Google's rel nofollow

This morning I read about Google's new initiative to end comment spam. They debuted the idea with support from all the major blog vendors, Yahoo, and MSN. As I read on I was amazed at the simplicity of the idea: include an attribute in the URL that tells web crawlers to not follow the link. When a spammer comes by to drop their URLs in comments or trackbacks, the blog software parses the URL, adds the nofollow attribute, and it's done; the URLs won't count toward the spammer's PageRank (or any ranking in any other search engine). The incentive for spamming is basically gone. Of course, they may not care, opting to just spam anything they can with the hope that the blog won't have the nofollow attribute implemented.

I went ahead and added support for this into blur6ex today.

URLs in comments and trackbacks will go from being like this:

< a href="http://romej.com" > romej < /a >

to this:

< a href="http://romej.com" rel="nofollow" > romej < /a >

I had worries that the new attribute was a big XHTML validity violation, but apparently the rel attribute can be just about anything.

Amid thinking about all this I discovered another use for the rel attribute. Simple things like this are helping cobble together a more semantic web. It's cool.

Talk about being late to the party: I finally got a Gmail invite today through the isnoop spooler.

Novell has been avidly hiring top talent from the open-source world. It also looks like Beagle will be available for Windows.