Flutterby™! : Intertwingled technology 2010-01-04 20:56:40.017164+00

Intertwingled technology

2010-01-04 20:56:40.017164+00 by Dan Lyke 6 comments

Johannes Ernst, of LID and thence OpenID fame, has started prognosticating: Upon 2020, musings on what the world might look like in a decade. His first entry is Technology's Future is Deeply Intertwingled, and I was going to comment there, but realized that this particular sort of navel gazing is better kept on my own site. So, two observations:

First: Back in the early Naughties I was at some weblogger gathering when "tags" and "folksonomies" were all the rage, and someone asked what I thought, and I scoffed and said "in Library school, the Dewey Decimal System is a full semester course". Of course I also went home and implemented a classification system for my blog, but even today I find that it isn't as useful as I'd hoped, and, further, that classification is (as Nelson points out) really really hard even when one person builds the taxonomy.

I think there's great value in "the semantic web", I'd love to see more people publishing more information that lets me use online data outside of the human interpretation context, but, alas, I also think there are huge gains in building heuristic systems (which may evolve into something "smarter") to interpret data in the context in which we wish to use it. Acknowledging, for instance, that people aren't likely to standardize on a format for dates of events, and instead building systems to parse text for event information.

Which brings me to: In your example, it'd be really nice if you could get the straight data stream for traffic and overlay it on to the display of your choice, but if the publishers of that data or the display are getting their revenue stream from alternate sources then they've got a financial incentive to make that data less useful to you.

I think the guy working on fully intertwingling teh intarwebs also has "flying cars" on his backlog.

But I find my search and Google's search (they have different results, let alone different orders to their results) generally work better than the topics.

Well, I'm running 2:1 tags over posts. I tend to be really quite freewheeling when it comes to tagging, and I think my favorite tag so far is "dead zombie languages rising from the dead" (yup—real tag I have and it references this post, but the link has obviously changed in the eight years, or I really had an issue with consultants in Chattanooga).

Mine has a keyword based topic picker, which is then available for editing. I usually do, though not always. Other front-page posters don't.

The problem I've found is that with almost 13k entries I'd have to go do a lot of splitting and reworking to make some of those categories useful. "Sexual Culture" is, at this point, way to freakin' broad, for instance.

I have tags for every entry on my blog, but they're added by hand (unlike yours, which I think is automatic). And yet, I still find it difficult to find stuff using the tags and I often have to resort to grep. And that's because the tags I pick at the time of writing are often not what I'm looking for five years later. I'm not sure what to do though.