by antipasto on Tue Apr 19, 2005 1:52 pm get it here: http://writtorrent.sf.net/
I tried to post about it here, figuring that all the Xlobby hackers would know what the heck XSLT was and how they could use it, but well, looking back, I barely communicated anything at all about my plugin.
The entire web is becomming more and more dedicated to Semantic Web technologies. This is basically XML documents that describe only the semantics of useful information. The idea is that if you provide only the web page's raw data, cataloged like the Dewey Decimal System, then you can have a global database that you can query directly!
That's a little abstract of a description, but basically, this is RSS feeds, podcasts, broadcatching, and it is also an especially cool technology when combined with P2P protocols like BitTorrent.
Xlobby, as you know, get it's X most likely from how wonderfully it implements XML documents to describe your media files, the screens you make, and the events and commands you program.
Well, this is semantic web technology! (Good job, Steven!)
The next versions of a lot of operating systems may well be designed just like the insight of our great guy here.
So, I wanted to be able to browse RSS feeds that had media inside of them (RSS *enclosures*) much like I browsed an interactive TV guide on my Wide Open West cable box.
So, it was pretty simple. XSL-T is document format (just like XML) that allows you to specify just precisely how you could change one random XML document you find out on the web into something you want, like, oh, an Xlobby database.
So now, by being able to create programmable actions that allow you to download and translate any XML document (given a personal library of XSLT documents) and then manipulate that information as part of your HTPC system.
Some things that I have accomplished with this plugin:
Videoblog downloading (The videoblogging people are cool!)
Blogdigger Media RSS Feeds (this makes a good 'TV Guide' for the Internet)
BitTorrent statistics (Azureus creates an XML document for its stats)
Flickr Tags / Popular (Browse random people's photography!)
Audioblogs / Podcasts
Since Podcasting has been centralized by Adam Curry at his site, I have a special plugin in the works for that which allows you to specify a different XSLT depending on what type of XML document the server returns. For podcasting, Adam Curry's site has an OPML which is a directory (in XML) of RSS feeds for podcasts!!! The RSS feeds then of course have an .mp3 enclosure for each show.
So, its pretty easy, once you've created a native Xlobby database on-the-fly
, to spawn all kinds of interface ideas like listening to a podcast in one room, and browsing raw internet media from the new site I'm helping moderate called http://www.OurMedia.org which will soon have RSS 2.0 media enclosures!
So, in my opinion, I don't need a TV capture device anymore with *my* Xlobby configuration. I have way more stuff from the general internet, all available from my remote, and right on the couch.
The plugin comes with a library of XSLT documents that I found useful, and a simple text file that describes the custom plugin command, which is made up of three simple things:
1. The XML document's URL (local or remote)
2. The Local Xlobby Database you are overwriting
3. The XSLT document that will transform the XML document into an Xlobby database.
It may seem cryptic, but its not... it's *semantic*!
The podcasting-specific plugin (coming soon) will be slightly more complex, much more usable as far as general XML document processing.
Best of luck, and feel free to email me or respond to this post or on any of the sites mentioned in this article. We'll get 'er done![/url]