XKCD — Standards

Randall Munroe has a comic for every situation, and this is certainly the case when it comes to the creation of standards. For the better part of a week I've been looking at the structure of digital textbooks and how they can be made a little more flexible, a little more portable, and a whole heck of a lot more accessible. At the moment the most common formats to find a digital textbook are straight HTML, PDF, or one of the common ePub formats. These have their own lists of pros and cons but one common thread that I've found over the years is that none of these solutions are particularly good when you're trying to make something to be used in a classroom. Solutions are often created in an attempt to appease the students with very little consideration made for the teachers. This is something that I would really like to change in the near future.

One of the more complex problems that I've been trying to address over the years is how to create decent teacher's books that incorporate all of the resources and metadata that might be required in a classroom. I've developed APIs that return this data in a consistent format for textbooks developed within the company, but third-party resources, such as those put out by Pearson or National Geographic, have long been a problem due to the lack of flexibility in the data formats they employ. Every resource is either locked away in a proprietary format, or provided as a PDF that then needs to be restructured in order to make it useful within a modern digital textbook system. What I would really like to do is devise a decent data format that could be published as an open resource that addresses the fundamental problems of working with teacher's books and see it adopted and built-upon within educational circles.

Scribbled in some notebooks next to me are a number of notes that, when combined, form a semi-rational structure that can be presented as a template for how a teacher's book might work. It takes into account things like sequencing, alternate lesson plans, audio resource metadata, video embedding, resource attributions, and more. When I apply this format to some of the textbooks used at the day job, I can generate a portable file that sits at between 18 and 65 megabytes in size, which contains the entire text and structure of the book, plus any additional resources required such as audio files, video files, images, and printable homework activities. All in all, the specification looks good. It hasn't yet been properly tested, though.

Over the next couple of weeks I'll be transforming some learning resources from their source PDF files to "proper" Markdown-formatted teachers books using the JSON data structure I'm hoping to propose as a standard. There will be a couple of books from Pearson and I'll probably grab something from a Canadian publisher as well just to see how complete the idea is. If I am able to fit just about any kind of textbook into the definition structure without trying to find workarounds to an incomplete design, then I'll know the idea has merit and can be shared. If the structure creates friction between the content and the smooth operation of a classroom lesson, then I'll know there is still much work to be done. That said, I would love to see this structure become an official standard and get used by schools and educational resource publishers around the world.