2016 05 19 Minutes

METS Editorial Board Teleconference Minutes

May 19, 2016
Attending: Karin Bredenberg, Tom Habing, Betsy Post, Leah Prescott, Tobias Steinke, Brian Tingle, Nate Trail, Robin Wendler Absent: Bertrand Caron, Terry Catapano, Richard Gartner, Jukka Kervinen, Andreas Nef, Merrilee Proffitt

Minutes from the previous minutes were approved.

Karin provided an update of the "PREMIS in METS Guidelines" revision. It is still a work-in-progress. The committee will meet again in four weeks. They are preparing a list of questions for the METS Board. We should revisit this at the next METS Board meeting.

Next there was a discussion of changes to the METS Tools and Utilities page. The Board received a report of broken links. The report was added to the METS Github Issue Tracker. The link to the JHOVE2 link was easily corrected to point to the JHOVE2 project's BitBucket site. The i2s link was for a digitization product from a French vendor. After corresponding with the vendor, Betsy, determined that the correct link for this should be this, http://www.i2s-digibook.com/products/digital-library-solution/limb-gallery/limb-gallery-2/. It points to LIMB which has superseded the yoolib product. There was also a broken link to a tool from U. of Alberta. Betsy will see if she can contact anyone there, but in the meantime the link should be taken down from the tools page.

There was a discussion of where the next face-to-face Board meeting should be held. The DLF meeting (joint with Digital Preservation 2016/NDSA) will be in Milwaukee, November 7. It was pointed out that the next meeting should be in Europe if we want to stick to alternating years. Several options were mentioned including DC in Copenhagen, October 13, and iPres in Bern, Switzerland, October 3. Bern could be expensive to attend. There was some preference shown for DLF, but not unanimous, but if DLF is the choice we would need to act fast. DLF extended the deadline for proposals to May 23, so we should act fast to make sure we can all attend. Betsy will Doodle poll to select possible sites; send any other suggestions to Betsy. The consensus was that we would not do a workshop this year, just a meeting.

Finally, the spreadsheet was discussed. There were several questions about the spreadsheet and the process.

First, it was clarified that we are analyzing all the METS sections, not just the dmdSec as initially described.

There was a question about what it meant when a column had "yes", "no", or was just left blank or in some cases was described as not applicable, etc. People seemed to be following different practices. Many left the value blank if the profile did not mention an attribute or element, but others seemed to use other conventions, especially for mandatory, optional, or repeatable elements, such as using numbers ranges like 0-1 or 1-n.

Regarding a profile not mentioning an element or attribute, it was decided to leave the field blank in the spreadsheet, and Leah volunteered to color-code or grey-out all the fields deliberately left blank to make this clear.

After a bit of discussion it was also decided to use the words Mandatory or Optional in combination with Repeatable for the cardinality of different elements:

• Mandatory == 1
• Optional == 0 or 1
• Mandatory, Repeatable == 1 to many
• Optional, Repeatable == 0 to many
• Not Used == 0 (only used if the profile explicitly states that element should not be used; otherwise leave blank)(suggested by Karin after the meeting)

If there are more complex rules for the cardinality, you can describe it as necessary, such as “Mandatory, but only if ...” or “Exactly two, one logical and one physical” or something similar. Because the spreadsheet is intended for human interpretation and not machine processes, this approach was deemed adequate.

There was also a discussion as to what the spreadsheet will be used for. After the community discussions at the iPres METS workshop, it was clear that users still preferred that an XML Schema-based version of the METS standard be maintained, along with the possibility of a new XML-based METS-lite version to make the schema more accessible or easier for a broader audience. The spreadsheet is intended to inform any decisions as relate to the design of lite version of METS, which elements and attributes are being commonly used and which are not. There was some discussion as to whether an analysis of the METS profiles would really map to use cases or not or if there might be better ways to discover these. The possibility of analyzing actual METS documents was briefly discussed. There was also a discussion of how the spreadsheet might be used as part of the next steps, such as how we might analyze the results by assigning different people to do short analysis and write-up for a particular METS section.

There was some criticism of how the spreadsheet was laid out in terms of needing to cross-reference yourself or repeat the same info in multiple different sheets or columns, but there might not be a good way to avoid that.

The consensus was continue to work on completing the spreadsheet with it continuing to be a topic in future meetings.

The next topic was introduced with the question, "Are there any new developments for PCDM?" Brian mentioned that much of the Hydra PCDM work is using Curation Concerns which is Ruby on Rails engine.

Tom announced that he had posted the raw XML for all of the registered METS Profiles in GitHub. This was done mostly to facilitate the creation of a new XSLT for rendering the prfiles as HTML; since it is just the raw XML it isn't really human-readable. There was also some discussion of the XSLT that Brian did some time ago (~10 years) which summarized the features of all the registered profiles at that time, Index to registered METS Profiles by Features Used.

The meeting was adjourned after about 40 minutes.