Ben and Mena at Movable Type just released a Trackback threading tool that will build an entire tree out of MT trackbacks. Here’s a page showing the Trackbacks from one of my earlier postings
Yes, this is what ThreadNeedle is supposed to do. Yes, ThreadNeedle is not finished.
When you enable TrackBack for a posting, Movable Type embeds a small piece of RDF in the page, such as the following:
<rdf:RDF xmlns:rdf=”http://www.w3.org/1999/02/22-rdf-syntax-ns#”
xmlns:dc=”http://purl.org/dc/elements/1.1/”>
<rdf:Description
about=”http://burningbird.net/cgi-bin/mt-tb.cgi?tb_id=17″
dc:title=”World Summit Weblog”
dc:identifier=”http://weblog.burningbird.net/archives/000476.php”
dc:subject=”Politics”
dc:description=”Many thanks go out to Farrago for pointing out that there’s an extremely well organized weblog covering the World Summit. Now, this is what a political weblog is all about. Take a look and tell me you’re not impressed with…”
dc:creator=”shelley”
dc:date=”2002-08-2514:00:27-06:00″ />
</rdf:RDF>
This simplified RDF can be embedded because all of the RDF data is treated as attributes — notice that there’s nothing outside of the Description element? Attributes don’t get printed out via the parsers in the browsers.
So, why can’t I do this with ThreadNeedle? Two reasons.
First, I can’t control the output of the weblogging tools, so if I give you a piece of RDF to embed in your posting, the weblogging tool will try to add break tags (< br/>) to the code to handle line breaks in the RDF. This screws up all RDF processors.
Secondly, anything more sophisticated then the example I showed you requires special handling to embed the RDF in HTML/XHTML. Surrounding the data with the Script tags will work — the parsers ignore anything contained in script tags. However, this still doesn’t solve the problem of weblogging tool munging.
I can generate RDF, and it’s very doable to create an application that finds the RDF, and follows the threaded entries to the new page and looks for embedded RDF and so on (as the Movable Type Trackback threading application does, except that it uses the Trackback data stored in our local data stores) but I can’t control the munging of the RDF by the weblogging tool.
Ben, Mena, this was cool. Really. And thanks! I hate to be greedy, but can you and Ev and Dave and the other weblogging tool builders give us a window in the weblog posting page to include content that is embedded directly in the posting, without manipulation by the weblogging tool? Then others, such as myself, can provide functionality — such as ThreadNeedle — that isn’t dependent on the weblogging tool and without having to go through extraordinary means of handling this markup munging problem.
I realize that webloggers can turn off line breaks (either for a weblog or a posting), but many webloggers don’t know how to include their own HTML line breaking tags. What I’m looking for is the best of all possible worlds — a separate window that takes text which is added to the bottom of a weblog posting without any processing by the weblogging tool, while still allowing tool processing of the weblog entry itself.
Pretty please?
In the meantime, I have the Movable Type Threading CGI application running (Access here, pass in the URL of the page with Trackbacks). Feel free to try this with pages that have TrackBack enabled. If it slows my server too much, I’ll have to pull it, but we can give it a try for now.