more (but not quite) compliant. Lots of work on the Unicode classes - Unicode12 moved to Unicode15
… uses grep and sed to generate the internal tables
… Some errors in Unicode spec found (e.g. omega and Ohm)

No changes to implementations by Norm, John, Bethan, MSM

No changes to testsuites

Issues and bugs

Issue #147- Norm's proposal adopted. Issue an erratum

<Github> https://github.com/invisibleXML/ixml/issues/147 : Are control characters allowed in an ixml grammar?

ACTION: Norm to issue an erratum for issue 147

<john> s/for this/for Issue #147/

<Github> https://github.com/invisibleXML/ixml/issues/147 : Are control characters allowed in an ixml grammar?

Issue #146 - awaiting Norm's action

<Github> https://github.com/invisibleXML/ixml/pull/146 : First draft of EBNF to BNF conversion document

Issue #139 - still awaiting more work.

<Github> https://github.com/invisibleXML/ixml/issues/139 : https://github.com/invisibleXML/ixml/blob/master/samples/URI/rfc-3987.ixml

Steven: the specifications aren't really good

ACTION: Steven to propose a route forward on Issue #139 (URI/IRI grammar)

<Github> https://github.com/invisibleXML/ixml/issues/139 : https://github.com/invisibleXML/ixml/blob/master/samples/URI/rfc-3987.ixml

Steven: the syntax is very loose indeed on some items such as IRIs that aren't valid URIs

ACTION: Steven to propose an iXML grammar for IRIs and illustrate the places where it might fail.

Issue #137 - open pull request by Norm (#167) awaiting resolution, constucting iXML grammar from sections in the spec.

<Github> https://github.com/invisibleXML/ixml/issues/137 : We should document the XML tag set that results from parsing an ixml grammar with the ixml specification grammar

<Github> https://github.com/invisibleXML/ixml/pull/167 : Recraft the gen-grammars PR on top of the latest master

<john> s/contructing/constructing/

ACTION: Steven to review Norm's pull request 167

Issue #13 - Renaming proposal - MSM not really ready to discuss - perhaps needs a meta-discussion on v.Next?

<Github> https://github.com/invisibleXML/ixml/issues/13 : Renaming proposal (for v.Next)

Steven: supporting renaming, but no specific ideas yet, though use cases exist.

MSM: two forms of renaming - static and dynamic. i) Static: The name is given in the grammar but isn't a non-terminal. ii) Dynamic: the name is read from the input (e.g. parsing XML)

Steven: the latter is a case of context-dependency, which is a huge issue.

MSM: If there is a schema, then it's possible, but for general XML we need to extract the name of an element from the input data.

Steven: Use case - recognise as A, but output as B.

ACTION: MSM - open dynamic naming as a separate issue

Issue #75 - combining grammars

<Github> https://github.com/invisibleXML/ixml/issues/75 : Combining grammars (for v.Next)

Norm: Doesn't think it is difficult but Steven reports encountering a number of issues, such as scoping, whether it appears in the XML.

Norm: there would have to be decisions on such scoping effects

MSM: Scoping issues: RelaxNG seemed to make a good attempt at such combinations. Perhaps having primitive 'combinators' from RelaxNG analogues

ACTION: Norm to produce a concrete proposal. John offers to assist

<john> Next meeting: 2023 January 10

Summary of action items

  1. Norm to issue an erratum for issue 147
  2. Steven to propose a route forward on Issue #139 (URI/IRI grammar)
  3. Steven to propose an iXML grammar for IRIs and illustrate the places where it might fail.
  4. Steven to review Norm's pull request 167
  5. MSM - open dynamic naming as a separate issue
  6. Norm to produce a concrete proposal. John offers to assist
Minutes manually created (not a transcript), formatted by scribe.perl version 196 (Thu Oct 27 17:06:44 2022 UTC).

Diagnostics

Failed: s/Person_Person//

Succeeded: s/Present+Person//

Failed: s/Person+Person//

Succeeded: s/(omega/(e.g. omega/

Succeeded: s/this/issue 147/

Failed: s/for this/for Issue #147/

Failed: s/contructing/constructing/

Succeeded: s/then the/then/

Found 'Next meeting:' not followed by a URL: '2023 January 10'.

All speakers: MSM, Norm, Steven

Active on IRC: Github, john, Steven