[onreact-com]

You are doing mobile SEO the wrong way. The industry best practice is to use one page with different stylesheets. Not a separate page, site or subdomain like Digg does.

You might want to check out a list of mobile SEO resources I compiled last week for a start:

http://www.seoptimise.com/blog/2009/07/33-mobile-seo-mobile-...

[Edit:] Also "blocking search engine crawlers from seeing the mobile view" will result in your mobile pages removed completely, from the mobile index as well.

[karanbhangui]

While the CSS spec was designed to keep in mind various media types, the industry standard is indeed a 'm' subdomain for the mobile version of the site. I think this need has arisen because the use of HTML/CSS has gone well beyond minimalist content markup of documents, to full blown GUI websites (i.e. sites with complex navigation bars), which doesn't exactly transfer well onto mobile. I think this is why only the simplest of sites (i.e. blogs, text search engine results) will be able to display an alternate stylesheet for acceptable mobile presentation.

More complex sites would need a more tuned site (i.e. m.wikipedia.org - notice how the edit stuff is cut out for example)

ps- i hope i've understood your contention correctly :D

[onreact-com]

OK, it obviously depends on which industry we actually speak of. The overall Web industry tends to do it wrong like in the example above. Creating two sites for the same content is fundamentally flawed.

From a SEO perspective we always try to avoid duplicate content issues and splitting up of your site authority. I'd recommend the current issue of the Search Marketing Standard Magazine (disclosure: I'm an affiliate of it) which features an interview with mobile SEO Cindy Krum. She explains the issue with separate sites/pages quite well.

On the other hand I have witnessed the negative impact of such subdomain based mobile SEO already years ago. It never quite worked with Google.

[ig0rskee]

Well, it's pretty much a standard:

cnn.com & m.cnn.com

facebook.com & m.facebook.com

gmail.com & m.gmail.com

So definitely disagree with the notion of "fundamentally flawed" as literally every significant site is doing it at this point

[onreact-com]

Yeah, but do we need Gmail or Facebook crawled? No, both are private. They don't get indexed at all.

The CNN site is quite messed up. Check this:

"2 U.S. Marines killed in Afghanistan" site:cnn.com

You have duplicate content twice + the mobile version does not even contain the article when you click on the search snippet.