Navigation:
Documentation
Archive



Page Tree:

Child pages
  • CI Documentation - Meeting Notes - 2013-02-08

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

Skip to end of metadata
Go to start of metadata

Attending: Bruce Barton, Tim Cole, Steve Masover, Bill Parod, Bob Taylor

Agenda

  • Goal of final documentation
  • Big picture: outline of final CI documentation
  • Candidate already-written text to repurpose
  • Section responsibilities, next steps, first review date

Notes

Overview

Wiki to be preserved at UC Berkeley – migration date uncertain, likely in March.

Discussion of ways in which Bamboo's work or work related to Bamboo's focus might go forward in the coming quarters and years.

 

Goals of final documentation

Preserve artifacts of BTP worth preserving, make them legible to folks who were not involved in Bamboo

Outline of final CI documentation

Tim: I like this outline.

Bill: I like outline too

Candidate text

Faculty Survey (Harriet Green's work)

Harriet can write a 3-5 page description of her work with charts, links out to other artifacts of the survey she conducted.

Book Model

Book model diagrams are out of date. Tim and Bill to munge text, not sure it's worthwhile to make an accurate set of updated diagrams. Bill doesn't think that the old (May 2011) diagrams ought to be used even to illustrate the Book Model concept; detail of diagrams would obscure the conceptual value.

Bruce suggests here that the CMIS binding of the book model as documented in running code on the Work Spaces end. But as important would be to record issues we thought outstanding, needing resolve; proposing some hints of where one might go in resolving these issues. To give the sense that this is point-in-time and not a complete, finished body of work.

Tim is concerned we don't try to do too much. Also because I'm not sure what of the things Bamboo did will be of most interest in nearest term. Conceptually our work in CI was interesting in two ways; validated that CMIS is probably worthwhile; and validated that there are ways to smooth out differences between repositories.

Bill believes there's a current Book Model description in text. (Found it here: CMIS Types and Paths Map for Book Model -- WORKING DRAFT)

Tim: Okay, starting to make sense, appears to be doable, to glue together from mostly-extant pieces. From Claire Stewart / Karen Miller model, through several iterations, to a somewhat reduced implementation of Book Model in code. I believe that there was modification of Perseus on the Tufts side in order to accommodate CI Hub's 'need' for "pages." We should describe, probably just a few sentences.

 

CI Hub

Bill: Good Javadoc needed.

Tim: Craig Evans indicated that he has not gotten far with this, even in the few resource classes he intended to document.

Bill: Architecture document opens a path into the code, but comes up short. Architecture document doesn't break cleanly into architecture and implementation sections. If not Javadoc, I think there needs to be something more on the implementation. Bill could produce this.

Tim: For the record, Mike and I will commit our working code here (next week); and whether Fernando's deployment of those bundles on services-qa will work. For Perseus and TCP at any rate (not Hathi yet because of API change, requirement of an app key). Connecting to Djatoka as an external resource.

Bill: Add some kind of source tree description – not exactly implementation description, but orienting to the source tree committed to the repo – into the architecture diagram (between CI Hub section and Configuration section). The trick would be to do this in a schematic way, not too wonky or detailed. Might be repurposable as Javadoc to add to the higher level classes – at package level – a middle ground between thorough Javadoc and none.

For future evolution of CI Hub:

  • Steve wonders if a set of notes from those who wish to contribute them might be better than trying to craft a coherent vision
  • Bruce favors this, believes there are/were different perspectives in the conversations to the point we took them ... would be good to let folks see us grapple with these questions.
  • Bill and Tim both like this too.

Client and user experience

Bruce: I'm on the hook to say something about the state of and intent of repository browser. Also about how we were using Fedora as a local object store, and work we did to fill out bits of unimplemented methods on the CMIS API. Then some thoughts about as the limitations of the repository browser, and possible ideas about where we might have gone if we'd gone forward.

 

Next steps

Humanist Scholars' Use of Digital Materials

UIUC (Tim, Harriet), and run by Quinn as well given her strong knowledge of Planning Phase artifacts & accomplishments. Can be done in ~ 2 weeks.

Bamboo Book Model

Tim can pull together a draft from extant pieces, but will not make much progress before ~2 weeks from now (booked until then). May not do a diagram, will see how the work develops.

CI Hub

Bill will take ~2 weeks to add an implementation section (for lack of better word) to architecture document, as discussed.

We'll each consider what (if anything) to contribute re: future evolution of CI Hub; Steve will loop in PLS and BMA to let them know their thoughts are welcome here.

Client and User Experience

Bruce to do this. Note this work is necessarily coming out of personal time given Madison commitments; still, will try to get some of this material drafted (including Work Spaces side of things) within a two week time frame

 

Where to do work?

  • Steve expects wiki to remain in place for February, migration sometime in March
  • People should use whatever documentation tool they like if it's accessible and portable to wiki – Google Doc works fine

When to meet next? 

Let's teleconf again once we have some material completed and ready for review. The teleconf would be discussion following review of others' work. Let's not commit to having this teleconf, but do so if it's needed – in about 3 weeks, toward the beginning of March.

  • No labels