A few quick items of interest form the web this week. Two offer a perspective of the process of making standards (looking at OAI-PMH); another is an interview with Brian Lamb reviewing the history of Learning Object Repositories.
Talking to DC [Washington] (Adam Bosworth, Adam Bosworth’s Weblog)
In a post based on his experiences with standards development, Adam outlines seven guidelines for good standards development
- Keep the standard as simple and stupid as possible.
- The data being exchanged should be human readable and easy to understand.
- Standards work best when they are focused.
- Standards should have precise encodings.
- Always have real implementations that are actually being used as part of design of any standard.
- Put in hysteresis for the unexpected.
- Make the spec itself free, public on the web, and include lots of simple examples on the web site.
Making Standards that Work (Dorothea Salo, The Book of Trogool)
Relfecting on Adam’s post, Dorothea relates some of those principles to her own experience and view of standards development in particular commenting on OAI-PMH.
OAI-PMH is an interesting example because it’s so widely used and, as Dorothea says, so simple. When it works, it works well (even if we might now like to change some of it to be more web friendly). However, metadata sharing works best in defined communities. When OAI-PMH doesn’t work, it’s a mess as frequently it’s the data harvesters who notice but who are dependent on the data providers (and potentially also their technical support) to change anything. Interestingly the OAI-PMH Static repository specification pushed some of the emphasis back onto the data provider – as their base information in xml had to be valid xml before it would be mediated by a gateway (but SRs are a whole other story with lots of potential but their own problems.)
The Sordid History of Learning Object Repositories or, a chat with Brian Lamb (Jim Groom, bavatuesdays)
One of the interesting things about the UKOER programme is how much freedom projects have to choose how they are going to store, describe, manage, and share their resources. They are using a wide variety of approaches, which include repositories, content management systems, and web pages of rss feeds. They’re also using a wide variety of ways to descibe stuff. All of the approaches though are some way from the sort of educational world which the original learning object repositories envisaged and which Brian Lamb reflects on here:
http://bavatuesdays.com/the-sordid-history-of-learning-object-repositories-or-a-chat-with-brian-lamb/. I still think repositories have a lot to offer the management of learning materials but they’re not the only option, are better as part of a wider suite of tools, are really hope they aren’t going to ask users about semantic density. To my mind Brian’s relfections highlight a number of reasons why the UKOER programme is implementation nuetral.
I’ve also realised that I’ve not yet pointed to a related resource from CC Talks about OERs A chat with Stephen Downes on OER http://creativecommons.org/weblog/entry/17860 I was also going to talk about Pete Johnston’s latest installment about Simple Dublin Core but that’ll have to wait for another day (http://efoundations.typepad.com/efoundations/2009/11/simple-dc-revisited.html)