Aswath - also being a VoIP veteran (no longer an oxymoron I suppose), I can guess where you are going with your three point list, but also suspect the reaction of many people would be how do those three items make my user experience better, cheaper, simpler, and/or more reliable? Perhaps you can expand or point us to an earlier post in which you've already done so?
Posted by gzino at February 11, 2007 07:37 PMAshwath
I agree with you 100%. ATA based architecture has not taken into account enormous potential of underlying VoIP technology and have tried to adapt a shortest route to market.
Indeed there are several features and enhancement possible to enhance and provide best user expereince using little bit of innovative thinking and stepping away from PSTN call flow model.
Posted by aNISH at February 11, 2007 08:03 PMYou're absolutely right.
I'm not sure that the reason service providers haven't pushed for more ATA features is due to cost. I think it's more like service providers (at least most of them) actually want a telco model, a service-provider centric model, and the matter of driving the cost down on the ATA comes out of that (if all we need is a dumb ATA, we want it as cheap as possible). I think you and I have both spent time fighting this battle, trying to show why edge-based (smart CPE) makes more sense, but for the most part it has fallen on deaf ears.
This creates a vicious circle, where manufacturers say "nobody else is asking for that" if you complain to the device maker about bugs or if you ask the vendor to produce devices that do more (something like the ability to "signal to each other directly" for example).
This traditional telco "smart network, stupid devices" world-view permeates the industry, and certainly is loud and strong among the most prominent players.
Posted by David Beckemeyer at February 11, 2007 08:19 PMGil:
The following are some of my thoughts. If the user signaling can be presented to the human user, then it will be possible to inform the called of the Subject of the session. If I know the subject of the session, I can decide a) whether to respond to the session invitation, b) if I could support multiple modes, then I could decide on the preferred mode. If I could change the mode in the middle, then we could add/change the mode. For example, to offset my accent I prefer to have an IM session going when I am on a call. I would think a properly constructed end device can make this happen seamlessly.
PSTN offers two forms of Transfer - blind and consultative. But in VoIP, we could offer different forms of transfer. I could do the consulting phase over an IM, share a document and whatever.
People talk all the time about directory based dialing. How many ATA's offer this capability? Another unfulfilled feature is click to dial. Why should I depend on Google or Skype. Why can't I click on my browser, which communicates to the ATA that in turn establishes a third party call setup?
As a group, we celebrated the breakthrough Apple made on the user interface front on iPhone. But I pitched for the VoIP industry to build a device with rich user interface around the same time Apple started that project. But the VoIP industry was content with the 12+1 button phone because it was focused only on PSTN replacement.
(http://gigaom.com/2004/09/29/open-letter-to-steve-jobs-how-to-heed-the-call-of-voip)