TEI Technical Council Minutes 2015-05

, Ann Arbor, MI

Initial Agenda: http://wiki.tei-c.org/index.php/Council_Agenda_2015-05

Participants

Council: Syd Bauman (SB), Lou Burnard (LB), Hugh Cayless (HC – Chair), Fabio Ciotti (FC), James Cummings (JC); Martin Holmes (MH), Stefan Majewski (SM), Paul Schaffner (PFS), Peter Stadler (PWS – Local Organizer), Raffaele Viglianti (RV), and Stefanie Gehrke (SG) intermittently via Skype,

Guests: Sebastian Rahtz (SR) for TEI Simple discussion. Lauren Romary (LR) for Standoff Discussion.

Thursday, 2015-05-28

TEI Hackathon report (JC)

SIG Reports

Fall f2f meeting

MH on Building the Oxygen Plugin

Roma Redux

Note: Council would like to thank Paul and Paula Schaffner for hosting a marvelous BBQ at their home — a fascinating place in its own right, and a happy birthday to Fabio (FC).

Friday, 2015-05-29

MH presents the oXygen plugin, the concept, features and build process; Action on MH: turn into TCW documentation after developing on wiki.

Standoff

Council had a skype discussion with Laurent Romary about standoff

TEI Simple

Materials:

Discussion by Skype with Sebastian Rahtz, project Co-PI, with Sebastian introducing the Simple Processing Model and its background. There was some worry about re-use of <rendition> for output. Council suggests <outputRendition> if project amenable.

Council suggests giving some examples that are not related to TEI Simple; e.g. manuscript description or prosopography, also to generate XForms interface. Sebastian suggests XForms might be more challenging but msDesc could be done with existing behaviours.

Council recommends to stop calling it a function library. Call it a method/signatures/rendering API of some sort. Would be good to get Guide on API (or behaviour library), and examples in multiple implementation languages. It must not look underspecified. Document that users of the processing model will need to decide how to implement.

Suggests a @mode attribute on <model> in order to keep @output a fixed list (e.g of ‘Web’, ‘Print’ etc.).

What will the Council have to do in order to support processing models? Do we need to support a reference implementation, for example? Yes, there is a reference implementation in the form of an XSLT function library that implements the behaviours defined in the PM ODD.

Note: There is an Implementation in XQuery of PM by Wolfgang Meier;

you can see the showcase machine that Wolfgang set up at http://showcases.exist-db.org/exist/apps/tei-simple/index.html (for editing user: admin pass: simple)

To install on your local machine, grab and run http://data.exist-db.org/eXist-db-setup-2.2-develop-3f8357b.jar

To recompile the whole thing locally, it requires the newest development build of eXist as described here https://github.com/wolfgangmm/tei-simple-pm

For example HTML function library at: https://github.com/TEIC/TEI-Simple/blob/master/polygon/html_functions.xsl

or LateX (incomplete) at: https://github.com/TEIC/TEI-Simple/blob/master/polygon/latex_functions.xsl

Council consensus was to agree to the TEI Simple proposals, with minor changes suggested below.

Actions for TEI Simple Project:

<xsl:attribute-set name="font">

<xsl:attribute name="fname">Arial</xsl:attribute>

<xsl:attribute name="size">14px</xsl:attribute>

<xsl:attribute name="color">red</xsl:attribute>

</xsl:attribute-set>

or <tei:funcVal><tei:funcName>fname</tei:funcName><tei:val>Arial</tei:val></tei:funcVal>

Pure ODD (presentation by LB)

cf. https://sourceforge.net/p/tei/feature-requests/530/

Council appreciate LB’s and SR’s sterling work on this spec and the associated processing.

msPart (feature request 505)

Finishing off msPart revisions (538; will assign to a group for discussion later)

@selector proposal (MH):

http://wiki.tei-c.org/index.php/Proposal_for_@selector_attribute_on_rendition

counts as a best practice model for proposals to the Council!!

Action on MH: wait for two weeks, then implement if no one objects.

SB points out that once @selector is implemented, we can deprecate the use of @render on <tagUsage> for this purpose. Once that’s done, there is no longer a need for @partial on <tagsDecl>, we can go back to insisting on one <tagUsage> for each element type within <text

I'm sorry, but I do not remember anyone suggesting that we allow this to alter the TEI Abstract Model. Rather, we spoke of the need for prose that says “don't do this” and perhaps formal checks to guard against it, because it breaks the abstract model. >.

Saturday, 2015-05-30

App. Crit.

See:

Robust discussion. There appear to be several options on the table:

The objections to #3 come down to:

Council proposes that:

Things to keep in mind for evaluating Syd’s approach:

Council Terms and Elections

HC pointed to a proposal on Board to change the date of elections and the starting date of terms. Consensus was that Council does not mind earlier elections, but wants terms to continue to start on 01 January.

Next telecon: Action on Hugh to make a Doodle poll to schedule.

Meeting adjourned

Council commended Paul Schaffner for doing an excellent job as local organizer.

Feature Request Review

Group A: PFS, SM, FC, SG

Group B: JC, MH, SB

Group C: RV, PWS, LB, HC

Amber Feature Requests

GRP # Summary Owner Creator Created Notes
A 556 Allow <hi> to be contained by <m> Hugh Cayless Caroline T. Schroeder 2015-05-27 hi is allowed within w because it was added individually to model.hiLIke. Easy enough to do the same for m. Would allow epidoc to be consistent in their use of hi to record such things; OTOH, there are alternative mechanisms available, e.g. use of seg. m and hi may seem to belong to different spheres but it cannot be denied that there are people who wish to operate in both. Group is divided, marginally in favour of allowing this change. Action: HC to implement
A 555 improve constraint constraints Syd Bauman Syd Bauman 3 hours ago Syd go for it, sounds sensible. If Syd has concern with it, please raise and explain. Action: SB to go for it [DONE]
A 554 guidance on use of @calendar and @datingMethod Gabriel Bodard Simona Stoyanova 10 hours ago Martin’s document (referred and linked to in SF comments) seems very clear and well explained. Could be followed. GB (again, see SF comments) has volunteered to implement. Voting for doing this. Action: Assigned to Gabriel Bodard to implement
B 553 JTEI schematron validation issues Lou Burnard 2015-05-05 Issues 1 & 2 are solved. Issue 3: Council mildly recommends making this a warning instead of error. Action on MH: Close ticket. [DONE]
B 552 clarify how to encode author of subset or part of document Kevin Hawkins 2015-04-15 SB Volunteers to add clarification prose to the note. Action on SB: Add clarification.
B 551 No abbr element in JTEI Lou Burnard 2015-04-06 Council recommends adding abbr (but not expan or choice). Action on MH: Add <abbr> element. [DONE]
C 550 Provide place to record the Unicode Normalization Form used rviglian Jens Østergaard Petersen 2015-03-24 This appears to be more a XML issue than a TEI one. Unicode also provides quick ways of checking for normalization: http://unicode.org/reports/tr15/#Quick_Check_Table Some clarification in the Guidelines may be needed anyway (eg add prose to projectDesc.) Action on RV: add clarification to guidelines
C 549 More flexible dateline for periodical edition tracking M. H. Beals 2015-03-22 We don’t think the example constitutes a dateline. → close, reject
C 548 Give more structure to abstract Laurent Romary 2015-03-15 We think Laurent’s request for structured abstract is reasonable, but are unsure whether to create a new <section> that can only contain <head> and <p> or to allow <div> but constrain it (presumably with Schematron) when in the header. (note: non transcriptional div and non transcriptional p for new element proposal)
A 547 Allow g in notatedMusic rviglianti 2015-03-13 Example does not appear to be notated music in the full sense at all. If there is a case for both marking the presence of notated music and transcribing to some extent the contents of the same, then two approaches suggest themselves: either allow a text-wrapper within notatedMusic (on the analogy of figure), e.g. ab or seg or both, which itself could contain either text or text-equivalents like <g>. Or allow a <choice> between <notatedMusic> and its textual representation. Rejected. If accepted, should probably be in the form of option 1, i.e. adding seg and/or ab or equivalent to notatedMusic, certainly not <g>. General rule should be that <g> should be allowed everywhere that character data is allowed, and also vice versa**. Should perhaps open a new ticket requesting the addition of a text-bearing element to the content of notatedMusic.** Action: close-reject (PFS)
A 546 <defaultVal> should be removed from all specs Syd Bauman Lou Burnard 2015-02-08 We are not happy to opine definitively on this issue, on which Lou, Sebastian, Martin, and Syd have already had a full discussion. That discussion seemed to be moving in part at least toward mutual understanding if not consensus. It became clear that default values express a necessary decision in the generation of schemas (not just dtd or xsd) and that simply removing default values effectively leaves that decision to the XSLT, without documentation in the ODD. The suggestion that a constraint be added to warn that if <defaultVal> is used when usage="req", or if the value of <defaultVal> is not in a closed value list seems welcome (and was implemented approximately 2 months ago). We suspect that the alternative suggestions for the main point (dropping defaultVal and replacing it with something else, or redefining defaultVal to be documentary only) have unknown implications and should perhaps be deferred. Action: MH to report on cases where defaultVal is used.
A 545 Deprecate oVar and pVar, Revamp oRef and pRef Stefanie Gehrke Laurent Romary 2015-01-31 We don’t have enough data on the usage of the less common forms (pVar and oVar), so we cannot judge confidently whether to drop them, i.e. whether there might not be people who will be affected. We may nevertheless support deprecation of *Var and extension of the content model of the *Ref variants. Adding @notation to pRef is a good idea. possibly in a class, have to check if @notation is used at other places. Action: MH to ask e.g. FN language encoders if they are using it. Action: SG to ask on TEI-L and TEI Ling SIG lists if anyone is using them.
B 544 align oVar/ @type with the general taxonomy of grammatical features Stefanie Gehrke Piotr Banski 2015-01-30 Agree and recommend using datcat attributes. Action on SG: If Piotr agrees to suggestion, implement. (but bear in mind above ticket 545).
B 543 Make the content model of occupation more elaborate Fabio Ciotti Laurent Romary 2015-01-22 Agreed. Make content model macro.specialPara; also should raise new ticket for bringing other similar elements (<residence>, <nationality>, <climate>, etc.) in line with this. Action on FC: Implement. Action on SB: create new ticket for other elements DONE (new ticket, FR #563, created 2015-06-04)
B 542 Make occupation part of the content model of author Fabio Ciotti Laurent Romary 2015-01-22 NO. Council sub-group believes strongly that <occupation> should not be allowed as a child of <author>. It should reference a <person> if you want this kind of detail. (Furthermore, there’s lots of cruft allowed inside <author> that should not be.) Action: Waiting on Laurent to respond to JC’s comments otherwise close.
C 541 Using `<zone>` for points and paths rather than shapes Stefan Majewski Martin Holmes 2014-12-21 Need clarification for use cases - e.g. what is the need to finding a line on an image as it has no area? → clarify definition of zone: should always be a closed area. Action on MH: Make a proposal for a path element
C 540 expanded recommendations on values of @xml:id rviglianti Kevin Hawkins 2014-12-18 Raff working on this. Will produce some examples illustrating best practices. Action on RV: Produce examples and recommendations.
C 539 New element annotatedU Lou Burnard Lou Burnard 2014-12-05 > to Friday 29 discussion
A 538 guideline example <msPart> Stefanie Gehrke Stefanie Gehrke 2014-12-05 Go for it, this appears to be very clear Action on SG: change to green, implement
A 537 Support for non-ISO schematron (1.4) should be removed Syd Bauman Martin Holmes 2014-11-19 Ask Syd for a list of changes this would require. We have apprehension that this might break many other things. Action on SB: Do something.
A 536 Make @source global James Cummings Hugh A. Cayless 2014-11-18 May be feasible to define a global @source broadly enough to replace all three of the local @source attributes currently in use; or at least two of them. Most of the arguments against making @source global would apply also to making @resp and @cert global, which we chose to do anyway. (But @source does have a particular bibliographic meaning.) Consensus very very skeptical about making it global. Action on HC: Produce set of examples for expansion of where @source should be allowed for reconsideration.
B 533 Add profileDesc as optional child biblFull Hugh A. Cayless Laurent Romary 2014-11-14 Laurent needs independent headers; our options are:* Add model.profileDescPart to the biblFull content model. * Make the content model of biblFull an alternation between its current model and the teiHeader model. * Reintroduce independent headers by making teiHeader a child of listBibl. * Why not just use teiHeader? * Just do nothing; Laurent can customise. Action on HC: Investigate implementation of #2
B 531 New element for secluded text Hugh A. Cayless Hugh A. Cayless 2014-10-17 We request that Hugh create an elementSpec we can look at, with the expectation that it will be approved. Action on HC: Create draft elementSpec and pass it by council for review.
B 525 translated from Syd Bauman Fxp33 2014-09-14 Council subgroup 2015-05-28: Assign again to SB to follow up, and with the recommendation that he develop a proposal for @translationOf, to be available on all text-bearing elements. Action on SB to follow up, but possibly create a proposal for @translationOf on a new ticket. Done, sort of: e-mail sent to fxp33 2015-05-29
C 521 Allow <lem/ > to Contain Chunk-Level Markup Hugh A. Cayless Gregor Middell 2014-07-31 > to Saturday 30 discussion
C 517 transcription of quranic "end of ayah"-sign Elli Mylonas Tobias J. Jocham 2014-07-07 Action on HC to poke Elli Mylonas and reassign if she isn’t planning to finish it.
C 515 re-think model.global's membership and use in <line> Lou Burnard Syd Bauman 2014-07-02 Needs more investigation by LB Action on LB: Investigate more.
A 511 New element <transcriptionDesc> as a possible child of <encodingDesc> Martin Holmes Thomas Schmidt 2014-06-26 Discuss along with annotation proposal. Action: On Original Poster, give us something we can discuss publicly.
A 507 create generic policy for the TEI on integration of external standards Syd Bauman Kevin Hawkins 2014-05-10 Subgroup wonders if this considers datatypes. Policy should deal with usages of W3C datatypes/CSS, for example. More contextual information (even if history cannot be used), examples, etc. would make it much clearer. Guidance/recommendations against relying on external standards that are not stable. Action on SB to pick this back up
B 505 Redefine <msPart> Stefanie Gehrke Caroline T. Schroeder 2014-04-29 We basically like the originally-proposed msFrag solution best; it’s clean and straightforward. We see no strong objections raised to it so far. Whole Council agrees. Action on SG (with MH): generate elementSpec for <msFrag>, and insert a subsection into the prose.
C 504 Replace @active and @passive on relation with @from and @to Hugh A. Cayless Hugh A. Cayless 2014-04-21 relation for rdf-like encoding is abuse. We should remove rdf-like examples from the guidelines and recommend better ways of representing rdf-like statements within TEI (eg with graph) or by embedding RDF-XML. Action: close-wontfix [DONE] Action: HC will create new FR for handling RDF
C 498 Permit new <persona> as a child of <person> Lou Burnard David J Birnbaum 2014-02-22 needs more investigation by LB Action on LB: Investigate more
C 496 `<taxonomy>` should be allowed as a child of `<category>` Martin Holmes Martin Holmes 2014-01-31 Action: MH to bug everyone on the Council list about this, and if no objections in 2 weeks, go ahead.
A 482 The term strikes back - terminology chapter Lou Burnard Laurent Romary 2013-11-06 Possible we just didn’t discuss it last time?
A 479 Adding the @place attribute to <head> and <seg> Sebastian Rahtz Ville Marttila 2013-10-29 Action: HC to close, since there’s been no followup
B 464 Need for some way to test the oxygen-tei package before release Martin Holmes Martin Holmes 2013-07-05 Agreed, we don’t see a practical way to have good test coverage. Maybe a manual process that is defined in a release-manual is enough. Action: MH to test straightforward things, and not think of really complicated tests.
B 459 warn user of dropped constructs Syd Bauman Syd Bauman 2013-06-17 Closed, is already implemented. Related tickets with partial aspects still open.
B 457 make explicit difference between tagUsage and ODD documentation Lou Burnard Kevin Hawkins 2013-05-24 Action on LB: Remind LB to have a look at bullet 4 of the last comment.
C 384 free-standing attributes -> class Paul Schaffner Sebastian Rahtz 2012-09-15 Action on PFS: Create list of changes needed.
C 378 Encoding of Standoff annotations Peter Stadler Javier Pose 2012-08-26 This ticket shd be linked to stdoff proposal.
C 360 New attribute @keepHyphen Paul Schaffner Martin de la Iglesia 2012-05-03 Action on PFS : to provide examples, one simple, one complex, so coded as to represent Lou’s preferred solution (two different glyphs for two different kinds of hyphen -- not the original suggestion of a kludgy @keepHyphen attribute. ) See comments on ticket.
C 326 an <object> element Sebastian Rahtz Sebastian Rahtz 2011-10-14 No activity from SIG., We propose ticket should be closed as now over 4 years old. Can always be reopened if action happens. Action on HC to close with comment.
C 366 rationalize content models of org and place (etc) James Cummings BODARD Gabriel 2014-01-26 Could close ticket until JC stops having real work to do.
C 324 Allow certainty etc. inside milestoneLike elements BODARD Gabriel BODARD Gabriel 2012-06-03

Amber Bugs

Group A: MH, PWS, RV, HC (on Saturday)

Group B: FC, SM, SB

Group C: LB, JC, PFS, HC

GRP # Summary Owner Creator Created Notes
A 754 Need for a standard way to warn of class changes affecting ODDs Martin Holmes 6 days ago > close. It’s the user’s responsibility to associate their ODD with a specific release, or deal with “nightly” builds.
A 753 <app> is phrase-level Hugh A. Cayless Marjorie Burghart 2015-05-18 Defer to tomorrow’s discussion.
A 752 Inconsistent extension on ODD template file Martin Holmes 2015-05-13 > possibly change from xml to odd extension; Action on MH: check with SPQR about motivation for odd extension. [DONE/closed; the problem resolved itself, apparently.]
B 751 <event> not allowed in <org> Syd Bauman Syd Bauman 2015-04-24 Council subgroup agrees;. Approach appears to be clear. Acton on SB: poke list to review branch P5/for-bug-75/1 [DONE, although I haven’t seen my post show up on list yet], and if no objections in 2 weeks implement
B 749 fatal bug in 2.8.0 apropos att.global.rendition Sebastian Rahtz 2015-04-14 already fixed, confirmed by reviewing the source. (ticket closed)
B 747 Updating Roma/ eXist on release day: improve the process Martin Holmes 2015-04-04 Action on PWS to fix the problem, making sure that Roma, and any other release scripts still work.
C 746 Adjustment of example at Guidelines 16.2.5 Joel Kalvesmaki 2015-03-25 Action : assigned to LB, who fixed it.
C 745 Menu item has bad link Kevin Hawkins Joel Kalvesmaki 2015-03-25 already closed-fixed (MDH)
C 744 Broken clarosnet/ Ashmolean links in relation example Martin Holmes Martin de la Iglesia 2015-03-13 Action: HC open new ticket, close this one
A 740 "freestanding event" in 13.3.2.2 Stefanie Gehrke Piotr Banski 2015-02-20 Action on SG: Check whether <gi> elements are missing and rephrase according to the proposal. mark green
A 738 documentation for <media> vs. att.media Peter Stadler Piotr Banski 2015-02-18 This is really a feature request, that overriding part of an inherited model does not mean you are no longer a member of that model, and the documentation should reflect that. Action on MH to open a BugUG and close the current ticket. [DONE]
A 737 explaining NFC Stefan Majewski Jens Østergaard Petersen 2015-02-18 Was just a clarification in the text, have fixed this commit r13223 [DONE]
B 736 compatibility characters among East Asian logographs James Cummings Jens Østergaard Petersen 2015-02-18 Accepted and fixed [DONE]
B 735 older strata Fabio Ciotti Jens Østergaard Petersen 2015-02-18 Accepted and fixed [DONE]
B 730 Unicode history Hugh A. Cayless Jens Østergaard Petersen 2015-02-18 Subgroup thinks that a better correction should be “ by the time the Unicode standard was first being developed”.
C 728 lamenting the passing of SGML SDATA Stefanie Gehrke Jens Østergaard Petersen 2015-02-18 Fair cop. Revise this section and rename it … maybe delete the whole thing. http://www.tei-c.org/release/doc/tei-p5-doc/en/html/CH.html#D4-45-2
C 727 dated XML processor info Stefan Majewski Jens Østergaard Petersen 2015-02-18 Comment SM: I find this information not dated, I think this still holds, even though many people use UTF-8 today.
C 724 part of the "Title" clipped by Roma, in a somewhat sneaky fashion Hugh A. Cayless Piotr Banski 2015-02-18 Is the Council able to support Roma? In any case, this should be transformed into a Github issue Action on HC: mark green, assign to MH (and duplicate issue on GitHub)
A 721 hors d'oeuvre unpalatable in 9.4 Stefan Majewski Piotr Banski 2015-01-30 **Action: SM implement (**Update the inline examples for what the referenced dictionaInries REALLY provide)Include example of use of @notation e.g. for sampa
A 718 dated info on XML processors Syd Bauman Jens Østergaard Petersen 2015-01-21 The first sentence could be slightly rewritten. We think the whole chapter could stand to be revisited, but this is low priority.
A 710 xml:lang/ langUsage Syd Bauman Jens Østergaard Petersen 2015-01-21
B 706 revisionDesc suggesting order of changes rviglianti Sebastian Rahtz 2015-01-17 Subgroup suggest to drop “Conventionally” or even to drop the whole sentence Action on RV: Fix all examples in the guidelines;Remove ‘Conventionally’.
B 703 nonsense in remarks of <specGrp> Syd Bauman Syd Bauman 2015-01-03
B 697 sequence of declaration info is lost when making an ODD Lou Burnard Lou Burnard 2014-10-31
C 696 macro.paraContent doesn't contain l James Cummings Hugh A. Cayless 2014-10-18 But it does already, HC obviously smoking crack. (HC had done previously, but forgotten to close the ticket) CLOSED.
C 695 distributed spec files point at wrong schema location Peter Stadler Peter Stadler 2014-10-12 This is already fixed
C 630 use of <listRef> in specs Lou Burnard Kevin Hawkins 2013-12-28 Low priority: LB to try harder
A 581 `<attRef>` needs better documentation Martin Holmes Martin Holmes 2013-06-24 Action: MH to improve documentation
B 563 inconsistent encoding of citations to sources of examples Paul Schaffner Kevin Hawkins 2013-04-12
C 548 use of modal verbs in Guidelines Kevin Hawkins Kevin Hawkins 2013-04-08 Action on JC: He MUST -- no, SHALL -- review uses of “should” (etc.) and distribute them among MUST (or SHALL), SHOULD, and possibly MAY (or alternative formulation altogether.) JC also to add a short discussion on Modal Verbs and their use in the Guidelines.
A 216 half title pages in TEI Tite Syd Bauman Kevin Hawkins 2010-08-09

Check Green Bugs / FR for updates on status?

Check RED Bugs / FR for updates on whether should just be closed or potential status change?