Adapt Learning Community: Databasing of Content and Real-time Delivery of Courses?

I'm just learning about the Adapt Framework and excited to see how the Adapt Authoring tool will pull it all together. (Any update on the timeline for the authoring tool. I read somewhere 2Q 2014 but the official site just says 2014) Anyway, I think you have a great vision and I'm anxious to see it evolve and grow. I have a couple of questions.

1. Based on my understanding components are built based on json statements. Those json statements essentially bring together the component and the learning content. Do you foresee that the learning content (e.g. titles, instructional text, paths to media) will be stored in a database and generated dynamically real-time to create the learning experience or will the json files be text-based files on the server? Certainly there are pros and cons to both options. If you are going with the text-based file approach do you see any significant obstacles if one wanted to leverage a database approach instead?

2. Related to the prior question, if a company considering the adapt framework required some additional principles of Centralize Templates and Real-time delivery of content what would you recommend? How easily could the framework be adapted to accomplish these additional principles? To expand the principles, we would want centralized templates that are used for all content that is generated dynamically real-time. That way, if we need to make a component update we make it in one location and all content would be updated. This would eliminate the need to hunt down published content in multiple locations. The real-time delivery principle would allow us quicker speed to market because we could eliminate the launch process and make quicker maintenance updates.

Thoughts?

James

Hi James,

Although the Authoring Tool does store content in a database during content creation, the published content is contained in a flat json file. I'm not aware of any plans to change this at the moment, and it has not been raised prior to this as a feature that we'd aim to support. Having said that, I'd be interested to find out what the rest of the community thinks about such a feature.

Your second question regarding centralized templates is a little trickier. A lot of work has been invested in the framework to minimize the space footprint of a published course, meaning that the code that drives components and other features is culled, minified and obfuscated as part of the publishing process. This means that - to a certain degree - the javascripts and templates that are used in a published course are unique to that course. This means that Adapt content is very streamlined and can be delivered very quickly over the network to clients. I believe a design change to support centralized templates would negatively impact this performance. Again, though, I think it's a very interesting proposal, and I'm curious to see what the other community members think.

While I don't think the core team will attempt to support these features in the near future, I suspect it might be feasible to fork the adapt-cli and perhaps the framework itself and implement those changes - I'm keen to hear from Daryl what he thinks about the framework's potential to deliver content in this way.

Thanks,

Dennis

Show parent | Reply

Hi James,

thanks for your feedback and for raising these questions. And thanks, Dennis, for your response. I wanted to add the following:

Do you foresee that the learning content (e.g. titles, instructional text, paths to media) will be stored in a database and generated dynamically real-time to create the learning experience or will the json files be text-based files on the server?

As Dennis said, the authoring tool will store the data in a database. However, the workflow we are implementing focuses on creating a published / exported module at the end of the process.

We're specifically focusing on the usability of course creation / editing and functionality to support this particular type of workflow (e.g. with the requirements to collaborate, review, preview etc), so the idea is that we make the creation process as easy as possible.

The other side of this coin is the delivery process. For now, this is not within the scope of our project given the size of the task we have set ourselves. Our thoughts at this time are that LMSs / VLEs and even regular Web CMSs (or LCMSs such as Xyleme) are more directly suited to this task at this time.

However, we are interested in exploring this once we have achieved our initial goal and vision. I believe the world of delivering learning is changing and with initiatives such as xAPI / Tin Can and various social learning platforms there may well be an option to branch out into handling delivery in the long term (especially if anyone were to express an interest to collaborate with us on this particular aspect).

I hope this helps. Please don't hesitate to get in touch if you have any further questions.

Thanks,
Sven

Show parent | Reply

Thanks Dennis and Sven for your detailed and quick replies.

I can certainly understand the reasons for the direction you are headed. I often consider the pros and cons of a packaged course verses dynamically created courses that are delivered from the authoring tool/LCMS. To me packaging a course and placing it on a LMS is like separating a child from its mother. Once on the LMS all layers of the course (skin, navigation, features, and content) are no longer easily maintained and so many benefits and capabilities are lost. For example, once packaged the benefits of social authoring and collaboration are lost. You also lose the ability to receive direct real-time learner feedback that can be acted upon immediately. You also lose the benefits of single source authoring where multiple outputs can be updated from a single content source.

The role of the LMS is also diminishing or changing. We are now in an environment where small bite content (perhaps just 1 "article" instead of 3) may be all that is needed to meet the needs and demands of the learner and the ability to dynamically serve up only what is needed at the time of need is very powerful. I believe that only through the dynamic generation of content will we be able to meet the mobile on-demand needs of our users. Also, in some cases, content now needs to be accessible outside of the LMS and in locations where it will better meet the needs of it user. Keeping that content on the LCMS allows for it be in many locations but still easily maintained through one content source and one template source.

My last thought is an idea pertaining to the blending of eLearning and performance support. Consider for a moment content that is not only responsive to the device but also responsive to the intent or need of the user. (I don't know if this is a new concept but if so perhaps I should coin the term "responsive intent"). Let me explain. We know that formal content is different from performance support or reference/help content but that there is often overlap between the two. Formal content contains stuff that I consider fluff (e.g. quizes, objective page, transitions, etc..) that would never be desired when accessing content in a performance support setting. Well suppose for a minute that the user wanted both the option of formal learning and performance support from the same content. To achieve such the user would shrink the responsive page to a small performance support view where they would only be presented with a tree hierarchy of the content structure as well as search, index, and topic options from which they would be able to access performance support topics (most likely articles in Adapt). And, when in this view, the nature of the content would also change from formal content to performance support oriented content. This performance view would have all the "fluff" removed containing only critical performance support content. So what do you think? Given the demands of performance support content with high maintenance cycles this is one reason why I think real-time generation of content outside of the LMS would be needed.

Anyway, thanks for letting me go on with this. Keep pressing forward. I look forward to seeing the authoring tool.

Take care,

James

Show parent | Reply

Hi James,

thanks for elaborating - and great to have this discussion.

I can see you have given this a lot of thought and there are a lot of points, which resonate. I started with writing a much longer reply but came back to one question most of the time. I hope you don't mind me keeping it short and asking that first. (I'm hoping we can keep this discussion going for a bit :-)

The functionality you describe, e.g:

not separating the content from the generation/editing tool (child/mother) for reasons:
- easy maintenance
- social editing / collaboration
- acting immediately on real-time learner feedback
content needing to be accessible outside of the LMS and in locations where it will better meet the needs of it user

sounds very close to regular Web CMS functionality. The main question, I wanted to ask is where does a regular CMS (e.g. Drupal, WordPress) fall down in terms of your requirements?

To half-answer my own question, I imaging this might be in two to three areas:

Tracking
Generating meaningful, easily interpretable tracking data on a changing course. The emerging xAPI standard may help with this.
Learning-related interactions
Do regular CMSs have sufficient, suitable, learning-related plug-ins
Publishing to other systems
Is it possible to publish (updates) to multiple delivery locations

As mentioned, I am keeping this reply short and look forward to your thoughts!

Sven

Show parent | Reply

Sven,

I think you hit the nail on the head with the areas you mentioned. I haven't explored CMS system such as Drupal or WordPress primarily because of their lack of focus in the learning space. Based on this discussion, perhaps I need to take a more in-depth look. Sorry, I don't have more to say but I'll give it some more thought.

Show parent | Reply

Hey James,

I've been meaning to get back to you on this and sorry if this gets a little too much of a brain dump.

I really like the comment :

"To me packaging a course and placing it on a LMS is like separating a child from its mother."

What I like about it is that it got me thinking about the connection between the courses and the backend. We rely so much on systems that give us ways to path out learning but then separate the learning out into SCORM packages.

Then on the other hand we have Treehouse and Code Academy who have an integrated system.

Both of these systems work for their intended usage: Managing users learning and giving a free rein of learning.

I like to think about your comment in the opposite direction - imagine a world (I guess it needs to be a world) where children were given the best mother that suits the childs needs. So the child gets everything they need from the parent without much thought or change in behaviour. Just an idea...but something worth thinking about.

In terms of Adapt - the child has been built with the learner in mind, with some pretty cool technology. I guess technology like the tin-can API can help pave the way for a more dynamic mother.

I'm not even sure this makes sense but thanks for the thought provoking comment - Keep them coming.

Thanks,

Daryl

Show parent | Reply

I think that Adapt, as an Open Source tool, is a step in the right direction to achieve a new level of dynamic learner platform.

Especially with the creation of the Authoring Tool, I can see that the dynamic creation of an Adapt Course in a learning platform being a real advantage and the ability to edit Adapt course content in real time is a viable solution.

The Adapt Framework is a solid core to start building dynamic, integrated, multi-device learning.

Chris

Show parent | Reply

the authoring tool will store the data in a database. However, the workflow we are implementing focuses on creating a published / exported module at the end of the process.

javascript obfuscator tool online

Show parent | Reply