On 5/1/06, Erik Moeller eloquence@gmail.com wrote:
On 5/1/06, Kelly Martin kelly.lynn.martin@gmail.com wrote:
None of Wikipedia's mirrors has anywhere near the pagerank weight that Wikipedia itself has. With the exception of a small number of already high-prominence sites, Wikipedia linkage is almost certainly bound to substantially alter the rank of a given page in search engine results.
I agree with this now that it is also clear that mirrors are deliberately reducing external links. It is clear that Wikipedia's impact on the ranking of the sites it references must be significant. I continue to maintain that tagging all links indiscriminately as "nofollow" is essentially an admission of defeat; it is like saying: None of our links are worth spidering. We're just a wiki. Please, search engines, go to some other trusted sites, not to us.
I'd say it's more like saying "Don't assume this link is worth spidering just because we link to it." That's quite different from what you're saying.
I don't think our links are bad enough to justify such an explicit statement with all the consequences it has. And the continual pounding of "Wikipedia is an encyclopedia" does nothing to change the fact that it is a resource used by millions of people, that it is a resource which is embedded into a larger network known as the WWW, a network which itself is accessed by most people through portals and search engines before they arrive at a content source. It is in our interest, for our own project and for moral reasons, to help people to find high quality educational resources, as opposed to those which are promoted from within typically corporate, "trusted" frameworks such as AOL or MSN.
Ultimately I think the search engines are in the best position to determine how best to accomplish the goal of helping people find the resources they're looking for. And I think it's a very open question whether treating links from a wiki differently from links on a site with more tightly controlled access, helps a search engine guide people to high quality educational resources.
Of course if Wikipedia had the time and resources to dedicate to such a problem, a better solution would be to have a relatively small group of individuals go through and mark links as followable or not followable (high quality educational resource, or spam) manually. But I don't think that's reasonable.
Maybe putting a time-delay on the attribute is a good enough compromise. I don't know. IANA search engine guru.
The fact that Wikipedia maintains community-created lists of topical links has always been, to me, one of its strongest points, one of the best ways to further explore and find relevant information on a topic. [snip]
Sure, Wikipedia is great for that. But I really don't see how that's relevant to the nofollow tag, which is a tag for bots, and more specifically search engine spiders, not humans.
Do we, essentially, want an invisible wall of separation between user-generated content and content from "trusted" sites? Or do we want a better solution? I don't like nofollow. It is a hack. It lacks imagination.
On this point I agree. Maybe Google or one of the search engines looking to compete with them would be willing to consider a better tagging system. <a href=blah type=wiki firstadded=20060501195702> or somesuch. "Nofollow" is definitely a poor terminology. The tag should describe the link, not try to dictate to search engines what to do with it.
Anthony