The archives browser userscript achieves this by first extracting a file barcode from the page and then sending this barcode off to a php script sitting on my server. Because the image locations are all based on the barcode, the php script can use it to build a Media RSS feed. The ‘View in 3D’ button launches CoolIris using the url of the php script.

The Mapping our Anzacs scrapbook uses a different approach. First of all I had access to the template, so I could insert a link directly rather than with a userscript. The link again goes to a php script, but this time it uses the Tumblr API to gather the info it needs to build the RSS feed.

In theory, you should be able to write a userscript that extracts all the information you need from a page of thumbnails to generate the Media RSS feed. Indeed, I’ve done this for the online photo collections at the PROV. The problem is that there’s currently no way of directly loading this feed into the CoolIris plugin. The best you can do is load it into a ‘flat’ slideshow using the CoolIris default javascript option. It looks quite nice, but it’s no 3D wall. If you want to see it in action, then install this userscript and go the the public transport photos.

If you want the full 3D CoolIris experience then you need something a bit more complicated. If the server is using GET to return search results, then it should be possible to use a userscript to send the current search to a php or python script somewhere else. This script could then do all the parsing, pull out the details and return them as an RSS feed.

The PROV site doesn’t allow this. So the only way I could see to CoolIris-enable it from the outside would be to have a userscript extract the image details then send them on to a server script somewhere to generate the feed. I had a go at this, but ran into a problem with the length of the querystring that was being passed.

Of course, the best solution would be to write and deliver the Media RSS feed from the PROV servers themselves! Or even better, make all the data available through an API or as Open Linked Data and leave it to users to develop all sorts of cool new interfaces.

Now I’m trying to work out if I can cut and paste an url such as http://proarchives.imagineering.com.au/index_search_results.asp into the cooliris url so that i can browse images off specific sites?

Cheers, asa

If you are out there and you get this…knew nothing about screen scraping a few days ago. Came across this http://newprosoft.com/web-content-extractor.htm Not bad for someone like me who has ZERO scripting/coding background!