Re: Google PR for mirrors

Personally, I don't care about the PR, just remarked it's has gone... it was 6 and in one day - NA I don't have a company to promote, so it was enough the answer "Google has removed duplicated pages". Finally, I agree that all the traffic must come to the main site and redirect to a local mirror In case the mirror is down - the user will be redirected to the second one or finally he will remain on the main site. Regards, Sebastian Adolfo Garcia Veytia wrote:

Well, the real truth behind this thread is that mirror admins just found out that the link to our websites included on the mirror pages just lost all value from googles perspective. Just admit it, there is nothing wrong. Let's remember that hosting a mirror for PHP should be about helping to spread our favorite language, not promoting our own companies. When we started hosting the .MX mirror back in PHP3's days, no one cared about page rank - you just hosted it. When google added some relevance to our site because of the mirror it was a welcome perk, but it was never the point of hosting it. Lately, when people noticed that hosting a mirror for PHP added a valuable link, offers started arriving to the mailing list from a lot of bogus companies not related to PHP and/or the internet industry. Companies only interested in the extra PR they might get (this is IMHO at least). By doing this, the mirror system benefits by being hosted by companies which are really interested in the language itself, not in the link from the mirror. just my $0.02.. Adolfo Garcia Veytia .MX Mirror Admin On Wednesday 01 July 2009 15:33:22 Hannes Magnusson wrote:

On Wed, Jul 1, 2009 at 21:24, Stream Service<info@streamservice.nl> wrote:

    
At this moment via de canonical link it sends traffic to php.net

(initiated by the php.net maintainers for as far as I can see). If someone for example from Moldova searches for a PHP function with (for example) Google it should be redirected to md.php.net/... if you ask me (probably the fastest result).

Search engines have no way of knowing that. The only thing search engines see are over 500 variations of the exact same page ([mirror.]php.net/[function.][strpos]/manual/[function.strpos.php][en/funct ion.strpos.php], we probably have more that I don't remember) which causes tremendous amount of duplicated content in the search results. By using rel=canonical (as we do now) there is only one page. One search result. "Collective page rank". We then take care of directing the user to alternative mirrors. -Hannes