Navigation:
Documentation
Archive



Page Tree:

Child pages
  • 2010-09-24 CI Conference Call

This wiki space contains archival documentation of Project Bamboo, April 2008 - March 2013.

Skip to end of metadata
Go to start of metadata

Collections Interoperability Conference Call, 24 Sept 2010

Attending (INCOMPLETE!): Tim Cole, Robert McDonald, Marlita Kahn, Steve Masover, Bruce Barton, Peter Gorman, Scott Prater, Jim Muehlenberg, Tom Habing, Jon Dunn, Bill Parod, Harlan Wallach (joining late), Charles Blair; Harriett Green; ...(Please edit to add missing names)

Introduction

-----------
Tim: Draft agenda – Ongoing discussion of CI Scope & Dependencies; Assembling and refining list of candidate standards, protocols, and specifications supporting / required for Collections Interoperability; picking times for regular calls

Marlita: scope definition most important; status reports also important

Bruce: Also, calls between CI and WS again? regularly scheduled calls? regular is Tues afternoons
Tim: perhaps representation from CI instead of joint calls

Notes from call (drafted by Steve Masover; edited by Tim Cole)

-----------

(Disclaimer: attributed text is paraphrased and approximate, not exact quotes; apologies for any incorrect attributions)

Discussion of CI Scope

Tim: proposal excerpts pertinent to CI at https://wiki.projectbamboo.org/x/YAA3AQ .
Content as cross-cutting concern. Rich enough descriptions, resource provenance, identifying objects – all relevant to enabling scholars' work with a wide range of content. What can Bamboo services operate on. Identify standards and protocols that need some fleshing out, e.g., CMIS
– a skeleton that needs flesh to be a fully an application profile.

Marlita: WS model for defining scope has been working well. Questions that needed to be answered to fully define scope, people answered these; look at use cases derived from colleagues and also from planning wiki (narratives); then outline at a high level some functional specifications. Get down on paper in some order to see what we've defined and what we've missed. Threaded
throughout: what are intersections with other areas ... design, tech implementation, schedule.
?? : Uses cases might be a good way to organize overlaps. A common set of use cases across groups might be useful.
Tim: Yes, I think also there might be additional cases that focus on interoperability. So perhaps three child pages or groups of pages related to
scope: questions, use cases (identified or new), functional reqs for collections interop. Two or three people within group on call to focus on each of these – draft in two weeks? Then larger group can respond to it.

Marlita: tag narratives on wiki with CI-specific tags? This was useful to WS team to relate use cases from BPP to what we're doing going forward.
??: Areas of intersection described by other areas? That might be a helpful starting place.
Tim: Yes, call with WS group. Also on Platform Requirements pages has some questions/dialog around CI. That should inform our work, certainly, but we should also take a fresh look and consider our experience in other projects.

Steve: Transformation services
Tim: For example tools that work on TEI docs ... so, perhaps some stylesheets that transform between TEI sub-standards. Increase reach of tools.
Marlita: Tools and Services registry pages – also some questions there.

Tim: Think our way through some questions before we make tags and start tagging narratives in BPP wiki. E.g., when and how to use persistent identifiers, when do you need a new identifier because an object has changed. Volunteers to work in these areas?
?? : Some questions may fall out of use cases. Is functional reqs first cart before horse? Common set of use cases?
Tim: Fair point.
Marlita: Planning phase use cases (narratives). https://wiki.projectbamboo.org/x/-YBw ... and WS Tech Wiki

Jim: Questions dredged up from early summer meeting?
Tim: Yes, please create a page.

Tim: Use cases?
Peter: I can do that.

Tim: I'll work on the functional specs, but I would like some help on that. Perhaps convene a call next week or the week after?
Jon: I could do that.
Tom Habing: I can join in.

Tim: Overlap calls with WS
Bruce: Maybe as needed rather than standing? We can give you a couple of weeks to get some traction....
Tim: That's good.
Bill: Standing back to review materials before joining in.

Harlan: announcing self

Tim: Steve, platform intersection?
Steve: Common provenance data, cross-cutting ability to expect access and core metadata to be harvested by BSP services. Specific data sets to be shared among areas of work (WS, BSP, TSR ...) that want guidance from CI about how to represent ...

Discussion of metadata standards, protocol, candidate collections, etc.

Tim: segue to 2nd item, standards, protocols, specification ....
Tim: Persistent identifier standards; identify resources and parts; versioning resources; metadata schemes; preservation metadata, rdf & structural frameworks; OAI-PMH; updates via Atom/RSS; remediation / transfer services ... So we may want to segment candidate standards/protocols/specifications into categories that reflect concerns on these levels. Then we need to narrow down focus. This may take some time and iterations.

Scott: A definition of collection, of a digital object, that is fairly rigorous might help focus and frame.
Jim: Somewhat like what WS is doing ....
Bruce: Looking at entities of interest in Work Spaces. E.g., person.

Then AuthNZ. What's the domain, then what are the rules of the road for that domain.
?? : What do you get when you follow an identifier for an object?
Metadata? Data stream, in which of various formats/protocols?
Tim: Moby Dick. Related identifiers for metadata, versions, TEI version, PDF version.
?? : Means of discovery for different objects and classes re: what is associated with it.

Tim: Much work on metadata schemes, but also issues with many of them.
Must look at what particular applications have done (app profile) in using these schemes Charles Blair: Jen Riley, map of metadata standards ... there are so many that apply ... difficult to present meaningfully.
Tim: Yes. A shame Jen will not be at Indiana going forward ...
Charles: Also a matter of where standards are authoritatively cataloged.

Bruce: Just looked at that map ... intimidating! ... are there "carrier standards" we can focus on, as opposed to standards that are interior to a discipline. By "carrier" I mean, what are standards for shipping content around from one place to another.
Charles: Packaging standards?
Bruce: Perhaps, yes.
?? : Standards change all the time. One thing I would be interested in working on is how to plug and play standards as they evolve. This is something we've been looking at @ Wisconsin.
?? : Classes of standards we need to support. Provenance, agent, descriptive metadata. Then look at candidates within each class. A way to hack away, categorize.
Marlita: Related to Bamboo deliverables. To bound scope of what we need to deal with.
Steve: agree. classes plus relevant to Bamboo a way to bound
?? : A way to manage change over time as well.
Peter: Scott and Peter
Tim: Great. Start with classes, get some comments on that, then populate.

Bill Perod: Collections to target?
Steve: ARTFL, Perseus, Google Research
Jim: Early discussions
Tim: HathiTrust
Marlita: Corpora Space team ...
Jim: There's a list of candidates.
Tim: Will find that list and forward link.

Future calls

Tim: Last thing before break is future calls. No real clear consensus from Doodle Poll ... would like to get on a regular schedule, can cancel or make it a sub-group meeting if there's not general business. 1st/3rd or 2nd/4th best?
Steve: Neil is doing a "general" Doodle poll. Which day and time works.
Tim: Will try after getting a bit more info about what times are blocked out by other groups/meetings. Twice a month schedule. Week of 10/11 most likely time for next meeting of this group. Recorded version of this available until next call.

Action Items:

Questions: Jim
Use cases: Peter
Functional Specs: Tim and Tom
Metadata classes: Scott, Scott, Peter

Marlita: Please embed comments in pages, with your initials and date. (i.e., rather than putting comments at bottom of wiki pages)

Tim: Thanks!
Call ends.

Back to Collections Interoperability Home

  • No labels