Technology Leaders Balk at Technical Guideline in Federal Grant Program

In response to Bernard’s comment, the IMS Board of Directors provided a pledge letter (guarantee of interoperability) to Secretary Duncan and other officials at the Department of Education in November of 2009 in anticipation of the AGI OER initiative. The letter contains very explicit standards that could have easily been included in the solicitation along with SCORM. SCORM even can be contained inside one of the standards (Common Cartridge) – SCORM and Common Cartridge are both based on something called IMS Content Packaging – so there is a clear route to compatibility.

I’ve uploaded that pledge letter to the IMS forums. Bidders may be able to make use of the pledge and select from the alternative standards. But, so far there has been no response from Dept of Labor or Dept of Ed on what would be acceptable. The pledge is available for download here:
http://www.imsglobal.org/community/forum/messageview.cfm?catid=58&threadid=595#1398

The open content idea behind the Federal Grant is a nice one: Create, Share, Reuse, Remix. But SCORM is not made for that and neither is IMS CC! Thats where I disagree with Rob Abel. Both standards are packaging formats. They are basically ZIP-Files containing HTML Files plus a manifest (index file) telling the Learning Management System in which order those HTML-Files have to be displayed. But both formats do not cover the “Remix” aspect of open content. Hardly any teacher is using teaching material from others “as is”. They will want at least their own design (with their school’s logo) and use their own order when presenting the content to their students. And both formats do not support that! So from my perspective they are both outdated (but, and thats where I agree with Rob Abel, SCORM is WAY MORE outdated than the very modern and more flexible IMS CC format!).

In my opinion the only way to go in the future is to create platform independent content. A very interesting project is e.g. Allyn Radfords idea for a structured content editor for open educational resources: http://wikieducator.org/Funding_proposals/Structured_content_editor_for_OERs (Disclaimer: you will also find our universities letter of support on the project website because we believe in this approach and want to support it).

Content today should be created in structured, open, non-HTML-format like XML. This way you can choose to transform your content into the format of your needs, this can be SCORM or IMS CC but this can also be ePub, if you want to display your content on an iPad or an HTML5-based format for Mobile Devices etc. (Plus other formats like PDF or .docx if print formats are important). At the University of Zurich we use last years IMS Learning Impact Silver Medal winner eLML (eLesson Markup Language, http://www.elml.org) for to create e-learning content. But there are several other similar initiatives like CNXML or OUXML and they are equally superior to SCORM or IMS CC in my opinion.

Why? Let me make an example: At our university (the University of Zurich) we have a learning object repository with nearly 700 lessons from about 50 major e-learning projects. We are talking about around 20’000 HTML pages of content. Two months ago our University introduced a new corporate design: new logos, new webdesign etc. If we would have created this content using either SCORM or IMS CC then we would practically have to change each one of the 20000 HTML pages “by hand” which would be the end for many of those projects. Using XML we can now create one new design template and with one mouse click create 20’000 pages in the new design (at least in theory, in reality it takes at least two clicks :-) Or we can transform the content into the ePub format and publish it on iPads. And the question SCORM or IMS is irrelevant for us because using eLML we can produce both packaging formats.

And when it comes to “remixing” content, structured XML is even more powerful than HTML because – as e.g. CNXML proved – you can create tools that offer teachers the possibility to freely remix content (this can be a whole page or just a paragraph or image). Thats something both SCORM and IMS CC doesnt offer because there you will always be stuck with single HTML-pages in the end (unless you agree that one HTML page is the smallest “remixable” object, but I would not want that limitation as a teacher).

So my advice: Use 1% of the budget to create an intelligent structured content editor as the one in Allyn Radfords proposal (based on the experiences made in projects like eLML, CNXML or OUXML) and then the applicants of the other 99% of the budget will have to use that editor to create their content :-) And to IMS: Invest resources into creating a structured language like CNXML or eLML and then we can all start building tools to “create, share, reuse and remix” based on one common standards. That would be nice!

In the 25 years I’ve been watching computing-related standards, the of the huge number that were developed by the government, something like 1% were ever widely adopted by those who weren’t forced to. Even IP (probably the second-most-widely adopted) was developed independently, by researchers supported by DARPA.

Though, to give props: the most successful government-developed standard, COBOL (still in use more than you think, 60 years later), was developed under the leadership of Adm. Grace Hopper, so at least there’s one good precedent for military development.

Joel – Cool comments – I also feel that we should spend 1% of the money building a structured content editor. And when I look at Allyn’s page, he is right on the money as to what the single most important use cases we need to solve. Where we diverge a bit is that you want to immediately invent a format that solves all the problems. And of course CNXML and ELML are interesting candidates – they sadly have not had enough investment yet to have assured broad appeal. I (not surprisingly) think that the first tool should be an IMS CC tool that can read and write IMS CC. Then lets get that tool in people’s hands and learn from it – at least they can use it with their current LMS systems. Then we go through iterative development taking inspiration from LMS’s, CNXML, and ELML to co-engineer a truly good solution that makes it so we can make one artifact and generate everything from ePub to PDF to flat HTML. Actually what we need is a simple, open source version of SoftChalk.

But the tool needs to start simple and be useful immediately in the current marketplace. The tool cannot be a research voyage of discovery where we are endlessly brainstorming about the most awesome possible way to do things.

If we had a little bit more money, I would actually put funding into deeper investment into desktop tools for CNXML, ELML, *and* IMS CC – and then I would fund those folks if and only if they worked together as a team. It is OK to work on more than one solution at a time if you have the money. Let us try a few things and see what clicks with the teachers and students. There is always real danger when technologists guess the use cases and then build the tools for the teachers without ever checking with those teachers.

I think the federal government should be congratulated at least for suggesting that these materials be developed against some kind of standard. None of the standards that have been suggested here have taken off in higher ed. At least SCORM did become a defacto standard in other sectors. Higher ed is not likely to embrace any of these standards if they are not encouraged to–perhaps by grants like this.