Robert_Charlton - 7:35 am on Jan 26, 2013 (gmt 0)

What they did do, is put links on every page of the site saying essentially, "view mobile site" (on the standard web version) and "view traditional site" on the mobile pages. For whatever reason, they used a tracking link (ie, /?source=mobile) to link back to the main site from the view traditional link.

LuckyLiz - The links back and forth between versions and the ?source parameter are often elements used when you have a mobile site on an "m" subdomain. Generally, the two-site approach is used in conjunction with user agent-based redirection (browser sniffing) to decide on the most appropriate version for the user, depending upon the size of the viewing device. It's got some advantages as well as disadvantages compared to the responsive design approach. Eg, if the desktop version of the site is too elaborate or might require too much bandwidth, a responsive design site might end up being slow on a mobile device. There are a lot of variations in how sites can be adapted to different sized screens.

The links between "Mobile Site" and "Desktop Site" are to allow the user to navigate back and forth between versions. If the user doesn't like the version he/she's on, say the mobile version, the ?source parameter prevents automatic redirection back to the mobile version after the "Desktop Version" link is used... and vice versa.

Do we need a no follow tag or something on the links, in addition to setting up those rel canonical and rel alternate tags on the appropriate places?

I haven't been involved with the situation you're in, but offhand I would not use a nofollow tag. I think Google will sort it out much better by spidering both versions.

I would follow Google's recommendation regarding the appropriate use of the rel canonical and rel alternate tags. Also from the Google developers page...

This two-way ("bidirectional") annotation helps Googlebot discover your content and helps our algorithms understand the relationship between your desktop and mobile pages and treat them accordingly. When you use different URLs to serve the same content in different formats, the annotation tells Google's algorithms that those two URLs have equivalent content and should be treated as one entity instead of two entities. If they are treated separately, both desktop and mobile URLs are shown in desktop search results, and their positions may be lower than they would otherwise be.

This assumes, btw, that the content is roughly equivalent. I'm not sure how to treat situations where content has been dropped from the mobile version... but I'm guessing that Google would figure things out better with these tags than without them.

Here's an excellent article on choices to be made among interface types and device destinations, and it covers some aspects of what your developer may have done.

The article also contains much food for thought about factors that might affect what type of mobile site you build....

Do Mobile And Desktop Interfaces Belong Together?
Smashing Magazine - Mobile
July 19th, 2012
[mobile.smashingmagazine.com...]

My emphasis added...

...This article looks at how a typical responsive website is targeted to mobile handsets and tablets, contrasting it with its desktop-facing sister website. It considers how such a website might also be deployed as a cached HTML5 Web app, and why you might want to do that instead. For various reasons, discussed here, you might decide to target your responsive design more narrowly as a hybrid for smartphones and tablets, keeping it separate from the desktop interface....