In a recent webinar that I hosted about how vendors can create FaceTime-like services on Android devices, I introduced the different approaches and questions that these vendors need to ask themselves. One of these is the issue of protocol selection:
If you are about to develop a new video calling product/service, you've got too many choices as it is already:
* You can go for 3G-324M on mobile handsets and be compliant with a huge number of devices in Asia and Europe. Alas, the video quality is going to be quite poor with only 64kbps for everything you send.
* You can decide to use H.323, which will immediately give you access to most of the existing enterprise video conferencing market.
* You can select to focus on SIP. Then connecting to IP PBXs or hosted VoIP solutions will become easier.
* XMPP/Jingle is another choice--it will let you use Google as your service provider and give you access to its Gtalk user base.
* You can always go for a proprietary protocol--Skype being the obvious choice as it is the only dominant player doing proprietary stuff.
* There’s H.325 (AMS), the new kid on the block, which is currently being standardized and might soon become a reality.
And now, you can also use Apple's "open standard" for FaceTime. But you actually can't, as connecting to it without being Apple is currently impossible.
And here’s the thing--we don’t need another "open standard", Steve:
* Couldn't you just select one of the existing solutions and just USE it?
* Is it so hard to use SIP for the registration process of FaceTime?
* What's wrong with the current authentication and certificates mechanism of SIP that you had to reinvent it?
* How come you had to push both media channels on the same RTP session? Couldn't you, for once, use the industry standard for communications?
What we DO need now, Steve, is to agree on how to do video calls over our networks. And once we do, we simply need to test it for interoperability. Anything else, and we fall short of the visual communications promise.
Don't believe me? Just check what others are saying and doing:
* GigaOm's latest research briefing about the http://pro.gigaom.com/2010/08/report-videoconferencing-unleashed/enterprise videoconference landscape places interoperability as the biggest barrier to the growth of this market.
* The IMTC's SIP Parity activity group is working on deciding how to utilize the many options of SIP for visual communication
* The UCIF focuses on how communications needs to be done in all of its aspects (hint: standardized, interoperable and modular).
You should note that both the UCIF and the IMTC are not trying to invent a new protocol, but rather to work with existing ones to achieve interoperability.
That's what we need, Steve. Can you help?
