New (sorta) Info From Google
 

I found a thread at another board that was started by Googleguy (a rep from Google) and it brought up some answers to some common questions posted over here a lot as well as debunking a few myths
http://www.webmasterworld.com/forum30/29720.htm

A quick glance through - important to remember we are in the middle of an update so some of this is slanted

He puts to bed all of the mis-information out there about toolbar PR and backlinks - its just as we've been trying to say here all along - dont use it for analysis

He also talks about the best way to ensure that you dont get into a split site problem i.e. with the www.domain.com and domain.com and what to do (which I believe we recommended over here some time back) when you do get in this situation.

Also finally came out and got rid of the myth that Google penalizes for frames and Iframes - just not true - while also mentioning the urban myth about java (like mouseovers) that we use causing some sort of problem - again - not true

most importantly - they are working on updating the WM guidelines to include things like - if youre gonna use pages with &id= with session ids youre just not gonna get crawled very well - if at all and other useful info like that thats been around for a while but was never included in the guidelines.

and lastly he did say that we are about half through the update and in another thread went into a little more detail saying that they almost have all the datacenters synched, but then would start adding in the filters and "quality enhancements" that they always do at the end of an update.

Hope this helps ya'll

Sweet.

I didn't see anything really new, but a very enjoyable read.

I guess the session ID thing was the newest and possibly most valuable, altho I never use dynamic pages myself, so it's mostly just an interesting new point to me.

The stuff on backlinks, backlinks updates and paherank updates was very good to hear confirmed.

I liked the comment on absolute vs relative links. I actually like to use both absolutes and relatives on the same page, so I'm glad to see that my sense of the additional value of some absolutes being confirmed.

Still always worries me when they slap on their 'quality enhancers' because you never know what that red flag is and it seems that often they are way too big chunks of the web getting 'quality assured' and far too many spam pages not :(

It is a good indicator that if you come up well for a few days after a reindex, then disappear, that you might have something on your page that they are filtering.

cd34 - exactly :)

Bill - the absolute vs relative links turned into quite a big issue for some of us back starting around Sept last year and really hitting hard in December - there were a few of us that lost just about everything in the SEs because of them - we all had "vanity domains - ya know the .net,.org spelling var. etc - and because we had relative links - and our vanity domains got botted by Yahoo and listed - Google found them and started splitting the pages of our sites between the different domains - it even happened in the case of some sites where the only difference was the www.domain.com and domain.com and the serverconfig wasnt set correctly to do the apache redirect it was supposed to be doing (the 301 thats built in to the serverconfig wasnt handling the requests right)
Its taken some of us until this update to get things fixed and back in to where we are doing pretty darn well again - and I can tell ya I never want to go through that one again :)

What is the proper way to redirect from site.com to www.site.com?

I hope this update is over with soon, my G traffic has turned into a bad joke :(

Put something like this in your .htaccess file.

RewriteEngine on
RewriteCond %{HTTP_HOST} ^domain\.com$ [NC]
RewriteRule ^(.*) http://www.domain.com/$1 [R=301,L]

cd34 - shouldnt this be taken care of in the server config with the servername set to www.domain.com - so that it always defaults to that?

Although I do agree that the above code for the htaccess works just as well to ensure that you dont have the splitting problem and takes care of the case where Google already knows about both versions :)

Well, the server config only does the redirect on certain conditions:

ServerName www.domain.com
ServerAlias domain.com

When a request for http://domain.com/ comes in, the alias says, serve it if you have it. So, it finds index.html (or index.php whatever) and serves it.

If you have a Directory like http://domain.com/directory/ it will again serve the index.html (or .php/.shtml/whatever)

However, if you have a link like: http://domain.com/directory when the server does the request it says, hey, directory is not a file, redirect it to the server, let me build the name, and redirects to http://(ServerName)/directory/

So, if the request that comes in is properly formed, apache will not redirect to the ServerName specified in the config.

hehe, nothing new at all