TEI Council Conference Call 27 Sep 06

Syd BaumanFor distribution first to Council, then to world via TEI website.

 

This electronic file is the source.

 

Converted to P5

 

Times are UTC unless otherwise noted.

Started ~12:04 with SB, SS, AK, LB, SR, JC, CW, CT, JW, DP, LR

MD had sent his apologies earlier.

 

Minutes from last meeting

    • done

    • Implemented except for French — agreed (LB dissenting) to go w/ LR’s proposal to rename fr/ files to same as en/. change French filenames, too 2006-10-11

    • completed

    • No one read the chapter and reported. CW suggests we can close this issue.

    • DB not present, deferred

    • SB reports this item has been completed. Furthermore, the changes to the Guidelines that were suggested have been enacted. The only outstanding issue from this round of discussion is that of certainty, accuracy, confidence, exactness, precision, etc. However, another issue has since been raised: how to encode regularizations of non(-proleptic)-Gregorian dates. P4, while recommending a specific format for (proleptic-)Gregorian dates permitted users to use other systems so long as they were documented in a particular manner. After some discussion on this issue we were able to reject some possibilities (e.g., that the W3C syntax be retained, but without the restriction to (proleptic-)Gregorian dates), but were not able to definitively conclude how to proceed. Both use of one set of attributes with their semantics declared in the header (ala P4) was suggested, and two parallel sets of attributes. summarize dating attributes issues for Council, with firm recommendations where possible. 2006-11-07

    • Done, see the result. However, after some discussion, we were reminded that there are really two different questions of interest:

      • what elements in the current system would be better if their content models included a limited version of model.phrase; this group includes some elements that are not found in the teiHeader (like figDesc), but excludes many that may appear in the teiHeader because they also can be used as descendants of text in a manner that would require the full model.phrase (e.g. p and name)
      • what parts of the teiHeader would be better if their content models included a limited version of model.phrase, regardless of whether or not it is currently possible (because they might also appear in the transcription, i.e. as a descendant of text)

      SB had interpreted the action item as only referring to the first of these questions. Council discussed both issues; it was decided that both are worth pursuing. Two mechanisms were identigied for verifying which elements can make use of the limited phrase class, and which require the complete model.phrase. The first was for Council members to scan the list SB produced looking for obvious errors or elements of concern. The second was to produce a schema that makes use of the limited phrase class, and for Council members to use this schema and large quantities of data looking for problems. Council chose the former method, without ruling out use of the latter later (did I get that right? —SB). consider list of limited phrase elements and propose refinements if needed 2006-10-11 Implement limited phrase system using the elements in the refined version of the list. 2006-11-08

    • LR indicated that he needed a better method of querying TEI to look for these. He was referred to ED W 84; a newer version of this is available from . post to Council a list of elements which contain either macro.glossSeq or model.glossLike in their content model, but really should have only desc2006-09-29

    • Done

    • CW reports he did so. JW reports he has some notes that he will get out to the group this week. Write a document discussing: 1. bibliographic elements (bibl, biblFull, biblItem, and biblStruct) 2. How to incorporate MODS into a TEI document 3. his own and perhaps others’ opinions on extent. 2006-10-01 He reports that MODS is much more of a cataloging approach.

    • discussion deferred, see below

    • Not done. Discussion deferred, see below.

    • discussion deferred, as MD not present

    • Discussion deferred, see below.

    • Discussion deferred, see below.

    • Discussion deferred. Note: this item was accomplished, DP & MD are now on the PB list — SB

    • discussion deferred as MZ not present

    • SB reports this was done

  • Not discussed. Note: this was not done — SB
WGs

PBDP reports that her comments were posted to the PB list, and that MM replied. DP has been thinking about this, thinks LB’s response echoes some of her own concerns. Consensus is to keep the conversation going, perhaps with DP demonstrating her system. It should be pointed out that marking up the formula limits processing as compared to an abstract representation.

PERSCW suggests another meeting, but we’re unsure of funds. ask MD if he is willing to take personography forward into places/ontology ascertain whether or not we have funding for a meeting

There is some overlap between this WG and the planned activities of the ontology SIG. SR suggests that group decide via e-mail whether or not moving forward would significantly change the existing work.

Document review

MaintenanceConsensus that DB’s document is pretty reasonable, and should be published. Some suggested that it should be presented in Victoria. revise DB’s draft document as a working paper for Council TC W 08 2006-10-20

TEXT + IMAGE encodingCouncil discussed the document recently posted by DP & CT, and then discussed somewhat on the list.

It was generally agreed that the document is divided into two parts:

  • linking text & image
  • incorporating SVG into TEI

The general feeling was that the first part is significant work that requires only a concrete formal proposal to be added, a bit more revision before being handed to the editors to be folded into the Guidelines (either as part of PH or FT). revise document somewhat, with an eye towards what parts are intended for Guidelines. by MM

As for the second part, the consensus was that SVG should not be incorporated in the TEI Guidelines directly, but that instead a sample ODD demonstrating how to do this be made publicly available.

JW suggests a name other than fax for the attribute that points to an element’s image.

Open issues

conformance— action still on MZ and JC; LR has points he intends to contribute. All Council members should continue to contribute to discussion. contribute to conformance discussion 2006-11-15 Will be on agenda for next conference call.

Other business

feature requestsLack of time precluded full discussion of this initial digest. LB asked for Council’s endorsement of his provisional prioritization: this was given, modulo some corrections (see appended list). Further topics noted:

  • Three uses: how big (for librarians), like biblScope, or structured data. Action on JW (see above) to discuss the issue.

next call 11-22; reports CW to discuss w/ editors, and post to council

Meetings

MMAs we are out of time, very little discussion of reports took place. CW will discuss them with the editors and post to the Council.

next callNext conference call is tentatively scheduled for Wed 22 Nov 06 at 12:00.

SF Feature requests

LB had posted to the list a number of items the following summary of issues for resolution. The agreed resolution is indicated as follows:

  • 1540221: Add age as a child of person analogous to sex (YES)
  • 1524368: Change content model of msIdentifier to reference class model.placeLiLike_sequenceoptional (YES, if MJD and DB agree)
  • 1442353: Add listPerson to att.typed class (YES)
  • 1550795: Change content model of catDesc to reference model.glossLike (MAYBE, at SB’s request)
  • 1058736: Add new stamp element to MS module analogous to seal (MAYBE; or generalise?)
  • 1007370: Add new theorem element as specific kind of ab (not clear in which module to put it) (MAYBE: council requested more information)
  • 1551357: Relax the content model of app to permit generic note element as well as or instead of the structured rdglem pairs. This probably needs a class. (MAYBE)
  • 1007369: Define a set of default values for rend to indicate block and numbering options (NO)

 

Call ended at 14:06.