metadata

Jeff Jarvis has a spot on post about using tags for information about people. People tagging creates a marketplace by matching content against people’s needs and interests. In the case of jobs – a tagged job requirement matched against a tagged resume. As it happens I will be launching a new product soon that includes some of these elements. I want to eventually take tagging one step further and use it to create any type of metadata, allowing people to create tag/value pairs in their own namespace and create RSS modules on the fly. The premise of this is that at the moment the name portion of the name/value pair in a tag is ‘category’ i.e. if you tag something as a ‘resume’ you are saying ‘category=resume’. What is interesting is when you allow people to invent their own ‘name’ portions of the tag. e.g. ‘category=sushi_restaurant location=new_york’. That’s what I’m…

Jon Udell looks at multiple metadata values within CSS class attributes. CSS is the logical place to put metadata in that metadata often, but not always, needs to be styled – e.g. a ‘headline’ might be big bold and blue. But the main problem for me is that by using attributes in HTML tags to define metadata you have to count nested tags in a parser, which is a total pain and can cause problems if fragments of information are syndicated. Wouldn’t it be much better if HTML were amended to include closing attributes? i.e. you could say: <pre class="headline">DOJ sues Microsoft</pre class="headline"> instead of: <pre class="headline">DOJ sues Microsoft</pre> That way people wouldn’t have to invent tags, which couldn’t be styled, such as: <headline>DOJ sues Microsoft</headline>

Marc Canter links to Peter Van Dijck’s excellent illustrated semantic web debate. This is a really captivating interface idea. Basically it is a set of quotes in the endless pedantic [sic] web debate next to pictures of people. It reads very much like a cartoon with speech bubbles next to faces and somehow comes to life much better than a threaded discussion. I would like to see the inclusion of a picture (and one-line-bio) of someone next to a trackback item so that you could better follow a distributed discussion that linked to an original posting via trackback.

It is obvious now that the major record labels are screwed. But what about this from Clay Shirky: “The internet has lowered the threshold of publishing to the point where you no longer need help or permission to distribute your work. What has happened with writing may be possible with music”. What would weblog software geared around musicians having weblogs where they distribute their samples of their music consist of? A weblog posting would have infomation about the song and a link to an MP3. A parallel RSS feed could have metadata about the song and the MP3 could be delivered as an enclosure. RSS aggregators would take advantage of metadata within the music RSS module. Shirky: The Music Business and the Big Flip via Scott Schrantz

Jeremy Allaire’s Radio “Lots of industry analysts have commented that ‘public web services’ (e.g. web services that can be accessed and used through Internet-accessible public APIs) haven’t really happened. When one looks at RSS aggregation sites such as Syndic8.com it’s quick to see that there are thousands of “web services” out there for people today.” Like most things on the web the grass roots phenomenon is often the more successful.

In the right hand side bar I have indicated my own one line bio within largely decorative <olb> tags. Any others that I see around I will be adding to my Blogroll. In the RSS 1.0 output I have added the <olb> and links to an HTML bio (not yet up) and an XML version. This metadata is added to each item. I will be replacing the RSS 0.91 output with RSS 2.0 which will include the ‘author’ module. <author:olb>Scottish, based in San Francisco, former architect at Foster and partners, co founder: Moreover, co founder: Origins.net, co author RSS 1.0</author:olb> <author:linkHtml rdf:resource="https://www.davidgalbraith.org/author/bio.html" /> <author:linkXml rdf:resource="https://www.davidgalbraith.org/author/bio.xml" /> More soon.

Blogroll entries with bio’s have started coming in – keep sending them. Some of the the one line bio’s describe the weblog rather than the author. When doing your ‘<olb>’ try and make it a sentence that best describes yourself – think of these as personal headlines that could eventually link to a full bio just as a news headline is sometimes written as mini digest of a news story.

If you look at my Blogroll you will see that all items have a one line description. Weblogs are about people and the first thing I do when looking at a new weblog is try and find the ‘about me’ info. In order to link weblogs to bios I am proposing a blogroll format of $name ($blogname): $one_line_bio (maximum 150 characters total). Where this becomes the headline link to either an XML bio or ‘about me’ page. So here is the deal, if you email me a blogroll entry in this format I will add you to my blogroll. [OK so not exactly much of a prize – but this is a kick-off to something much more interesting, I promise]

Help! Is the following legal HTML or, more specifically XHTML? <span class="foo">bar</span class="foo"> I want to write a ‘dumb’ metadata parser that can parse fragments of markup and doesn’t need to count in and out of nested div or span tags. Being able to explicitly label attributes in closing tags would help, to say the least.