It’s interesting seeing Dave Winer playing with search as a new primitive on his “blog” (I put it in quotes because his blog is a mega-archive and leviathan body of work)…

A new dimension has opened up. Until now writing has been more or less Of The Moment. As Google has punished me for not supporting their various hijacks of the web, the writing has become more ephemeral, kind of like Twitter – once written – hard to find again. With Daytona all of a sudden my memex has memory. My writing is upgraded a billion percent.

Makes me think about ways to upgrade search as a more powerful primitive on my own site….

Right now I’m using a version of Craig Mod’s search script for static site and it’s fine. Functional but not really transformative.

Instead I’m thinking about a new /search page that pulls together:

  • Blog archive (obvs)
  • Notes archive via microblog API
  • Full tweet archive (I’d self host a copy of my own tweets to do this)
  • My wiki and saved notes

Making everything searchable feels like fun and it might surface some nice things….

Update 12/14: Hmm I just downloaded by twitter archive and tweet.js is 50MB so it’s not going to be practical to host and search that everytime someone uses my search box….