The past several years in enterprise communications are starting to look like a transitional period, from the circuit-switched TDM systems of the past to the true multimedia IP environment into which we're now moving. The era of the IP-PBX was a time of hybrid systems--platforms that used IP as their underlying network and treated voice and video traffic as IP streams, but controlled and delivered these multimedia traffic streams in such a way as to mimic the structures of the old TDM world.
There were very good reasons for doing it this way, chief among them continuity. As I've mentioned before, those of us who were around when voice over IP moved from the consumer/hobbyist space into the enterprise, will recall that the early proponents of enterprise VOIP took great pains to assure enterprise end users that IP networks could deliver reliable voice. At that point in history, reliable telephony was still the benchmark; I can remember Craig Hinkley, then a top networking exec at Bank of America and an early VOIP adopter, using his 2005 VoiceCon keynote platform to pledge allegiance to what he called the "god-given right" of enterprise users to get...dial tone.
Less than a decade later, dial-tone is in fact still a business requirement for many if not most enterprises. Yet the idea that it's a "god-given right" to be able to sit at your desk and pick up a C-shaped chunk of plastic connected by a pigtail cord to another, wedge-shaped, chunk of plastic--well, that idea just doesn't register like it used to. Sure, whatever enterprise system you rely on has to work--but the idea that such systems must include a desktop telephone just isn't operative any more. The very idea of "dial tone" is antiquated--cell phones don't make that noise, and neither does Skype. Probably half the kids in America, if they accidentally picked up a landline phone and heard that noise coming out, they'd think something was wrong with it.
And yes, I understand that Craig was using the term "dial tone" in the abstract, to refer not to the noise itself, but to the guaranteed service availability that the sound announced to the user. But guess what? The concept of "five-nines" is an aspiration, not an assumption for current-generation communications networks.
There was a profound sense at Enterprise Connect Orlando last month that things are changing, that the early adopters at least are getting ready to leave this transitional phase behind us. Whether it's enterprise execs hoping not to buy any more PBXs, or the rise of WebRTC and software-based systems, the old verities really do seem to be losing their hold on the industry--starting with end users and their insistence on driving change.
At least some folks who make a living designing networks are likewise rethinking old certainties. One of the pleasant surprises at EC Orlando for me was getting to re-connect with Sorell Slaymaker, a longtime friend of the show and of the No Jitter site. Sorell followed up our chat by sending along a provocative blog post that questions some of the basic assumptions about how to run IP networks that carry multimedia traffic.
You should go read the whole post; Sorell challenges some of the traditional wisdom about classifying traffic, provisioning WAN links, and more. His argument is that the economics as well as the technology has changed, and that efficient deployment of enterprise resources may require new ways of thinking about best practices. The subject is worthy of discussion and debate, and the Comment section at the end of Sorell's piece is there for just that purpose.
This is one of many new conversations that I think we'll be starting in the industry.
Follow Eric Krapf and No Jitter on Twitter and Google+!
@nojitter
Eric Krapf on Google+
