Post details: MediaWiki software and rel=nofollow
Jan 08, 2008 : MediaWiki software and rel=nofollow
In case any of my readers run a MediaWiki site, you should know: By default, MediaWiki is configured to use rel=nofollow on links. This means anyone who sets up a MediaWiki site, not just wikipedia and its siblings. Here's more info and how to reconfigure it. That makes me mad. They should have better defaults.
Comments, Pingbacks:
Comment from:
Kyle Adams [Visitor]
· http://www.kyleandkelly.com/
I don't understand... isn't rel=nofollow a reasonable default in wiki software links, given the potential for abuse by spammers?
01/09/08 @ 07:51
Comment from:
Dave Brondsema [Member]
· http://brondsema.net/
Maybe, but I think it's a weak argument. The proper way to combat spam is to lock down your pages, implement appropriate CAPTCHAs, and/or review all changes. I think most wikis are set up to do that, and given that, it is a good default to have regular links.
Wikipedia doesn't use nofollow to avoid spam (thats what they have captchas and oversight for), they use it because they are a huge site with a lot of google juice and want to reduce the temptation to spam it.
Wikipedia doesn't use nofollow to avoid spam (thats what they have captchas and oversight for), they use it because they are a huge site with a lot of google juice and want to reduce the temptation to spam it.
01/09/08 @ 08:53
Comment from:
Disaffected Wikipedian [Visitor]
I thought you might be interested to know that I conducted an experiment to find out if bots actually follow these "nofollow" links. They do. Someone else did a similar experiment earlier (http://www.seo-blog.com/rel-nofollow.php), but mine placed the test link within Wikipedia. First of all, I found a stub about a subject I'm interested in. Then I wrote a piece expanding on the subject and placed it in an orphan page on my site. The only way anyone could find out the page existed was if they followed the link I placed in the Wikipedia stub article.
I got the same results as the previous experimenter. Googlebot turned up within 7 days, but didn't add the page to it's listings. Yahoo! Slurp crawled the page within 4 weeks. It returned again a few weeks later, after which the page appeared in Yahoo's search results. In the meantime the page was also visited by:
1) Mozilla/5.0 (compatible; archive.org_bot/1.13.1x +http://crawler.archive.org)
2) MJ12bot/v1.2.0 (http://majestic12.co.uk/bot.php?+)
3) voyager-hc/1.0
It's not impossible that other sites copied the link during that period without rel="nofollow", but if so, they never showed up in the referrer logs. However, several months later, I noticed a referral from another wiki that uses Wikipedia material without rel="nofollow". At that point I considered the experiment finished and stopped scanning the logs.
I think Wikipedians must be incredibly naive to think the only reason for link spam is to get benefit from PageRank. There are articles online by 'black hat' SEO consultants explaining several other incentives for spamming Wikipedia. For example, to attract visitors to product pages. You can find recent examples of product page spam here:
http://en.wikipedia.org/w/index.php?title=Web_counter&action=history
I got the same results as the previous experimenter. Googlebot turned up within 7 days, but didn't add the page to it's listings. Yahoo! Slurp crawled the page within 4 weeks. It returned again a few weeks later, after which the page appeared in Yahoo's search results. In the meantime the page was also visited by:
1) Mozilla/5.0 (compatible; archive.org_bot/1.13.1x +http://crawler.archive.org)
2) MJ12bot/v1.2.0 (http://majestic12.co.uk/bot.php?+)
3) voyager-hc/1.0
It's not impossible that other sites copied the link during that period without rel="nofollow", but if so, they never showed up in the referrer logs. However, several months later, I noticed a referral from another wiki that uses Wikipedia material without rel="nofollow". At that point I considered the experiment finished and stopped scanning the logs.
I think Wikipedians must be incredibly naive to think the only reason for link spam is to get benefit from PageRank. There are articles online by 'black hat' SEO consultants explaining several other incentives for spamming Wikipedia. For example, to attract visitors to product pages. You can find recent examples of product page spam here:
http://en.wikipedia.org/w/index.php?title=Web_counter&action=history
02/16/08 @ 04:31 Leave a comment:
