I was speaking with a colleague recently who gave me the age-old luddite question: what’s really new with voice mashups? Haven’t we been doing voice applications for years? I’m having a hard time seeing what’s really new here.
Here’s what’s new: the application platform’s changing. The stuff under the application is different, and in nearly every respect. The history of all computing, but in telecommunications in particular, has been dominated by applications written on top of some sort of hardware, with a layer of software on top of it to allow it to access the hardware. If you stop for a second, it’s easy to see my colleague’s point, because at the end of the day, that’s all that could possibly be. You’ve got to have hardware SOMEWHERE, you’ve got to have a piece of software to let you talk to it. The part he’s missing is that in the old world, we are forced to use a particular piece of hardware (the one in that closet over there), or a particular piece of software (the one that my program linked to). In the new world, you access these network resources using web services, and everything about generation of these services is transparent to you. You might not even use the same piece of hardware twice. As you sit down an plug in your laptop, do you know what generator connected to the power grid made the electricity you are using? Even if you could, would you care? Of course not. Do you really believe that it’s going to matter who terminated your phone call in one hundred years? That might be happening already.
You may claim that telephony services aren’t nearly as generic as electricity. I challenge you on that assumption (and I suppose I’ll post about that later), and web services provide the interface to plug into that network.