Dir23.com – My New Search Aggregator

I miscalculated the amount of stress using MySql to store search urls would put on the server. Now it is hard-coded javascript so it puts the stress on the browser side. Hopefully this one stays up. It’s working now, but I won’t be able to add to it until the weekend.

I want to create a script that will spit out opml files for the search queries next. Since a lot of the search engines will have feeds for their search results, I can wrap these up in an OPML that can be imported into a feed reader for listening.

The second version that is up is basically open because it is written in Javacript. It’s rather simple. Go to each of the search engines perform a search that includes a space. You need the space because you need to know what each engine uses as a delimiter (%20,%2B,+,_). Once I had that, I built the site to grab the query, insert it into the search urls and then replace the spaces with each engines delimiter.

But I had a trick, an unconventional one. I store the urls in a Filemaker database and use calculated fields to create the javacript on the fly as I add new engines. Then I export the javacript for the site.

The initial version used a mysql database for the sites and php to serve the results. That added unnecessary work for my server when all I used the database for was convenient storage for the data.

I plan on expanding it as I use it. Still coming up with ideas.

It saves time. Google doesn’t provide everything. If I have written a post about “twitter”, I can find other targeted conversations that I can join and perhaps bring a few people back here to read my post. It beats going to each search engine separately.

I am using javascript on Dir23.com now. It uses all client side resources taking the load off of my server. As far as serving actual feeds the same way, I have been looking at Flash and someone had suggested Silverlight to me.