Forum Moderators: open
For instance, I came across a site that has a PR of 5 for www.domain.com, and a PR of 3 for domain.com - both on the index page!
Does anyone know why this is the case?
I think what you were seeing results from a quirk of timing... the timing of when you make changes on your page vs. when Google visits. I think it's most likely to happen if Google follows someone's "non-www" link to you, and later follows a "www" link, (or vice versa) but you've changed something in between. In that case G. does see them as different. If the content remains unchanged for a while, Google is more likely to recognize them as the same page, as CIML said.
I don't think it matters which one you choose, but try to get your link popularity focused on one version or the other, not split between the two. Use the same form consistently in your own internal linking, and when you're trading links with other webmasters, tell them which version you'd prefer.
Can't hurt, might help!
> Why does Google not just assume that www.cnn.com is the same as cnn.com?
Because this is not always true; www.cnn.com is a subdomain of cnn.com.
While it is true that many - even most - domains are set up so that www- and non-www- hosts are identical, this is not always the case. The search engines must take extra steps to verify that pages on www- and non-www- are identical before simply assuming that they are. They do this 'as a courtesy' to webmasters, and in order to eliminate redundant information from their indices.
As ciml pointed out, many search engines will compare these contents, merge the backlinks and PR, and drop the less-popular URI, but it takes awhile. To avoid the problem outlined by buckworks, you can implement a 301-Moved Permanently redirect from the non-preferred to the preferred URI.
Jim
If you do it be aware that your site may completely disappear for at least a few months. I'll never do redirects for PR cosmetics again.
Does anyone know if this is true?
Barry Welford
What is an AREA?
And does your answer mean that Google is only dealing with URI's and not at all with URL's?
Thanks for your explanations.
Barry Welford
URI is a slightly more general term than URL....Think of a URL as the kind of URI the average person is far-and-away the most likely to deal with.
I have found this to be consistent over time - the pages do not merge and one page does not drop out of the index that is. The reason (that www is a subdirectory) is explained already in this thread.
I have implemented the 301 redirect mentioned by jdMorgan and never had any problems with it - there's a lot of threads in here describing the method, so i won't double post it here. It works so that no matter which version you are entering in the address line you will always only see one of them. In my case, inbound links for "www.example.com" gets redirected to "example.com" but it might as well be the other way round. I did it this way out of principle and because the non-www version had higher PR (i would have done it anyway sooner or later, PR was just an extra incentive).
The latest redirect i did of this kind was for my most important site - it was around a week ago, in the middle of the "Fritz" update. Both versions are still found in the index, but the non-www version tends to get displayed in the serps and it seems the www-version is slowly on it's way out. All this is exactly as expected.
/claus
