In-Feed Feedback

I’m playing second fiddle to Danny again right now, implementing his Reader Provided Blog Enhancements as a wordpress plugin. Currently I’m posting to a local MySQL table, from which I can pull the relevant information and create different views later.

This is a great example of some code that would be nice to do with XmlHttpRequest: rather than having the post go to a redirect (which is only going to work if the user has referrers on right now, otherwise it just brings to a single page that says it was completed), it could all be done in the client, and the user would never have to leave.

However, there’s a couple problems with this.

1. RSS Aggregators are not web browsers, and depending on the level of the implementation they are using for displaying HTML content, they may not support Javascript at all or not completely. I’m hoping that HTTP POST will actually do something useful for most of them, but even that is a guess.
2. Online aggregators such as LiveJournal oftentimes strip out Javascript to prevent malicious cookiestealing (and for good reason).

So, unfortunately, javascript is out.

Couple changes that will be happening in the meantime while I work on this: RSS feeds will be limited in size to 1 or 2 posts, so that you don’t get change-flooded every time I turn the plugin on or off to test something, and you may see the review boxes appear or disappear.

Anyway, nothing much to see yet, but I will be doing RDF export of annotations provided, so the data isn’t going to be lost, and I will be working to clean up the code and make it “just work” with a WordPress plugin, hopefully. They are surprisingly easy to write. I didn’t realize how simple some of the stuff was. Keep your eyes on the prize!

Oh, and Danny? Your RDF in that post is broken. Missing rdf:RDF, and one of your close tags is missing a /. Thought I’d let you know 😉

6 Responses to “In-Feed Feedback”

  1. Jimmy Cerra Says:

    You are correct about the mistakes. Cooincidently, the rdf:RDF is optional according to the validator at least.

    P.S Why is the one image upside down? A new way of getting attention in Flickr? j/k 😉

  2. Danny Ayers, Raw Blog Says:

    Semantic Web Blogging — Danny @ 11:08

    crschmidt’s now playing with In-Feed Feedback, implementing the thing as a WordPress plugin (”doing something us […]

  3. Christopher Schmidt Says:

    Jimmy: the rdf:RDF is optional, but only with a slight change to the syntax. Instead of using rdf:RDF as the root element (as Danny has done), you would have to use rdf:Description as the root. This is suggested by many schemas such as DOAP. No matter what, RDF/XML must be valid XML, so you can’t start with one thing and not end with it.

    The image is upside down because it was a transfer from my gallery install, which for some reason flipped a few things, and because it amuses me to have it upside down, so I haven’t changed it.

  4. Jimmy Cerra Says:

    Chris, you can use any arbitrary element as the root. Try this xml fragment with the W3C’s validator:

    <Action xmlns=”tag:example.com,2005:” />

    That is a valid RDF/XML document! You are correct that it must be valid XML.

  5. Christopher Schmidt Says:

    Jimmy: Yeah, that was really all I was saying. (Damn asynchronous conversations ;))

  6. Danny Ayers, Raw Blog Says:

    I’ve not had time to play with code on this (still fiddling with my backend 😉 but Christopher Schmidt and Russell Beattie have provided some useful info by including forms […]