This slipped past me, but way back in December, I attended the CNI fall meeting in Washington, DC. Knowing that it never works to fly on a December morning to make an afternoon meeting, I cleverly took the train the day before, only to endure five and a half hours of delay as Amtrak worked to resolve engine problems. But I did get there for the kickoff, when half the attendees did not (fog socked in the whole eastern seaboard).
The opening keynote went straight to a subject that was repeated throughout the conference: MOOCs (see Lynn’s post that covered other MOOC sessions). Lynch addressed a number of issues with MOOCs that I haven’t heard about elsewhere:
- A lot of schools are setting up MOOCs without a clear definition of success for the program
- MOOCs will generate a lot of student data, and it is not clear who owns it and what rights they have to use it, and what access students will have to their own records. It may be very valuable content for data mining.
- MOOCs have the potential to cause a lot of disruption in some doctoral programs, because we fund a lot of those doctoral students by having them teach the sorts of classes that are becoming candidates for MOOCs. Humanities doctorates in particular may be affected.
- Our ability to teach 100,000 students at a time is probably outpacing our ability to grade their work
Other points: many schools are including alumni in library licensing agreements. We want to do that here, but we have to wait for some campus-wide systems to come online that will let us identify alumni accounts.
Encouraging court judgments in favor of fair use and rights of users to access copyrighted material is counterbalanced by electronic licensing terms that increasingly give publishers more than their traditional first-sale rights. As Lynch put it, “Will your children be able to inherit your e-books?”
Cliff’s opening keynote can be seen at http://www.youtube.com/watch?v=Fvys5VZrjsI
Two other sessions I’ll comment on: in “Debunking Myths and Establishing Guidelines for the ETD Lifecycle,” four speakers addressed issues in long term curation and preservation of electronic theses and dissertations. Of particular interest was Gail McMillan of Virginia Tech, who presented “Do Open Access ETDs Effect Publishing Opportunities in the Sciences? Findings from the 2012 Survey of Academic Journal Editors“. This follows a similar survey of editors in the humanities and social sciences in 2011. This presentation gives substantive support to the argument that publishing an open access ETD does not greatly affect an author’s ability to get books or articles based on the ETD published. The slides are well worth a look.
Because of our investment in cloud technologies, I was particularly interested in “The Truth is Out There: Preservation and the Cloud.” David Rosenthal, Chief Scientist of the LOCKSS program, basically asked the question, “Can you afford to digitally preserve content for 100 years on the Amazon cloud?” And ultimately answered it, “No.” For long-term storage, you can rent space (think of a journal subscription that you have to commit to for 100 years); subsidized with ads (like Google Mail, but generally a non-starter in academic environments), or endowed, with an up-front payment considered to be enough to store the content “forever”. Because the cost of hard drive storage has been dropping by half every two years for several decades, preservation endowments have gained some traction.
But Rosenthal sees two problems: first, the cost of hard drive storage will level out. Disk prices will continue to decline, but not so quickly. Second, while commercial cloud vendors periodically drop their prices, they don’t do so as fast as drive prices go down and customers increasingly fund their profit margin. This makes the monthly payments to Amazon (and its successors) over the next century increasingly hard to support.
Rosenthal’s proposed alternative is low-cost local storage, specifically building storage “pods” from designs published by online backup company Backblaze. They provide plans for constructing units capable of storing 135TB, expected to last for four years, that cost about what Amazon would charge for one month’s rent on 135TB of space. There’s obviously a lot more to factor in, but in the end it’s a compelling case. Long term storage won’t make economic sense in the commercial cloud.