I’m at a talk by David Weinberger, whose upcoming book is about categorization and knowledge. You could think of it (as far as I can tell) as the book that tries to put a finger on why tags (despite their shortcomings) often feel as good or better than the heirarchical, outline-based view of data that all of us in the Western tradition have been brought up with, and how heirarchies and tags and all the wonderful metadata we have will play together going forward.
I’ve actually seen the slide deck before, so my mind drifted a bit to libraries after he showed a slide of them, and how I’d love to have all my books catalogued independently of their defacto data record in Amazon. I’ve mentioned here librarything, and (way back), alexandria. It would be sweet for Alexandria to get information from librarything instead of Amazon. It would be cooler still if it imitated Delicious’s coolest feature and read barcodes via a webcam and v4l. The code pieces are all there; it would be great if someone put them all together on Linux.
[Maybe the best insight from the talk, by the way- when you order a beer, and it isn’t the perfect beer, you don’t put it down and say ‘not the perfect beer’, and go to the next bar. You decide ‘good enough’, drink the beer, and move on with life. Data is often ‘good enough’- but habitually we often want the perfect information, and our tools (like heirarchies) have often either given perfect information or no information at all. Tools that help us find the ‘good enough’ data, verify that it is good enough, and move on with life, are really important, and tags and other metadata, well-used, can ideally help with that.]