Future of interoperability: food for thought…
I spent Monday and Tuesday in Bolton, for a meeting of the CEN/ISSS Workshop on Learning Technologies and a CETIS day on the Future of Interoperability Standards.
I arrived late on the first day, but it seemed to proceed like a regular CEN workshop meeting: I have doubts about the relevancy of some of the work, but things do progress and have some traction.
The second day attracted more people than expected: the good news is that quite a few people seem to care about the future of interoperability standards. The bad news is that the day was organized because of the feeling of dissatisfaction with how standardization of learning technologies is taking place. Seems like quite a few people share that feeling then…
Of course, the standardization process is far from optimal: it is slow, doesn’t always lead to results, or at least not always to results that matter to folks outside of these meetings. On the other hand, I am an optimist: in ARIADNE, we now have more than 400.000 learning objects with LOM metadata: that is not even near where we want to be, but it is much better than 5 years ago when nobody had more than 10.000!
In the introductory session, I mentioned some of my concerns:
- fragmentation: too few people work on too many different things in too many different organisations – there may well be more organisations than people in this area! I think that one of the reasons is that involvement in the development of a standard can be listed on a cv, whereas the work behind the scenes to develop infrastructure is often … rather hidden behind the scenes. A standard should really not be a goal, but more a last resort when we have a problem to make things work together.
- consensus: This is inherently a difficult type of process: it is a lot like “herding the cats”. In other areas where consensus is the norm, things move forward in a rather messy way too. I kept thinking throughout the day about the United Nations or the Global Warming Summit. Maybe the standards process is a bit like democracy in that respect: more kind of the least bad system… At least, the future of humanity or our planet doesn’t depend on what we do. (Then again, I happen to think that learning is important…)
- Don’t stop too early: Many of the “standards people” move on to something else when a standard is finished. (Incidentally, around 80% of the participants said that they were paid to develop standards. That feels odd to me: it seems to position standards as a goal rather than as a means…) I think that we should pay more attention to what happens after a standard is finished: LOM was finished more than 6 years ago, but we are still very much working with all kinds of communities on the deployment of application profiles, back-end infrastructures and front-end tools. And making progress – slower than expected maybe, but real progress: more on that later.
There was a lot of interesting discussion over the day – you can actually follow it quite nicely on twitter. I’m not sure that the discussion helped us to make a lot of progress. I certainly had a strong “them is us” kind of feeling: if we want to get better at developing standards, we need to agree on how we will make that happen… And we need to move beyond the “we should all work better together” stage…
In any case, there is some Good Stuff in the position papers on the web site and I’d love to hear your comments and feedback: what are your ideas on how to improve the standardization of learning technologies?
Erik Duval's Weblog 
So other than agreement that everyone was dissatisfied with the standards creation process (though I didn’t see many people who are on W3C or IETF working groups), no one really said what a good standard is? Seems to me the Web is filled with good standards, but to mention a few: HTTP, RSS/ATOM (or is this a standard?), XML, etc. Perhaps the dissatisfaction with standards is one for the lack of success that domain standards have had e.g. we are trying to create ones that already exist? Rather than try and further our own personal standard cause, shouldn’t we be contributing to standards that are successful on a WebScale (not just an academic domain scale?). Just did a “how to” search on youtube and got back 28,300,000 results (let alone the number of hits). Wouldn’t it be better to adopt the tech standards that these successful WebScale platforms already offer, e.g. adapt LOs to the Content Model YouTube provides? Perhaps this has been done, haven’t followed LOs for about 3 years.
Hi David – I was there, and I’m in W3C and OWF!
The feeling of dissatisfaction may say more about the ambitions people had and promises made than about the actual impact of standards, which have been uneven but probably no worse in terms of their “hit rate” than W3C and IETF (possibly better actually!)
The big difference is the LET domain has far fewer resources (expertise as well as money) to spend on standardisation, so we care perhaps more about our effectiveness, and so wasted effort is more keenly felt.
I completely agree with your remark about standards. I am currently writing my thesis about the semantic web, and it is uncanny how many standards and libraries exist in a new field. “The great thing about standards is that there are so many to choose from.”
It is especially ridiculous when you need to know all of the standards to use the technology in a meaningful way. Discussing on IRC always turns into a “Wait, what syntax are you using?”, since there are more than 5 different notation standards. Writing a new standard should always go hand in hand with deprecating the old one, and merging the people still using the old standard to the new one. In a way, this is a social construct; software is only as good as the number of people using/developing/supporting it.
Hi David – I was there too and I do work in W3C and other places that relate to it.
I have to agree with Erik’s comment about there being too few people doing work in too many organisations. I addressed this in my paper on the even web site too. In fact it was difficut to find a focus for the paper because of that.
I would argue that the way to move forwards from that fragmentation is more joined-up work – across the bodies. projects making disparate work interoperate. But that’s only half an argument because I think given the drivers of work (not always ones everyone could be proud of) its very difficult to achieve.
I’m not at all dissatisfied by where we are – I just think its early in the process – we are at an immature stage and are learning IMHO. -andy heath