OHS Launch Community Minutes: August 8, 2001

Eugene Eric Kim <eekim@eekim.com>

Version 1.0
August 9, 2001


Attendees   1  (01)

Inviting More Members   2  (09)

After Murray and Lee enlightened me on heffalumps, the meeting began in earnest. The first order of business was bringing in a second wave of members.    2A  (010)

I reiterated the reasons for evolving the group membership slowly, and the characteristics that I sought for potential members. Most importantly, our objective is to grow slowly so that we can more firmly establish our community's values and norms, and so that we have an infrastructure to bring new members up to speed quickly. We haven't quite achieved either of these goals yet, but we are well on our way. Thus, it makes sense to infuse some new blood now, people who are already familiar with Doug's ideas and our group, and who can both test our existing infrastructure and contribute to it.    2B  (011)

Everyone generally agreed with these principles. Murray emphasized that we keep this second wave of people small, and that we assess each candidate based on the stated criteria. I noted that while that all of the current members are local to the Bay Area, several of the new candidates are remote, which adds to our challenge. Lee added that these remote members won't change meeting dynamics too much, since they won't be able to attend them.    2C  (012)

The next group of invitees are:    2D  (013)

I had reservations about Grant, based on recent actions, but the rest of the group felt strongly that we should invite him and, through peer guidance, help him become a more constructive member of the community. We agreed that his invitation will be contingent on him agreeing to our guidelines to my satisfaction.    2F  (020)

Doug also listed a few people he would like to get future consideration:    2G  (021)

These three will most likely be among the invitees for the third wave of members.    2I  (025)

Integrating Our DKR   3  (026)

I described our integration efforts for our group's DKR. We're doing a good job of making all of our content addressable so far. Our biggest challenge, not surprisingly, is integrating this content properly. We're going to approach this several ways. I'll be generating an IBIS map of the e-mail discussion using Jack's Nexist, which is coming along quite nicely.    3A  (027)

Our main approach is to capture the group's ontology, or conceptual understanding. Howard, our ontology master, has been developing ontologies of our papers and discussion using Protege. However, Protege currently doesn't have great facilities for collective ontology development.    3B  (028)

Howard proposed a methodology for using a Wiki -- which I recently installed for experimentation -- and Protege in concert to generate our ontologies. Essentially, Howard would develop ontologies using Protege, export those ontologies in native format and in XHTML, make the Protege ontologies available for people who want to play with them, and put the XHTML documents on the Wiki. Using the Wiki, everyone in our community could edit the ontology. Howard would act as sort of a human validation mechanism, pointing out ways to keep the ontology consistent, and giving us hints on improving the ontology. At some point, Howard, would sync the content on the Wiki with ontologies he maintains on Protege. And the cycle would repeat itself.    3C  (029)

We also need to experiment with better ways of representing the ontology. I explained that Howard's ontology for Lee's NODAL is a great way of understanding NODAL. However, exported as XHTML, its structure may not be as evident. Ideally, we'd like a graphical and addressable representation of these ontologies.    3D  (030)

A Presentation on NODAL   4  (031)

Lee gave a presentation on NODAL. His slides will be available on our web site, but here's a very brief summary of the talk. Lee gave us his pitch presentation. We weren't necessarily the target audience for this particular presentation, but Lee wanted to show the elements that have elicited positive feedback from others.    4A  (032)

He started by describing the big picture. The problems we are trying to solve as a group are to decrease the barriers for ubiquitous collaboration. NODAL is a candidate design for an OHS "filesystem."    4B  (033)

Lee then described NODAL in more detail. He talked about the constructs for NODAL data models, and explained that a node in NODAL is the basic unit of granularity for metadata, such as a unique ID, its version and transaction history, and a permission record.    4C  (034)

Throughout the talk, there were a few questions. Murray asked about Lee's term "node types", and suggested that "collections" was a better term. I said that this seemed to be a relic from Groves terminology, and that renaming it might help clarify its meaning.    4D  (035)

Howard asked about whether "root node" implied tree structures. Lee said no. This was essentially the same point made at a previous meeting, where we suggested that "origin" may be a better term.    4E  (036)

Doug stated that he liked NODAL, but wanted to make sure that Lee make explicit NODAL's place in the larger system. He asked Lee to rename the title of his presentation from "A System for Ubiquitous Collaboration" to "A Filesystem for Ubiquitous Collaboration," a request with which Lee complied.    4F  (037)

Finally, Lee explained his current situation. He's moving in a few weeks, and will be working for SRI on a part-time, hourly basis. He's trying to get funding to work on NODAL, but is planning on working intensively on NODAL during his free time at least through the end of September, funding or not. As he said, he's moving, but he's not going away.    4G  (038)