from Denham Grey's KMWiki, a large collection of links to various knowledge tools, including

11:11:42 AM    

So I finally got a chance to actually look a little more closely at the feeds from the existing repositories that I posted last week and am slowly starting to get my thoughts together. Here are some reflections.

Not all <links> are the same

A seemingly small thing that jumped out for me is the different interpretation of what the <link> element should point to in D'Arcy's CAREO feed - his is the only one that uses the object's URL instead of the metadata record's URL. I'm interested to hear from D'Arcy's whether this was intentional and if so, the reasoning behind it. There aren't any rules yet as far as I can see; at first I thought that pointing to the record made most sense because un-extended RSS doesn't do a great job of displaying any of the existing metadata schemas I know of, but then I realized that this might just be anticipating extending RSS 2.0 through namespaces and so in fact the <link> field pointing to the object's url might in fact be the right way to go.

Some metadata filters seem more exciting than others

I'm glad I placed multiple feeds from the same source that used different criteria together onto this page. While I can see some cases where presenting the latest entries to the LOR might be useful (and I understand that proofs of concepts need to start somewhere), streams like that are likely to be less interesting than ones that use more pedagogically relevant metadata fields - e.g. "latest history objects," or "biology objects containing the word 'blood'."

Which led me to my next thought - unlike the way RSS currently supports news distribution where for the most part a 'channel' is defined by the content author , RSS as a distribution mechanism for learning objects repositories seems to get really interesting when the channel can instead be the latest results of a user-defined search using specific meta-data fields. So if you have user accounts in your LOR, a user could perform a search, be they an instructor wanting to know when any new objects are posted that might help in teaching fluid dynamics to 2nd year engineering students, or a student wanting to see new objects that could help with their term paper on mass production and the Civil War, and then be able to create a persistent URL that was the RSS feed for additional results for this search that could then be aggregated in their 'news reader' or whatever their personal learning object aggregator might be called.

Like I said, still early days in puzzling this through, but wanted to put something down. Off to bed (unlike Brian, I need some sleep before I die ;-)  - SWL

11:10:21 AM