VoIP Interop: built-in or bolt-on?

Every couple of years, I wonder when voice/data integration envisioned in this 1994 BYTE cover story will finally materialize. Two years ago, I was trying out the Vonage system. This comment jumped out at me:

I spoke with the Vonage folks today about the prospects for getting direct access to the digitized voice or the SIP signalling. It's possible in theory, but in practice the packet headers are encrypted (a nice security precaution), and they haven't documented how to reproduce the decryption and packet assembly that currently happens inside the ATA 186. Right now, they're focused on building out a basic voice service. But they absolutely agree that lots of useful integration can, and eventually will, be done. [Jon Udell: Bellheads versus Netheads]
It caught my eye because lately I've been investigating Skype, which turns out to be far more proprietary than Vonage. It's true that massive adoption, such as Skype is now experiencing, has a way of trumping standards. It's also true that a viable voice-call alternative "that just works" is a necessary precursor to the new breed of voice-and-data applications that we need and want. But if cheaper calls aren't the endgame, and if it's the apps that ultimately matter most, then shouldn't the mechanisms for creating those apps be built in rather than bolted on?

Former URL: http://weblog.infoworld.com/udell/2004/09/15.html#a1076