New Canonical Tag from the big 3

[kgun]

Originally Posted by Clint1

"If it ain't broke, don't fix it".

You can't even fix a W3C invalid site.

Errors found while checking this document as HTML 4.01 Transitional!

16 Errors, 8 warning(s)

All on the markup with the Google map link.

of Kongsfjordmat.no After changing & to " on the Google map link:

"This document was successfully checked as HTML 4.01 Transitional!"

Problem, the map is then over USA and not over Italy.

So I have to live with that invalid link

Google standard is Google standard, ...

They require more of their clients than of themselves.

[wige]

Wouldn't changing the "&" to "&" correct the errors and still let the link work?

[kgun]

No, I changed & to &

The site validated, but the link did not work as intended and I had to change it back to the invalid markup.

I have experienced this before, with other Google and affiliate link markup.

That is the reason that I don't change affiliated links. If you change affiliate links to valid markup, you can not know whether the link tracks correctly.

It should be easy for you to test it. It should not be difficult to locate the link on the site.

I use this validator (first hit on Google for that word).

[Webnauts]

Originally Posted by wige

If the destination page has a noindex tag, that has no impact on the distribution of pagerank at all, because Google still follows the links, and thus must calculate outgoing pagerank. (See the above Matt Cutts quote on this issue.)

I am back. And this time with some updates.

I have link on page A http:// www. seoworkers. com/goforit.html?bots=nocrawl
The destination page B has outbound links to other site pages or sites. So it is not a dead end.

In the robots.txt I have this:

User-agent: Googlebot
Noindex: *bots=nocrawl

Tested already several times, and it is sure that destination page B does not get indexed.
And even if there is an IBL from other internal pages or external sites, it does not get indexed.

If I would disallow Googlebot to access the destination page B for example like

User-agent: Googlebot
Disallow: *bots=nocrawl

I would have created a dead end, since I would block Google accessing the destination page, as we discussed above.

That would nothing different as blocking pages with the "Disallow" directive in the robots.txt. Don't we create dead end (dangling) pages that way?

Therefore, I strongly believe that is the best alternative to avoid dead end (dangling) pages, if I can make sure that the destination pages have at least one outbound link.

Don't you think?

---
Notice: Googlebot supports the robots.txt directive "Noindex" since a long time, it is just not official.
I updated my robots.txt. If you look at it carefully, and you agree with the above, you will find it very interesting: http://www.seoworkers.com/robots.txt

I also updated the robots.txt of a customer with his permission to test, which is even more interesting to look at: http://www.gameshop.gr/robots.txt

[Webnauts]

I want to add to the above post the following, to be more specific:

I have link on page A http:// www. seoworkers. com/goforit.html
The destination page B has outbound links to other site pages or sites. So it is not a dead end.

If I would disallow Googlebot to access the destination page B for example like

User-agent: Googlebot
Disallow: /goforit.html

... I would have created a dead end, since I would block Google accessing the destination page, as we discussed above.

I went more into more details at another forum: http://www.seoers.org/BB/general-sea...9483/#msg59483

[Clint1]

Originally Posted by Clint1
"If it ain't broke, don't fix it".

Originally Posted by kgun

You can't even fix a W3C invalid site.

I was only speaking in the context of that post of mine.
New Canonical Tag from the big 3

Errors found while checking this document as HTML 4.01 Transitional!

16 Errors, 8 warning(s)

All on the markup with the Google map link.

of Kongsfjordmat.no After changing & to " on the Google map link:

"This document was successfully checked as HTML 4.01 Transitional!"

Problem, the map is then over USA and not over Italy.

So I have to live with that invalid link

Google standard is Google standard, ...

Well that can be confusing.

They require more of their clients than of themselves.

Yeah that's for sure. Their own pages don't even validate.

[Webnauts]

I finally had time to express my thoughts in a manner that everyone should understand: Bot Herding: The Ultimate Tool for PageRank Sculpting

[Webnauts]

Originally Posted by inertia

What are your opinions on this Rand article: SEOmoz | 12 Easy Mistakes that Plague Newcomers to the SEO Field?

Specifically these points...

Could it be that Rand changed his mind about point 7?

7. Blocking Bots Rather than Using Nofollow
I think there's been some confusion about how PageRank sculpting works. You should NOT block bots access to pages you don't want to send link juice to - in fact, this doesn't even accomplish that goal. If a link points to a page, even if that page is blocked via robots.txt or meta robots noindex, it still accrues link value metrics. The only way to stop a link from passing juice is to use a nofollow (or to make it via an external Javascript redirect or embed in a non-HTML-parseable object). Please be careful about what you noindex or block to bots and why - you don't just hurt that page, you can hurt downstream areas of your site by walling off navigation paths too (and even stop link juice that flows in from flowing out).

"Please be careful about what you noindex ... (and even stop link juice that flows in from flowing out)".

That is the best joke I ever heard, since I am involved in the SEO industry. And I just added a comment there for further understanding and/or perception: SEOmoz | 12 Easy Mistakes That Plague Newcomers to the SEO Field

- Look at the date of my comment here SEOmoz | PageRank Sculpting: Parsing the Value and Potential Benefits of Sculpting PR with Nofollow .

- Then look please at the date I posted the article at SEJ Bot Herding: The Ultimate Tool for PageRank Sculpting .

- Then look when Rand posted this SEOmoz | Headsmacking Tip #13: Don't Accidentally Block Link Juice with Robots.txt which it is not in its original edition, since his claims were wrong and he have updated after the comments of Andy Beard.

With all respect ... better too late than never Rand.

Just to make things clear here.

[kgun]

John did you note the related video by Matt Cutts.

YouTube - Matt Cutts Discusses Webmaster Tools

If it don't function on YouTube (like her today) it functions on my site:

Global resources for webmarketing, branding and digital ad at AdSchoolworld.com

I still have a question about your opinion on the rel="nofollow" attribute. I have noted that W3Schools have put that property on affiliated links.

I would personally not do it for the following reasons:

1. You may end up giving away free ad on your site.
2. You don't know if the link tracks clicks correctly if you click on the ad link. As explained in post #151 - #153 above, I have noted that changing & to & may change the target and functionality of the link.
3. I think it should not be difficult for SeBOTs (at least the big three) to identify affiliate links.

Any opinion?

[Webnauts]

Originally Posted by kgun

John did you note the related video by Matt Cutts.

YouTube - Matt Cutts Discusses Webmaster Tools

If it don't function on YouTube (like her today) it functions on my site:

Global resources for webmarketing, branding and digital ad at AdSchoolworld.com

I have two points I would love to ask Matt. I will start a new thread about this video and I will add the link at Youtube.

1. He said that one of the best solutions (with a smiley) is to protect areas via .htaccess password protected area.

My concerns:

The IBM patent explores dangling nodes in depth, and provides a list of pages that may be treated by search engines as dangling nodes. And one of them is:

- If the page requires authentication.

What happens when a without a nofollow attribute or masked client/server side link is pointing to the protected area directory?

2. He mentioned the use of the noindex meta tag, and that is not the best solution. Damn... am I dreaming awake here?

My concerns:

a. Why didn't he mention the noindex robots.txt directive? Because it is unofficially supported?

b. If I disallow through my robots.txt a protected area with a htpasswd via .htaccess, I have a dangling/node issue, or?

Wouldn't noindex via robots.txt be better alternative? I am still thinking about it...

[Webnauts]

Originally Posted by kgun

John did you note the related video by Matt Cutts.

YouTube - Matt Cutts Discusses Webmaster Tools

If it don't function on YouTube (like her today) it functions on my site:

Global resources for webmarketing, branding and digital ad at AdSchoolworld.com

I still have a question about your opinion on the rel="nofollow" attribute. I have noted that W3Schools have put that property on affiliated links.

I would personally not do it for the following reasons:

1. You may end up giving away free ad on your site.
2. You don't know if the link tracks clicks correctly if you click on the ad link. As explained in post #151 - #153 above, I have noted that changing & to & may change the target and functionality of the link.
3. I think it should not be difficult for SeBOTs (at least the big three) to identify affiliate links.

Any opinion?

I was strictly against the nofollow attribute until I figured out some days ago that it can be useful if you are in a situation where technically is your only choice.

But how do I implement the nofollow attribute. Just rel="nofollow" and end of story? No!

Lets say you are have a masked link on your site page A (attributed with the nofollow attribute) with destination an external web site.

For example:

My site has a server side (i.e PHP) masked 301 redirect hyperlink like this <a href="http://www.seoworkers.com/redirect/adschoolworld" rel="nofollow"> which is pointing to adschoolworld .com/affiliate.php?id=1024

Then somehow Google discovered that masked link elsewhere without a "nofollow" attribute. What will happen? Googlebot, Yahoo and MSN will return the reference without a snippet. Just a URL. Remember what Matt said in the video you posted? I know that already since years.

So I came up with a 100% perfect solution for Googlebot.

I simply add a Noindex directive in my robots txt this:

User-agent: Googlebot
Noindex: *bot=noindex

and then I change the above redirect link to this:

<a href="http://www.seoworkers.com/redirect/adschoolworld?bot=noindex" rel="nofollow">

For Yahoo and MSN I add a disallow directive which I did not check if it realy works yet:

User-agent: Slurp
Disallow: *bot=noindex

User-agent: MSN
Disallow: *bot=noindex

Now that you have noted that changing & to &amp; may change the target and functionality of the link, don't you think it can be masked or redirected via .htaccess mod_rewrite or server side scripting?

I am sure that the big bots can identify affiliate links. There I have no doubts at all.

I did not understand what you met with giving away free links.

[kgun]

Originally Posted by Webnauts

I am sure that the big bots can identify affiliate links. There I have no doubts at all.

Hopefully yes, so you don't have to care about affiliated links through known and serious providers.

Originally Posted by Webnauts

I did not understand what you met with giving away free links.

I simply meant:

1. Suppose you sign up with an affiliate provider that don't have an option to put the rel="nofollow" attribute on the link.
2. Then you change that link and add the rel="nofollow" attribute on the link yourself.
3. You change the link semantic and it may not track correctly any longer. In the worst case no clicks are registered on your account or they are registered incorrectly. You don't get commission and give away free ad (links) to the ad merchant.
4. This has been my concern since I first mentioned that DreamWeaver 2004
   solved this in this way:
   <div id="advert">
   
   Ad link here
   
   </div>
   
   IMO a much better solution, since the link is not changed. And as you know, in programming the devil is in the details. You may know the effect of a comma error. And it may vary from SE to SE.
5. But you have answered the question. No need to worry about affiliated links. Keep them intact. The SeBOTS are able to identify them and handle them in the proper way.

John, this is and advanced topic. You introduce a new dimension with redirects.

[Webnauts]

Create a PHP file, calling it for example index.php, and edit there the following code:

PHP Code:

<?php
&lt;?php
$link = "";
$m = $_GET['m'];
$so = $_GET['so'];
$st = urlencode($_GET['st']);
$valid = "false";

if ($m == "") {$link = "http://www.kgun.com/";$valid="true";} // Homepage if no link given

//Sites
if ($m == "affiliate") {$link = "http://www.affiliate.com/example.html?id=kgun";$valid="true";}

if ($valid == "true") {
  header("HTTP/1.1 301 Moved Permanently");
  header("Location: ".$link);
  header("Connection: close");}
else {
  header('HTTP/1.0 404 Not Found');
  include("/var/www/vhosts/kgun.com/httpdocs/errors/404.html");
}
exit();
?&gt;
?>

Then create a folder and call it for example "redirects" and add in there the PHP file.

Then add an .htaccess in the redirect folder editing the following lines:

RewriteEngine On
RewriteBase /
Options +FollowSymLinks

### Prevent directory bowsing ###
Options -Indexes

RewriteRule ^m/(.*)$ /redirect/index.php?m=$1

Then add in your robots.txt the line:

Disallow: *bots=nocrawl

Then use as an affiliate link this:

http:// www. kgun. com/redirect/m/affiliate?bots=nocrawl

How about that?