Feedsearcher - search feeds by tag or by location
This site was created for the "Online Publishing" course in the Cultural Studies Master provided by K.U.Leuven Faculty of Arts with the help of my co-students Philip Mostmans and Nancy Fornoville.
Our objective for the course was to research ajax and analyse the key concepts, on a technical and theoretical level, including aspects such as future possibilities and usability. At the same time we were asked to create an example page that shows off some ajax functionality. Hence feedsearcher was born. Feedsearcher uses no server side scripting at all and relies purely on javascript and xhtml.
Feedsearcher uses ajax at two levels:
Interface (local data):
The site is driven by 3 xml files:
All the xml files are loaded with the dojo XHR (XMLHttpRequest) method.
Results (cross domain data)
- The results of the search by tags are fetched by using the Google AJAX Feed API
- The results of the search by location are fetched by:
- Google Maps API to set up the map, find a location and define the bbox (bbox stands for bounding box and holds the 4 coordinates that make up the 4 corners of the map viewport).
- a JSONP request with dojo.io.script from the Flickr search API.
- Flickr doesn't provide any rss feeds by bbox at the time (they do offer geo feeds, so use those if you find a good match by changing the url). They do provide bbox in the api and yahoo pipes is a great tool to mash up data from the web to custom feeds. That way the site provides a Flickr feed by bbox (it's slow though, so if you're planning to use the actual feed, have some patience, and try a few times or try the flickr geo feeds).
So, feel free to play around with our feedsearcher and leave a comment if you like.
Hans
