This post was originally going to be a response to my friend Stevens post on the same topic "Isn't it odd.." (Steven Westwell on social networking) so I'd read that first in case I've missed something in the overwrite to a post here.  His response turned in to an essay, and so has my own.

I'm eternally divided on the topic of social and collaborative software and the workplace, and even in leisure time.

The experience people currently have with social networking is really zeitgeist and faddish behaviour, people check their facebook over personal email at the moment because it's new and exciting, it's something they've not had previously, though month after month since about November 2007 facebook and to an even greater extent myspace have seen a steady reduction in both traffic and user activity.  I'm convinced that today's social networking, from the walled garden of the MSN network, sorry, Facebook ( ;) ), to the focusless indirection of twitter aren't actually setting the mould for community communication, they're a "web two point oh-ey" version of phpBB and usenet more than anything.

I'd actually say spam currently poses a greater threat to traditional email than social networking could possibly manage.

That said, it's correct to identify user driven communities as the direction things are and should be moving, but I see the success in this is directly linked to the maturity of unified online identities, rather than scattered ecosystems.  It'll take a killer app in the form of Windows Live (even based on the success of XboxLive), or OpenId (I prey for the latter but I'm reasonably certain it'll be the former) to act as the overall identity broker before people start to establish solid content driven identities on the Internet.

The biggest barrier obstructing the path of social networking becoming as ubiquitous as email is the walled gardens it builds around itself.

The Mecca of the "web 2.0" concept really is distributed social networking, driven by network events to publish stories, events, "friend requests" and other unspecified interactions (RRS and Atom feeds serving as the first implementation of this "half-push" technology).  A user should control their own identity, have the right to elect the authority who controls their identity (see OpenId), or control their own.  Until the social networking sites stop trying to wall in their users (like AOL, ICQ, MSN, Yahoo, local BBS', The Microsoft Network, Facebook, Myspace, etc before them) we'll spend a few years in the wilderness of disparate online identities and large communities achieving moderate success before vanishing.

Social networking platforms need to concentrate on adding value to the user experience, rather than trying to consistently reinvent the wheel.  If they were software project propositions this kind of behaviour alone would doom them to failure.  Innovation should be celebrated, and developers certainly shouldn't have to produce "yet another user management system" with all the features of their direct predecessors, just a little better, or a little more shiny.  The bread and butter of social interaction should be standardised protocols for communication between unrelated innovative "value added" sites.

The kind of interactions suggested, distributed file storage, notifications and calendars exist in products like SharePoint, and mediawiki, iCal and traditional email, but until they're tied together correctly they're doomed to be noise.  I really believe that until collaboration technology matures to be platform independent that it'll struggle to permanently compete and shape networking on a broader scale.  I work entirely in the Microsoft ecosystem and I still feel it's very important that Microsoft should not tell you who you are online, nor should Facebook, Blogger nor Apple.

Interestingly there's a different set of challenges facing the adoption of technologies like instant messaging in the workplace, and they're all entirely human.  Staff training and trust with communication mediums that can be painted as time wasting will always struggle to garner acceptance.  That said, I've used instant messaging for years in the workplace, but I'd imagine most technical companies will be ahead of the curve in this respect.  It's interesting if you look at studies of how wasteful people make checking email on a daily basis, after you account for disruption in flow and response times.  It's actually (at least in my experience) far more efficient to have email delivered on a pull basis, in intervals measured in hours.  People get more done when they only check their mail four times a day.

I'm really really excited about the future of social networking and its place as the main content driver of "web 3.0", but I like to imagine it more in the vein of the Ainsible networks of Orson Scott Cards "Enders Game" than the cheap message board hacks of myspace and their application ecosystem.  I'm striving to unify my online identity and I'd implore anyone else to do the same.