Welcome to WebmasterWorld Guest from 18.104.22.168 , register , free tools , login , search , pro membership , help , library , announcements , recent posts , open posts Become a Pro Member
Extended listing links. The sa=X parameter causes a Google Bug! InsideSEO msg:769474 10:41 am on Apr 25, 2006 (gmt 0) I found a Bug in Google's additional Links that are shown sometimes for the top ranked site (Navigation Links). Whe you click on it Google appends '?sa=X' on every Deeplink. Try it. Most of the webpages have no problem with this additional parameter but in some case it leads to a broken link. Even Google itself is affected by this bug! Examples: [ ...] The Link 'New features!' and 'Google Accounts' google.com or [ ...] The Link 'Valid XHTML 1.0!' google.com
Maybe we can find some more famous examples where this bug causes trouble
tedster msg:769475 6:41 pm on Apr 25, 2006 (gmt 0)
Welcome to the forums, InsideSEO.
This sa=X business was mentioned a while back, but I hadn't noticed that it actually breaks some links -- even for Google itself. Good observation!
Why they are keeping that part of the query string, even after passing the click through their own tracking script, is very confusing to me. I sure hope they don't start indexing those urls.
abates msg:769476 10:35 pm on Apr 25, 2006 (gmt 0)
Huh, that explains some 404 errors I've seen in my logs.
Has anyone written an .htaccess rule to redirect people to the proper URL?
Herm msg:769477 6:54 am on Apr 26, 2006 (gmt 0)
I'm also affected. It seems the old google behaviour was to make these hrefs as follows: /url?<google-parameters>&q=<an_extended_listing_url>
Now they malform the url with two "?" by constructing it like this:
I would love (and kind of expect) a fix at their end.
InsideSEO msg:769478 7:54 am on Apr 26, 2006 (gmt 0)
Wow, that was fast. Seems like Google fixed this bug. No problems with my sites and the above mentioned any more. But I haven't checked every datacenter Herm msg:769479 10:00 am on Apr 26, 2006 (gmt 0)
Fixed for me too. Google now encodes the second?. Which is a very good thing...