By Trevor Baca, VP Software Engineering.

During my upcoming presentation at ETel, Voice and the Web: The New Terrain, I’ll be examining how the global telephone network evolved from a completely closed system to where we’re headed when the global telephone network finally becomes available to applications developers everywhere.

In part 1 of this post, I examined by much of the 2.0 hoopla isn't about voice.

Part of it may be that voice doesn't actually make sense in all the social contexts that we, as telecom innovators, might hope. Maybe flickr is a case in point. If browsing the world's photos means that we're looking mostly at photos taken by people we've never met from time zones away, maybe voice just isn't the right way to reach out and make an introduction.

Another part of the answer may be fear of integration. Up until very recently, if you found yourself hosting a well-trafficked site with a large user base, it wasn't at all clear how you could offer up voice to your users, even if you wanted to. This may be what's going on with myspace. The obvious interaction guffaws and the site are the stuff of legend in the usability community. That could point to any number of things, of course, but one likely culprit may just be the risk of integrating anything at all with site growth that rapid.

And then there may be a genuine lack of interest on the part of the some of the most successful of the social sites. I can't be certain, but I suspect this to be the case with craiglist. Perhaps Craig himself doesn't care. Or perhaps nobody's approached him. Or perhaps it simply isn't clear enough yet that voice is a genuine possibility on the Web.

Where voice simply isn't the right tool for the job -- flickr, perhaps -- then we can stop asking questions. But where voice is simply disadvantaged -- either through the lack of interest, or integrations difficulties -- we owe it to ourselves to look past these proximal causes and go at least one layer deeper.

Consider, to start, that voice has, at least traditionally, cost money. The public network didn't come about for free. Then compare those decades of centralized state planning and control versus the free, drop-in Web components -- think shopping carts and comment boards -- as well as Google Maps and Feedburner-type web services, and it's easy to see why voice may not be the first thing that springs to the minds of talented Web developers everywhere. VoIP may, of course, turn "costs money" into a type of "free", but then we run into the fact, whatever the outcome of the religious war as to the uptake of VoIP handsets, what users really love is wireless. Which puts us squarely back in the "costs money" domain of the PSTN.

Of course, costing money isn't the end of successful innovation. But it probably doesn't help that the Web has evolved as an almost exclusively transaction-driven economy. Click here. For a search, for an API call, for an image, an article or a book. Doesn't matter what it is -- on the web it's the outcome of a mostly stateless, mostly timeless transaction.

But voice? Voice has always been about minutes, unlimited local calling, nights or weekends notwithstanding. It might help us to ask how we can turn voice into the type of billing the Web expects. That is, a billed transaction rather than a bunch of minutes.

And, last -- and probably most fruitfully -- we can tackle the question of integration and just how hard it is to use us. Just how hard should it be for a Web developer to start or stop a telephone call? That’s something we’ve been tackling at Jaduka. And I’ll be talking more about our API at ETel which will give developers direct access to, and all the inherent benefits of, the world’s highest-quality, ubiquitous public switched telephone network (PSTN).

Client-side installs are a barrier to innovation, not a help. Web developers and users alike hate Flash and Java downloads, and it seems unlikely Skype will change these feelings in any significant way. So why shouldn't control of voice on the Web look and act just like everything else on the Web, that is, as a web service?

So I ask again: how did social computing get so social without voice? Maybe part of the social Web doesn't need us. But clearly other parts of the social Web will. Whatever the case, it will be up to us to package our services, and to bill for them, in ways that Web developers everywhere understand, appreciate and will explore.