Google SEO News and Discussion Forum

You can bet they know. Most all threads from WebmasterWorld hit the SERPS with uncanny swiftness. Yep, they surely must know.

"Perhaps we may think of a way to bring this thread to google's attention."

[edited by: lawman at 10:25 pm (utc) on Mar. 11, 2005]
[edit reason] No Specifics Please [/edit]

Wouldn't an email tell us much of what we need to know?

[edited by: lawman at 10:25 pm (utc) on Mar. 11, 2005]
[edit reason] No URL Drops Please [/edit]

Also, we can find out what site is "hijacking" us, the problem is that in many cases we can't do much about it.

[edited by: lawman at 10:27 pm (utc) on Mar. 11, 2005]
[edit reason] No URL Drops Please [/edit]

[edited by: claus at 7:15 pm (utc) on Mar. 11, 2005]

I was the one who sugqested the <base href=""> tag and I saw a few questions about it so I will explain it's function a bit and how it could 'possibly' prevent hijacking.

There are 2 ways to crosslink within your site.
1: Absolute links
2: Relative links
Absolute linking is when you use the full cannonical url in every single crosslink within your site. There is no question where any link points to.

Relative crosslinks: If the page examplea.htm is in the same directory as exampleb.htm you could make a link from a to b by simply typing <a href="examplb.htm">link on page a to b</a>
or if the target examplec.htm is in the directory above the directory holding a and b and yoou want to link from a to c you could use <a href="../examplec.htm">Link from a to c</a>.
The link is relative to the page it points from.

The <base href="absolute canonical URL of this page"> is a header tag. If you use relative linking you really really should have a base href on every single page, even if it works without it, this gives the browser a reference to use regarding the relative links on that page. The links are relative to that base.

As far as using this META tag along with absolute crosslinks, kind of pointless but it MAY I repeat MAY help prevent hijacking, or at least cause the hijacker some work, because he would have to remove or alter that META tag in order to place the page in any other domain. Also he would have to fix all the relative links because they would mean nothing in another domain. And we all know hijackers avoid any kind of work for themselves, that is why they hijack.

As far as the 302 redirect problem this META tag "MAY" help. Because when google is deciding which page is the real page it also has this base reference which is absolutely clear that this page absolutely belongs at this canonical URL. But google only takes a 'picture' of the page so even though I use this META tag I'm seiously considering adding some dynamic content to my pages (a visible hit counter "MAY" work so that every visit to the page changes the content) that way every picture google takes is different in some way.

Welcome back to the 80's

[edited by: lawman at 10:30 pm (utc) on Mar. 11, 2005]

[edited by: lawman at 10:30 pm (utc) on Mar. 11, 2005]

I'll spare you the part where I quote Marx, LOL.

[edited by: activeco at 10:44 pm (utc) on Mar. 11, 2005]

this also probably won't fix sites that have already been hijacked.

The main question remains if Google implemented it.

Pages indexed droped 80%, ofcause visits also droped 98%, my PR came back 2 month ago, but no changes in serps, googlebot has also not seen much of my site, even if I have cookies for it :), I also can not see my site when I ad the 0filter, so I think I will first see improvement when googlebot coming back for good.

It is quite a way to look at things, search engines would go out of business if we blocked them from indexing our sites. How about everyone in the theoratical union blocked all search engines but union certified ones, and of course, to become a union certified search engine you have to show union members websites at the top of the listings and not some fly-by-night scraper sight. Your right, were the webmasters, were the ones creating content, why are they the ones in charge. What right do they have to show our content under anothers name, none.

Reading your post reminded me of something I had completely forgoten about. About 4 months ago I was looking around at DVD rental sites comparing deals and I came across one that had some other name but it was the blockbuster website. It wasn't at the top of the serps but about 20 down. I spent about 5 minutes scratching my head trying to figure out why blockbuster would allow someone else to use their name site that, seemed very odd. I had completely forgoten about it until I read your post about amazon, err, big river site.

While the theory would be to redirect when the bot is arriving via a 302 redirect, google's bot won't have this referrer information when it views the page, and there's no way the server could detect that the link was using the 302 method.. AFAIK

It has nothing to do with referrer or any other header info.
You simply do your own 302 redirects five times in the raw, with the last one having the content.
If someone does 302 to you he can't pass all the 302's in the loop, of course if "five redirections" rule being implemented.

This isn't something I'm expert on, but would there be any negative impact for the users / bots if they went through a 5 stage redirect for every page they accessed?

has any testing been done?

*Crosses Fingers*

That would sure throw a wrench into the spammers/hackers works, they ALL have multiple domains with 302's all over the place. It's part of what they do.

I like this idea and it is something that could be easily implemented I imagine. On a further update I got a canned response after complaining about three (3) 302s that said follow the google guidelines for good site design. They think I'm mental or something, that is an insult to send me that. I would have preferred no response.