Minutes from October 14, 2011

Minutes of the meeting of the SIG on Libraries 14 October 2011 ZHSG 1.005, Universität Würzburg

Kevin Hawkins (University of Michigan), co-convenor, called the meeting to order at 9:09 a.m. Central European Time.

Kevin introduced himself and his co-convenor, Michelle Dalmau, who was unable to attend. Other attendees introduced themselves:


 * Syd Bauman (Brown University)
 * Laurent Romary (INRIA & HUB)
 * Maud Medves (INRIA)
 * Martin Andert (Martin-Luther University Halle-Wittenberg)
 * Andreas Münzmay (OPERA, University of Bayreuth)
 * Christof Schöch (University of Würzburg)

We were later joined by Stefan Majewski (formerly University of Vienna and now Austrian Academy of Sciences) and Morfudd Jones (Llyfrgell Genedlaethol Cymru / National Library of Wales).

Kevin summarized work done on the *Best Practices for TEI in Libraries* ("BP"), released the previous night. Laurent suggested that the BP be recast as guidance not only for libraries but for anyone doing mass digitization.

Kevin summarized his work on aligning the version of TEI Tite used in AccessTEI with the canonical version in SourceForge. He also summarized his and others' work assisting a Google engineer with automatic generation of TEI markup conforming to something between Level 3 and Level 4 from books scanned through their library partnership but said he wasn't clear on Google's plans for making their scanned books available as TEI.

Syd asked whether Google's use of the BP was documented. Kevin said he doubted that Google would produce it but said that producing such documentation based on what we can discern might be worthwhile.

Kevin noted two suggestions for future activities from past meetings of the SIG:


 * Creation of stylesheets to convert Tite to BP Level 3 and an unofficial BP Level 3.5.
 * Support for FRBR modeling in TEI

Syd suggested that the SIG recommend how to express in a TEI header in a machine-readable way which FRBR Group 1 Entity (work, expression, manifestation, or item) is the object of encoding.

Laurent suggested creating a registry of all modifications or implementations of Tite and the BP so that people beginning new projects don't need to reinvent the wheel in making modifications for their use. Kevin and Syd asked whether it would be appropriate to have a registry of all TEI customizations, not just those based on Tite and the BP. Laurent said he felt that the needs of mass-digitization projects are specific.

Kevin offered some reasons to and not to turn over the BP to Council for ongoing maintenance. After discussion, it was agreed that the SIG should submit the BP to Council so that the levels can be listed on the Customizations page on the TEI website and included in Roma. We should also make sure they are included in oxygen-tei. We can tie the ODDs to particular versions of P5 to ensure that they won't break as new releases are made. Most importantly, maintenance by Council will ensure that there is coherence across the various TEI customizations and that and that the Council can find people to make improvements to the BP in the future.

Kevin suggested creating a wiki page listing TEI Analytics, TextGrid's Baseline encoding, and the levels BP -- all of these are meant to be allow for interoperability across text from various projects.

Laurent suggested evangelizing for the TEI in libraries. Kevin said that it was unclear how much use libraries were actually making of TEI but said that libraries that help faculty with technology should know about it to recommend it to users looking for support in a digital project.

Syd said that creating a stylesheet for converting Tite to the BP should be easy. Laurent suggested incorporating into OxGarage (by creating a tei-xsl profile in SourceForge). Syd will work on this based on some documents encoded according to Tite that Kevin will provide.

The meeting was adjourned at 10:25 a.m.