I've come across what appears to be very poor behavior in regards to SEO with the way newer versions of DNN are handling non-existent child pages.

As example: http://domain.com/about

Navigating to that domain would result in seeing the content for the about.aspx page.

However, it appears that navigating to any of the following will also result in seeing the content for the about.aspx page, but listed at different urls... Note: None of these are actual pages in DNN.

http://domain.com/about/company
http://domain.com/about/xyz
http://domain.com/about/about/about/about/about/about
http://domain.com/about/stuff/and/things

Correct me if I'm wrong, but won't google see this as duplicate content?  It will see each one of those pages as having the exact same content.

I realize that the likelihood of the bogus url's being indexed is slim to none, but what is to stop someone from creating url's on his or her website to point to http://domain.com/about/derogatory-terms-about-this-company/?

I would think this could have messy seo ramifications...

You can test this on dnnsoftware.com by going to the following:
http://www.dnnsoftware.com/about/why-this-company-is-the-best-company-in-the-world

To solve the problems you mention, i develop openurlrewriter (available on codeplex)

In enhanced 404 mode, only the know valid url are rewriten.

This make it possible to return 404 statut for all url's corresponding to non existing pages (or modules).

Valid url's are url corresponding to pages or generated by open url providers which exist for different modules.

http://openurlrewriter.codeplex.com/

Forge projects : Open Url Rewriter - Open Content - Open Form

My opinion regarding 404 management and SEO, is there was a mistake that was make a long time ago in DNN.

The mistake was to rewrite all urls (query parameters). This has generate a situation that make it impossible to DNN to know if a incoming request correspond to a real url or not.

In OpenUrlRewriter, i have give the possibility to not rewrite all urls, and rewrite only urls corresponding to pages and urls corresponding to modules with a provider.

It that way it is very easy to return 404 for non existing pages.

The difference is that all urls for modules without provider generates urls with query parameters www.mysite.com/mypage?itemid=123

But in therms of human friendly, a url like www.mysite.com/mypage/itemid/123 is not really more human friieldly in my eyes.

And for SEO Google gives you the opportunity in the Google Webmaster tools to tell if the parameter itemid modify the content (to tell google if it must be indexed as a different page or not).  And by default Google find it automatically.

This can maybe a alternative path for the core rewriter...

Regards,

Forge projects : Open Url Rewriter - Open Content - Open Form