Tag Archives: aaronsw

Metadata seven years later

Aaron Swartz talks at the Creative Commons launch party:

“Right now you can only ask a search engine one question: ‘What pages have these words in them?’ When pages include RDF metadata, you will be able to ask more advanced questions like ‘What’s the current temperature in California?'”

Aaron, thank you for being optimistic. Someone still needs to be.

Back in 1996, when RDF was more an idea than an acronym, I worked on a side-project with my friend Travis Weller. It was based on RV Guha’s MCF and hosted at the domain metacontent.org. We demoed the first part of the software at Mactivity/Web, and I still have the slides for the presentation (click the logo to advance). It was a web server plug-in that served a site from an object database (the prototype used an embedded version of Userland’s Frontier database, but the idea was to eventually provide object-relational mappings to other more common databases). We called the web server portion Rendezvous, because it gathered pieces of content and metadata and assembled them together to serve a page. Apple likes that name too.

We also designed parts of the admin interface, which was to be the killer app to enable thousands of web designers to make metadata an integral part of their web site. You sell users on the product by providing a great interface for managing an entire site’s content, and then handle organizing the metadata behind the scenes.

Somewhere along the way, we realized the magnitude of our goals and grew disillusioned. Or maybe we just found better day jobs. Either way, the metacontent.org domain expired and was taken by someone else, we never shipped any software (although I still have the code on a backup disk somewhere), and the W3C’s Semantic Web effort eventually emerged with a ton of smart people trying to solve this problem.

Yesterday I noticed that the metacontent.org domain was available again, so we took it back. Maybe I still have some optimism left in me after all.

Reading and typography

The weather turned cold here yesterday, and that just contributes to my blogging apathy after the Thanksgiving weekend. I’m just too lazy to blog, and the backlog of unread items in NetNewsWire was over 150 this morning. Time to trim the subscriptions again. There’s too much to read, and hardly any of it really matters.

Reading text on the screen continues to be a challenge for most people. A recent newsletter article from Human Factors discusses optimal line length:

“What can we conclude when users are reading prose text from monitors? Users tend to read faster if the line lengths are longer (up to 10 inches). If the line lengths are too short (2.5 inches or less) it may impede rapid reading. Finally, users tend to prefer lines that are moderately long (4 to 5 inches).”

Aaron Swartz reviews “The Elements of Typographic Style”:

“What I’ve realized since reading it is that publishing documents on the Web, no less than preparing them for publication as a book, is typography and deserves the same care as that noble craft. There are some differences, to be sure, but the core it’s about making the meaning of the text shine through the words, a craft that has been practiced for ages.”

He’s also put some excerpts from the book online.