2010/7/18 Ray Saintonge <saintonge@telus.net>

Perhaps it's because these systems are designed by obsessives for obsessives. 


It's all very nice to have organizing metadata, and meticulous accuracy in our text, but in our search for perfection we lose sight of our objective of making the information available to the greatest possible public in a way they will understand.  We want a place for scholarly comment and annotation of text in an environment where the user has a reasonable confidence in the accuracy of the root text.  We want to enable wikilinks, and added footnotes.  We want the user to be able to know that alternative treatment of the material is available at the click of a link, and sometimes on the same page.


Using the metadata in a card catalogue may help me to find some of the material in the library, but unless you can supplement this with the rich rewards of browsing the shelves.  Telling people to RTFM is no substitute for the hands-on efforts of trial and error.  It would surprise me if any more than a small fraction of Runeberg's users even knew that Dublin Core existed.  As the manual which they must read it confuses more than it helps.



I apologyze if I'm going a little OT, but I think that source needs en enhanced incategory tool, so simpifying a lot classification of works with good, regularly updated lists. I tested DinamicPageList extension into a small external wiki, and I'm enthusiast about. This existing tool (used mainly into Wikinews, but into at least one Wikibook too) is, in my opinion, what's needs to use our present data at their best.

I'm searching from long a incategory trick (using categories generated from templates!), and I've a script that does it by category intersection, but this means to re-write again and again the same list pages, where DynamicPageList doesn't wastes cronologies and - I guess - it's much more efficient than a series of API's queries by a bot.