Fall 2016 Face to Face, Vienna

Face to Face Meeting, Vienna, 25–27 Sep 2016

Present

Hugh Cayless (HC)
James Cummings (JC)
Peter Stadler (PS)
Elisa Beshero-Bondar (EB)
Magdalena Turska (MT)
Elli Mylonas (EM)
Raff Viglianti (RV) [remotely from DC]
Syd Bauman (SB)
Stefan Majewski (SM)
Lou Burnard (LB)
Martina Scholger (MS)

Agenda Items

TEI Simple Finalization—Monday 11:00
Report on first stylesheet hacking meeting
Stylesheets Hacking / Knowledge Transfer—Sunday Afternoon
Release Planning: MT & JC release technicians with EB documenting the process
Build a new (Pure)ODD processor?—Sunday Afternoon
Support for ODD->PDF output?
F2F Meeting in Prague? Any presence from TEI Council for the XML Prague itself?
Whither TEI Session / Martin's Paper / What Council does and should do
Council Meeting Scheduling
Feature to suppress examples from Guidelines in customized doc — SB

Sunday 09-25, 09:30–17:00

Location: ACDH Meeting Room

Morning 10:00–13:00

PS reports on the CMC SIG - they are interested in some/a new elements. Should put in a feature request. Council has homework to read the TEI journal article.
- - http://wiki.tei-c.org/index.php/SIG:Computer-Mediated_Communication [beware: current ODDs are project customizations and need to be merged/gerneralized for a FR]
- http://jtei.revues.org/476
Next release: when?
- Work on making releases less cumbersome
- Not ready for release before Members Meeting.
- LB wants to prepare anyXML as last stage in PureODD before next release--this may not need much work.
- Also clearance for TEI Simple. LB: this may take longer.
- http://teic.github.io/TCW/anyXMLproposal.html
- Proposal for release: 1st week of December
  - http://teic.github.io/TCW/tcw22.html (editable at https://github.com/TEIC/TEI/blob/dev/Documents/TCW/tcw22.xml )
  - MT & JC release technicians with EB documenting the process
    - EB should document the requisite tools/components in use during the release
    - SB: add a step to the process: Review of Release Notes, together with Freeze (on release branch)
    - Discussion of branches assoc with release HC: Master is always the latest stable branch. Release branch = candidate to become Master.
      - Scripts need to be properly identifying the branch--the release branch for each release has a dedicated number: the numbering is hard on the Jenkins maintainers.
      - Naming process of release branch needs better documentation. Give it one name without a number, like releaseCandidate. HC: not good git policy but easier for Jenkins maintainers. RV: It's potentially risky to keep reusing the same branch name.
      - SB and HC: reverse concern: what if old release branch wasn't properly cleaned up--and these merge with your current release candidate branch?
      - What is possible to make the process really streamlined, up to the point of having just a single place to fill in the release number?
      - (We decide to defer discussion as we prepare this release)
- Perhaps appoint people to tidy up the Release Notes before the release
  - PS: print out git logs and commit messages from last release. https://www.atlassian.com/git/tutorials/git-log/formatting-log-output e.g. in mark down output `git log --pretty="* %cn: %s [%h](https://github.com/TEIC/TEI/commit/%H)" --since=2016-04-01`
  - SM: label the commits to filter out the trivial ones
  - All of us on Council need to
    - read the Release Notes (TCW22)
    - RV: List which issues have been closed, and which aren't related to the release. LB: This used to be done years ago, when LB wrote the release notes--it's hard to work out which commits are relevant.
    - HC: If closed tickets are assigned to the release milestone that will make this easier.
    - How to write a git commit message: http://chris.beams.io/posts/git-commit/
    - Try not to pile lots of different alterations in one commit!
    - Use git commit --amend to correct a mistaken commit message (unless you’ve already pushed of course)
Issue handling recommendations (now included in TCW22):
- Assign tickets to release milestone
- Use ticket reference in any commits addressing or fixing the issue
- At least for the final commit addressing the issue try to prepare the commit message so it can be used for Release Notes
- Label the issue as Release Note (green label) if it’s important to include the note about it in Release Notes
ANY XML, discussion of LB's proposal
- http://teic.github.io/TCW/anyXMLproposal.html
- See also discussion of https://github.com/TEIC/TEI/issues/1399
- New element spec for anyElement.xml. It's an empty element, like textNode.
  - Currently we can define the <content> element with RelaxNG, but in new proposal we cannot use RelaxNG, only TEI ODD. This is a major change.
  - RV: This is going to be a problem, especially with MEI which needs RNG-defined datatypes (and element content e.g. wit rng:list). We still need Relax-NG. LB: We have macro.schemaPattern to define data types. RV: Testing necessary(!)
- First this needs to be tested, but if it passes, we should deprecate macro.anyXML
Lou's suggestion for a template: <xsl:template name="matchanyElement"> <element xmlns="http://relaxng.org/ns/structure/1.0"> <anyName> <xsl:choose> <xsl:when test="@except"> <except> <nsName ns="{@except}"/>  </except> </xsl:when> <xsl:when test ="@include">   <xsl:attribute name="ns"> <xsl:value-of select="@include"/> </xsl:attribute> </xsl:when> </xsl:choose> </anyName> <zeroOrMore> <attribute> <anyName/> </attribute> </zeroOrMore> </element> </xsl:template>

Discussion: Need to write this code to be as generic as possible Where does this belong in the Stylesheets?

Afternoon

Stylesheets!

SB reports on the first meeting of the Stylesheets Self-education Group
- Subgroup worked on https://github.com/TEIC/Stylesheets/issues/161
Need to improve the documentation to give a better idea of which stylesheets and what parts of them do what
Need to know which versions of stylesheets are where (oxGarage, Saxon, oXygen)
This link is for general users, but we need to find what generates it (the Makefile generates it!): http://www.tei-c.org/release/doc/tei-xsl/#start Click on any of the profiles.
- Group by component, or group by mode.
- We need high-level documentation of modes, and mapping of functions and purposes of each stylesheet [Kind of TEI Stylesheets Lite, Getting Started Manual]
- Need documentation of the order of processing.
- Knowledgeable people: SB, Martin Holmes, HC, JC, LB
- EM: How do we write this documentation that tells us how to begin?
  - SB: Martin Holmes gave a presentation that provides a good start on this.
  - Possibly ask him to reprise and start doc based on his presentation
- Methods to document: internal comments and/or external overview document. Write in the xsd:prose element so it's output as documentation. What, why, input, output.
make test runs tests for the Stylesheets. There is another makefile in /tests that can be used to test a subset of the xslts (e.g. for odd, html, docx, etc)
make doc puts the documentation into release/xslcommon/doc/tei-xsl/index.html
ACTION ITEMS: LB produces high level overview using Martin’s presentation from SC and other materials as a starting point. http://hcmc.uvic.ca/people/martin/tei/anatomy_of_a_transformation.pdf
Build a new (Pure)ODD processor (discussion that was part of the self-education session)?
Perhaps consider issues related to these tickets ( #1462 , #1473 ) when discussing this (RV) .
Discussion about running TEI and Stylesheet make scripts locally - on Mac and Windows respectively.
ACTION ITEM: Hugh writes doc for installing components to allow making stylesheets on the Mac
ACTION ITEM: [Someone] writes doc for installing components to allow making stylesheets on Windows.

Monday 09-26, 09:00–17:00

SM unavailable, MT dividing time between Council meeting and eXist-db workshop

Morning

Ticket Triage

TEI Simple Discussion

+ LB: If Council will support this, it needs good documentation; Getting examples from EEBO would be good. The other question is about marketing: should this replace Lite? Consensus: No. + JC: Nexts steps to check documentation and go forward as we are. + LB: - Is this meant to support EEBO TCP? * Not solely - To support 90% of users 90% of the time? * No. that is more lite’s job + MT: meant to support a large number of projects + JC: meant for modern, western printed books ‘printed books by dead white men’, (e.g. printed: not manuscripts, books: not weird non-codex objects, dead: not modern weird textual objects, white: vaguely western European in background, men: expressing that it is a limited scope that doesn’t cope with a wide range of text or authors. Clearly, this description isn’t meant to be exhaustive. + LB: revised checklist of its aims: - it’s aimed at print not manuscripts - More early modern than other periods, not modern - Western European traditions and characters - Larger scale, shallow encoding projects, e.g. for libraries, archives - As soon as you need to do deeper encoding, then you should be using your own customisation - It can of course be used outside of these parameters, but not what it was mainly designed for + RV: Worried about how a beginner transitions from Simple to where they need to go + RV: what Roma should be like is to have this help in making decision where to start your customization: asking questions about the character of material and project + Discussion of entry points and the difficulty of establishing "prescriptivist" approaches for people new to TEI and how it might be a bad idea. + Side note from LB: helping people decide on their starting point in the TEI world with a ‘checklist’ or ‘decision tree’ approach + Relationship with TEI BPL. Should the header chapter in Simple match recommendations of TEI BPL? Probably not, but would be good to get Libraries SIG view on the TEI Simple Header. - https://github.com/kshawkin/Best-Practices-for-TEI-in-Libraries - Minutes from group meetings http://wiki.tei-c.org/index.php/Workgroup_to_revise_the_Best_Practices_for_TEI_in_Libraries + Action on Council members to invite SIGs and community members to contribute their exemplary ODDs + Action on Council members to speak to Board about new home page for TEI website, re how people "start" with TEI (meeting tomorrow): - Re how TEI Simple is presented - Other options for beginning (including redesigned ROMA when it's ready) + Council agrees that it is not the marketing arm of the TEIC, while they are happy to help the Board with some technical issues Council have more than enough Guidelines issues, software development issues, etc. to keep them busy for a long time. (Especially post-SPQR, and now with LB leaving Council this will also have an effect.) + Council should advise Board that new web maintainer should have some experience with UX work.

Lunch break:

Lou, Hugh, James - Film backup video for Sebastian’s Event before joining others for Lunch. (We did this. Hopefully, it will never see the light of day.)

Afternoon

Work on tickets (see table below)

Tuesday 09-27, 12:30–17:30

MS and PS at the Translatathon workshop SM unavailable MT partially unavailable in the afternoon for the XQuery workshop

skype link up with http://www.oerc.ox.ac.uk/sebastian-rahtz (10:50 their time) -- Hugh, Lou, James to present; Dress Code: Loud shirts.
Then tickets

12:15–14:00, Joint lunch / meeting with Board

http://wiki.tei-c.org/index.php/Board_Agenda_2016-09

14:00–17:30

Note: This is Lou’s last face to face TEI Technical Council meeting (since he didn’t stand for re-election), Council wishes to thank him for his long service to the TEI Council. It fully expects to consult him as a member of the community.
Tickets

Tickets for TEIC/TEI

Number	Status	Title	Notes
1502	Go	<desc> and remark in TEI Guidelines	Closed
1499	Go	Misuse of @extent in an example	MS: in progress
1478	Go	Make attribute values in Guidelines examples match suggested values in attribute class (related to #1397)	EM - doesn’t need discussion
1475	Go	Deprecated attribute @render discussion needs revision	EB: Will close once I remove the commented-out portions.
1468	Go	<content> examples need pure content	MS: No discussion needed
1447	Go	Exemplar schema for ENRICH includes duplicate ids for Schematron constraint	JC: No discussion needed - bug - not investigated why yet.
1446	Go	Solve base-uri problem in Stylesheets before next release	Closed
1441	Go	the future of the Debian packages	Ask Stefan (EM pinged via git)
1436	Go	consider adding @xml:space as global attribute in Tite	MT: doesn’t need discussion
1428	Go	Refine elements that can contain <secFol>, <signatures>, <catchwords>	EB: With commit 59f1f5a we added a deprecation warning, but modifications won't be made until end of deprecation period (2018-09-30). Changed status to Blocked and Deprecated.
1427	Go	Restructure chapter 11	Do we need to discuss this? Not yet ready for prime time inmho.label changed to orange. Awaiting draft
1425	Go	Allow <bibl> in <witDetail>	Prod MT; MT feeling duly prodded
1420	Go	<valItem> should allow <remarks>, too	MS?
1419	Go	<lg> and <gap>	[SB to give quick report]
1417	Go	Define useful constants for classSpecs, e.g. $MEMBERS	RV: SKIP, low priority
1414	Go	ogrophy elements should be in att.canonical	MT: it’s Go to propose broader solution for dealing with ogrophy elements SB: decision (reinforced today) is to go ahead with <idno> 1st child(ren) w/o waiting for other ography element re-org.
1407	Go	abstract model constraint problems on model.pLike elements	SB no need to discuss (I had forgotten we already made it a “go”.)
1373	Go	Need for some way to represent macro.anyXML in Pure ODD	Progressed yesterday: waiting for trial implementation in stylesheets
1349	Go	odd4odds should be in Exemplars	SB — no need to discuss (but y’all should periodically send me complaints that this isn’t done. :-)
1344	Go	data.temporal.w3c—taste the pain	SB—needs brief discussion; I had not realized it was assigned to me, and am not sure what next step should be.
1333	Go	Link checking in Examples
1331	Go	examples combines @from with @notAfter	SB — more discussion needed. I thought this was done & closed but, not surprisingly, the constraints we added last time were objected to by at least on person: MH.
1293	Go	hors d'oeuvre unpalatable in 9.4	Waiting on SM
1290	Go	dated info on XML processors	SB— no need to discuss. SM and I are in middle of working on this
1267	Go	TEILite has @calendar, but no `<calendar>` element it can point to	Fixed in source. Need to check that schemas for exemplars are generated correctly during build
1196	Go	<attRef> needs better documentation	MH did. Closed.
1175	Go	use of modal verbs in Guidelines	JC: Since this is a periodic check, I'll go through large parts of more newly added prose and review our use of modal verbs again.
1082	Go	update ODD documentation on www.tei-c.org and in Guidelines	Nothing to discuss here. Move along.
1045	Go	Most attributes lack good examples	No visible progress. Close? Mark it Blocked
960	Go	i18n revision due
869	Go	half title pages in TEI Tite	No progress since March. Ask Kevin if we can reassign to someone else. Or mark it blocked.
618	Go	@expand example in att.lexicographic is incorrect.
564	Go	create <schemaRef> element to give pointing to ODD.	JC: Creating <schemaRef> as agreed by Council, just need to time to do.
548	Go	Give more structure to abstract	Closed for now
543	Go	Make the content model of occupation more elaborate	HC: Done
537	Go	Support for non-ISO schematron (1.4) should be removed	SB — requires brief discussion in how to proceed.
536	Go	Make @source global	JC: Complicated, agreed, nothing to discuss, just need to do it.
533	Go	Add profileDesc as optional child biblFull	HC: Done
527	Go	Replace current eXist for Roma with updated version	Closed
498	Go	Permit new <persona> as a child of <person>	No progress on this since April. Prod LB!
486	Go	deprecating members of a content model	Closed (yay!)
464	Go	Need for some way to test the oxygen-tei package before release
379	Go	retaining punctuation marks in the text of a TEI document	No progress since March: blocked? Now assigned to EBB & EM

Groups for Monday

A: PS, JC, MS
B: LB, RV, EBB
C: HC, EM, SB

Groups for Tuesday

F: JC, SB, LB
G: EBB, RV, EM, HC

Group	Number	Status	Title	Notes
A	1505	Discuss	Use of `@next/@prev` versus `<join>` with `@scope/@result` and `@exclude` needs clarification	Use of @next here is unusual, but not forbidden by Guidelines, and we think we shouldn’t change the current practice. The current suggested practice by Guidelines and examples is to be preferred. No need seen for additional clarification of join/@scope unless martinH produces concrete proposal.
B	1500	Discuss	A plea for behaviourSpec	Needs more discussion and elaboration, we dont understand it
C	1497	Discuss	Don't impose horrible algorithmic burdens on implementors	Lots of agreement on this one, it’s GO, and Syd writes minority opinion on ticket
A	1496	Discuss	Don't punt on model semantics	Should still leave to implementor what happens but strongly recommend that any processor will at least warn of ambiguous model elements. It is probably an error in the ODD not in the implementation so the user should be made aware of this rather than the processing be made automatically happy by hiding it. [Council argues for a schematron rule to forbid this unlikely occurrence.]
B	1495	Discuss	Confusing note on modelSequence	We think we understand Magda’s comment, we’re not sure if we agree
C	1494	Discuss	Clarify modelGrp semantics	Recommend: make it GO, change to modelSequence, fix wording
A	1493	Discuss	Allow <subst> in <line> please	Subgroup thinks more interpretative elements should be removed from line, not more added. To add more is madness. Council agrees completely that where <substJoin>, <add>, and <del> are allowed, <subst> should be, too. However, there is significant disagreement as to the solution: allow <subst>, or disallow the others (and <app> and ...)? Generally council thinks it should remove all this bumf from line.
B	1492	Discuss	Allow note in rdgGrp	We agree that note shdnt go in rdg unless it’s a part of the reading itself. Problem is deciding what is being annotated: sequence is not very good way of linking note with thing annotated. But its used elsewhere in this way. We agree to allow note in rdgGrp. STATUS GO
C	1489	Discuss	datatype of prefixDef/@ident	Recmend: Make Go, Syd fixes with Martin’s suggestions STATUS GO
A	1487	Discuss	Schematron rule disallows perfectly good markup	Deprecation mechanisms not followed because we didn’t think anyone would be doing this and it isn’t deprecation, it is extra constraint through schematron. And we think Martin shouldn’t be doing this. STATUS CLOSED
B	1486	Discuss	Location of ODD validation file for spec development	We agree with Hugh . STATUS GO
C	1484	Discuss	person and place elements should allow 'name' child similar to org.	Needs discussion; Why shouldn't Simple just change the content models to permit <name> STATUS GO
A	1483	Discuss	parameters to the TEI media type (generalizing their usage + a repository)	Seems like a good idea, assigned to Piotr STATUS GO
B	1477	Discuss	Merge 11.3.1.6 and 11.3.4.4	We agree with Gerrit STATUS GO
C	1476	Discuss	Source links for examples sometimes missing; all have disappeared from HTML output	Make it GO. Add schematron warning, divvy up 1514 examples and find sources or specify no source or unknown source Our thinking is to divvy ‘em up into sets of 100 un-sourced <egXML>s, and hand each Council member a set to do.
G	1474	Discuss	Need to improve GIS/gazetteer encoding support in TEI
B	1473	Discuss	PureODD may need other forms of datatype restrictions besides regular expressions	We would need a new pure ODD mechanism to do numeric testing. Needs work on a proposal. Or we have to do it in schematron.
C	1471	Discuss	new section on deprecated datatype macros	Let's have the discussion on deprecating datatype macros!
F	1470	Discuss	Example from address spec file is weird
B	1466	Discuss	Add @place to <head>	We agree. Add head to astt.placement, as an alternative to @rend. Example looks good. STATUS GO
C	1465	Discuss	Xenodata oddities	Closed
G	1463	Discuss	Relation of @style and @rendition needs clarifying	Suggest that we accept HC’s wording
B	1461	Discuss	Add new element <unit>	We are reluctant to add new element. Whats wrong with rs, or an ODD solution? STATUS NO
F	1459	Discuss	Description of <textNode> is insufficient	Dunno.
G	1453	Discuss	`<path>` should exist alongside `<zone>` for non-closed areas in facsimile	No, don’t do, look at SVG and IIIF, close or leave for discussion
B	1452	Discuss	@class attribute on <msContents>, <msItem>, and <msItemStruct> should all be added to an attribute class.	We agree. And note that other elements could be added to this class e.g. <p> STATUS GO
F	1431	Discuss	clarify how to encode short-form citations	SubGroup thinks it is daft - you should use <ref> and if not you could use @corresp. Update existing examples to use `<ref>` and not `<bibl>`.
G	567	Discuss	Allow <relation> to occur intermittently within parent elements	Prod SB
B	562	Discuss	allow paras in various children of ographies	Which elements currently containing model.phraseSeq should become model.specialPara ? LB says all, for simplicity’s sake.
F	539	Discuss	New element annotatedU	Prod LB
G	507	Discuss	create generic policy for the TEI on integration of external standards	Changed to blocked because not enough time
F	386	Discuss	free-standing attributes -> class	Prod EM + SB https://docs.google.com/document/d/1Punte188LwdqBko8m8-f9LzevP7EKoqhddUBlnXVSiA/edit
G	377	Discuss	Improve guidance and restrict usage of biblScope	Leave as is, but file bug report with zotero? JC notes they get it from us: http://wiki.tei-c.org/index.php/ZoteroToTEI As documented here: https://www.zotero.org/support/kb/tei
F	374	Discuss	Encoding of Standoff annotations	Prod PS
G	367	Discuss	rationalize content models of org and place and person	Prod SB
F	299	Discuss	att.editLike should not bring att.dimensions & att.ranging	F: Group suggests it should be GREEN: GO; Do it with modification that Alexey recommended. (I.e., we think the list of actual elements should be reviewed, and pending that review this should just be done w/o any further ado — no further posts or deprecation needed.)
	1501	(Discuss)	Cannot generate valid DTD within oXygen	This is a release blocker!!!
	1400	(Discuss)	Remove numbered divisions	Agreed to make it a P6 issue
	1350	(Discuss)	some documentation wrong/missing for correspAction	PS: Already done
	1081	Blocked	fDecl doesn't allow att.datcat yet
	941	Blocked	GLs say add 'ns' decl, but roma objects
	820	Blocked	"Add new attribute" errors, etc.
	672	Blocked	video html to tei
	519	Blocked	precision should be a member of att.datable*
	482	Blocked	The term strikes back - terminology chapter
	470	Blocked	att.measurement and att.dimensions overlap
	325	Blocked	Allow certainty etc. inside milestoneLike elements
	293	Blocked	add SourceForge feeds to http://www.tei-c.org/Activities/
	290	Blocked	att.canonical for model.persTraitLike	CLOSED due to being overtaken by other tickets