I almost missed this post by Scott Leslie expressing his mixed feelings about the standards talk coming out of Alt-i-Labs. Lucky for me, Stephen picked it up in OLDaily. Scott writes:
Part of me really wants some of these developments to come true, to deliver the promised 'plug and play' elearning environments described herein, and in my rational moments I know that 10 years really isn't that long for a field like this to coalesce around an open set of interoperability specs. And yet it would be hard to fault a newcomer looking at these presentations for wondering if this represents what is still to be done, how anyone manages to develop quality online learning experiences now (and how many PhDs will be required to operate the CMS of the future)?
The thing is, the newcomer would be correct to wonder. While a good teacher can cultivate quality learning experiences (online or otherwise) with a rock and a stick, the truth is that today's online learning environments don't really help them much. The majority of online courses basically use the LMS as a discussion board (a.k.a. "the rock") and a file sharing system (a.k.a. "the stick"). There are several reasons why interoperability standards should help this situation a great deal.I've made the long tail argument many times before, but to recap, we need many specialized learning applications to teach specific skills within specific disciplines using specific teaching methods. It's easy to trivialize this argument as talking about a bunch of edge cases. But that would be a mistake. The whole idea behind the "long tail" is that the total number of edge cases outnumbers the total number of core cases. So we are possibly ignoring the majority of needs when we ignore the long tail. And to give a sense of just how long the tail is, consider that most LMS's still don't have blogs or wikis built in. These mainstays of online communication everywhere except the LMS are apparently too far out on the tail to merit commitment of the highly limited development resources by the LMS creators (both Open and proprietary).
But the long tail argument probably doesn't even go far enough. I would argue that even many of the "core" LMS applications, properly designed, really should be situated software, meaning that the applications should be tailored to the needs of the individuals rather than created one-size-fits-all. For example, most online gradebooks suck because they have to account for many, many different possible grading schemes. Which means they do none of them simply/well. Most faculty are happier just using spreadsheets, even though it means they have to manually enter their data. Spreadsheets are situated software. They are designed for a particular situation, developed inexpensively, and thrown away when they are no longer useful. One could make the same argument about discussion board interfaces.
In fact, one did. Here is the paper I presented this week at Ed-Media 2006. Among other things, I argue that interface details of a discussion forum matter and are differentially suited for different teaching purposes. Can you design a discussion board with different display options? Sure. But that general strategy can only be stretched so far. (Witness the gradebook.) The only way to deal with this problem is to encourage the proliferation of plug-and-play teaching apps. To do that, you first need to define plug-and-play.
As for Scott's other concern, regarding complexity, I don't understand TCP/IP, HTTP, POP3, SMTP, IMAP, or MIME at all, and I understand precious little HTML, and yet I'm able to use the web and my email just fine. If these standards are properly developed and implemented (always big "ifs" when it comes to standards), they should be transparent to the users and relatively simple for the developers. You don't need to understand the guts of your web server to write a good web page.