Feed Sign in with OpenID OpenID

Simon Willison’s Weblog

Creative commons query

Aaron Swartz has been talking to Google about indexing Creative Commons licensed works:

From Google the news was mixed. He said he wouldn’t start indexing .0 URIs, which includes the URIs for all our licenses. He also said that he wouldn’t parse RDF for at least six months, since it required involved changes to their system and added overhead (which you need to keep down when parsing 3B pages). However, he did say that if we added <meta> tags for license information, they’d add a new search key like link: or inurl: right away, since they already had a meta tag parser.

Did they say anything about using the link element instead?

Posted 2nd March 2003 at 11:34 pm

This is Creative commons query by Simon Willison, posted on 2nd March 2003.

View blog reactions

Next: The importance of titles

Previous: Dependencies suck

3 comments

  1. Umm, I can't see anything about google on that weblog entry. Did you mean to link to somewhere else?

    Jim - 3rd March 2003 19:29 - #

  2. Nope, it's a link to my suggestion that the Creative Commons use a <link rel="copyright> element to indicate the license under which copyright is covered, rather than embedded RDF (hidden in a comment). If Google could index documents based on their link elements it would provide a very quick way of finding CC licensed content. Of course, the same is true for a meta tag but I find links more elegant.

    Simon Willison - 3rd March 2003 19:39 - #

  3. Nah, I was talking about the link to Aaron's site.

    Google in some cases is pure gold, but in others is absolutely brain-dead. In cases like this, I find that they ignore specifications and just go with what is currently happening on the web. That may be a good approach in the short-term, but if they paid more attention to semantic information like this, it would give a clear incentive for people considering using it, and thus benefitting google in the long-term.

    As an example, I believe google pays attention to file extensions when spidering a site - so your html pages might not be found if they use a non-standard extension such as http://www.example.com/index.info - even if it's correctly labelled as text/html. Sure does wonders for stable urls :(

    You really should add a preview page to this btw... :)

    Jim - 3rd March 2003 22:35 - #

Comments are closed.

Previously hosted at http://simon.incutio.com/archive/2003/03/02/creativeCommonsQuery

A django site