Why we're using Tin Can
Core to the Coracle platform is our Learning Line: a powerful social media style activity stream that records users' learning activity, including non-course based learning, over time.
This article attempts to demystify our approach!
SCORM allows courses to be built up from smaller modules using a Learning Management System (LMS) that allows progress tracking and reporting.
Unfortunately SCORM is a heavyweight and inflexible standard. It's relatively hard to make course material 'SCORM compliant' and it becomes inflexible when it is. Learning that takes place outside of courses (such as watching videos, attending seminars, private reading etc) cannot be recorded in SCORM.
This is where Tin Can comes in. The Tin Can API (or the 'Experience API' if you're the American Military) is being developed by the keepers of SCORM. This, in essence, is ADL, the educational arm of the USA's Department of Defense. The friendlier sounding, community focussed branch of this is Rustici Software, who were charged by ADL to replace SCORM, and thus created the Tin Can API.
Tin Can and Coracle
As soon as we heard about the Tin Can API we realised it gave us an epic reinforcement of what we wanted to do anyway: to be able to give people control over the records of their life-long learning experiences. We now had a standard with which to do it and the belief that in the future we will be able to easier share the courses and the data that we create.
Tin Can follows the social model. Interactions with a Tin Can LRS ('Learning Record Store') only require a short description of what the user or an agent acting on his behalf wanted to record at a particular time. This is in the form of a slightly mechanical but semantically correct short English sentence: "I DID THIS". No reference is made to what you did or how you did it, so Tin Can is opened to any form of learning the world can offer.
The great thing about this is, and the one we're really looking forward to, is the challenge of returning the data back to users over a period of time in vibrant and useful ways. For instance, we could show a user that he achieves better results when reading books than watching videos. We could show them that the population in general gets better results when playing games than reading books. The possibilities are really interesting.
Under the Hood
One of the first open learning features we implemented was our 'I Learned This!' bookmarklet. Here's what's going on to make that possible:
The developers of Tin Can have researched their choice of vehicles well. As well as JSON, the Tin Can specification makes use of OAuth 1.0 (a very good choice, since one of the key developers of OAuth 1 whole-heartedly retracted his support for OAuth 2 last year). Internationalisation also gets a mention, as does REST, GUIDS and the ISO8601 date format, all of which gives the feeling that the developers are clued up, and have made good, reasoned choices when considering their options. In other words, they've done what I would do.
The core of Tin Can is serialising the 'I DID THIS' concept. The 'I' being a learner of some kind, 'DID' being a well-defined verb describing how the user interacted with 'THIS', a learning activity; a traditional e-learning course, or a web page, a game, a simulation, etc.
Outside of this, there are efforts to describe the cast, who are referred to as Agents. There's a fair amount of talk about this, as the Agent who made the statement is not necessarily the subject of it. For example, a tutor may make a statement on behalf of their students. There's some structure to the way Agents are defined, but it's deliberately left fairly loose. Agents are uniquely identified by email addresses.
Learning activities also get a mention. They are called Activity Profiles, and are used to transmit information about courses, games, etc.
What features would you most like to see us implement? Please feel free to let us know...