Jon Udell: Google index synchronization

Jon's Radio : Jon Udell's Radio Blog

Updated: 8/6/2002; 12:33:52 AM.

Note: Jon's Radio has moved to InfoWorld

Monday, July 29, 2002

Google index synchronization

I planted the phrase "reverse anarchy" on this page the other day. Now my GoogleBox finds this page instead. But interactive Google doesn't find it at all.

Hmm. I guess I'll send this URL to Google's self-improvement department. It seems churlish to complain, given that Google does so much, so well, for free. But since they're inviting questions, I'm curious to know how the different indexes are managed.
10:55:19 AM

MT-RefSearch, and client- vs server-side scripting

A clarification from Eliot Landrum:

Thanks for the link and mention on your site about MT-RefSearch (http://radio.weblogs.com/0100887/2002/07/27.html#a357). There’s a couple of corrections I’d like to make though.

First, MT-RefSearch is a PHP script that searches Movable Type (a Perl engine) MySQL data. Basically, if you use Movable Type and have it store the data in MySQL, then my script can search the data and display results. Movable Type itself is written in Perl, but outputs static HTML that can also include PHP/SSI/whatever scripts. I’m not sure if the last three sentences actually cleared things up or not…

Secondly, I believe that Brad is using my script on his site (which is Movable Type powered, not Atomz, as far as I can tell).

Again, thanks for the mention. I hope more people will start to use reference data in interesting ways like this. I’m surprised no one has done things like this before.

Thanks, Eliot. I've tried Movable Type, and I knew it was Perl-based. But I haven't tried its MySQL storage option, or done any PHP includes from its templates. And since the site uses Atomz to search itself, I guessed that was the mechanism used to extend the search.

One reason there is not wider use of reference data is that not everybody has access to server-side scripting. Radio pages, for example, are dynamically generated but statically served -- a strategy that I like very much on the whole, but sometimes chafe at when nifty features like TrackBack and your search extender show up in MT.

Given Radio's decentralized and client-based architecture, such features are usually implementable in client-side JavaScript. Going forward, I'd like to see browsers become more able consumers of web services, and server-side and client-side scripting merge into a common paradigm, so that deployment of such features can happen in a more homogenous way.
10:28:16 AM