IST-Data Services Blog

At Educause’s “Advanced CAMP” (Campus Architecture and Middleware Planning) this week the sharpest lesson I learned is how little is solved in the problem space addressed by this year’s CAMP meeting: “Registering, Discovering, and Using Distributed Services in Academia.” Bob Morgan from Univ. of Washington put this in context right up front: though the organizers of the conference like to call the group “Advanced CAMP” Bob thought a more fitting name would be “Advance CAMP” — a group of IT architects seeking solutions at the edge of what is known and/or possible in EDU-space. I presented on what we’ve been hearing at Project Bamboo workshops (my presentation will be coming to the conference wiki soon).

For me the most interesting topic was service discovery, and the take-away from Thursday’s break-out discussion on that topic is that it’s more natural for central-IT providers to think about mechanics of finding a URI for and method of engagement with a known-to-exist service in a defined domain than the broader discovery question, “I want to do X, is there a service out there that’ll do it for me so I don’t have to code it up myself?” (That last is closer to the question a scholar of the type engaged with Project Bamboo might ask, and so is the question in which I’m most immediately interested.) Despite the scattered focus there were interesting suggestions in this space. A few of them: visualizations of services mapped from RDF graphs describing service relationships to each other (Loretta Auvil of SEASR showed some examples); a “bring back gopher” suggestion (with tongue only gently implanted in cheek, from Ken Klingenstein of U. Colorado @ Boulder), with a scholarly-domain flavored partitioning to replace the geographic / institutional partitions that characterized gopher back in the day; the idea (from Mark Morgan of University of Virginia) that assertion of identity might provide an appropriate context for discovery in a bounded space (leading me to think that maybe that’s one way PB can begin to understand the “Facebook for Scholars” types of suggestions we’re hearing at our workshops). The question of how to provide incentive to digital resource providers (be they providers of source material or of applications that find, analyze, manipulate, organize, or annotate it) to attach useful metadata to their contributions to the digital commons remains a cultural question that is central to success of discovery across ‘the usual borders’ but can’t be solved wholly by technologists.

Other thoughts worth sharing:

• “Not Invented Here” (NIH) syndrome appeared to have left the building for attendees at Advanced CAMP: everyone who presented expressed strong interest in not reinventing wheels, in building off of common components and services.
• Users learn semantics more readily than syntax, e.g., as Mark Morgan observed based on his experience with the Genesis II grid-computing project, “double click” has easily- and broadly-understood implications in a world modeled as a file system.
• Tracking digital artifacts as they replicate and morph across the internet is a hard problem that universities — with interest in provenance and chains of credit — will have a great deal of incentive to solve, and it’s probably worth looking at Disney and WGBH’s DAM initiative to see how this problem is being addressed in other spaces (Robert Clark of Duke and Michael Pelikan of Penn State Univ. were principal contributors to these discussions).
• Cross-border licensing and privacy challenges are something to keep a very close eye on: with European laws on the books or in the works that may disallow certain PII (personally identifiable information) to be stored on U.S. servers due to lax privacy laws, everything from federated service-composition to cross-border participation on wikis may be affected in ways that require serious modification and mitigation to projects in which Berkeley participates (thanks to Ken Klingerman of U. of WA and Mary Beth Lavagnino of Indiana University for presenting on topics of privacy and policy).

On the “juicy news” front, we learned from Jens Haeusser of University of British Columbia that USC has joined the Kuali Student Systems effort at the founder level. And Nigel Watling of Microsoft provided some guarded pre-announcements about a plot brewing in Redmond to offer outsourced ESB (cf. biztalk.net) … your services (and the data they carry) linked together on a Microsoft-run bus … and you thought Google Apps was scary! Nigel was a nice fellow, but there was plenty of nervous laughter when he told us that “Microsoft is absolutely committed to open standards” and that the company now understands that if its products are not based on open standards they’re “not relevant.” In response to skepticism about the advisability of putting confidential information out in the cloud, Nigel was less than convincing — assertions like “it’s out in the cloud already … people are now more tolerant of personal information out in the cloud” weren’t the kind of responses likely to satisfy this audience.

This entry was posted on Saturday, June 21st, 2008 at 1:18 pm and is filed under Data Services, News. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.