Daily Payload

SIP: The Never-Ending Hype Wagon

June 18, 2007

We ran across an article that exemplifies the problem we see with the continued hype surrounding SIP. Day after day and year after year, we continue to hear people talk about SIP as an "emerging" protocol that is web-centric, simple, and posed to take the industry by storm. Somebody, please give me a pair of wading boots. Planets and life emerge, but are we creating a protocol for the next species that occupies this planet?

Trying to be more objective reporters, we sometimes step in and comment on such ridiculous misstatements by companies that are clearly trying to position a technology because—well, it's what they have to offer. Technology is ever-evolving and, in the case of SIP, it has been evolving for 11 years now. The bottom line is that SIP fails to meet market requirements, has been very slow to progress, and alternatives are already finding their way into the market. Skype has more traffic than all of the SIP islands combined. The IAX2 protocol from Digium was created because there are clear shortcomings with SIP. Lastly, the ITU SG16 is now collecting requirements for a new, forward-looking multimedia system called H.325 that will focus on enabling communication between any kind of device and any kind of application, taking a big step forward in terms of defining what "multimedia communication" has historically meant.

Aside from new protocols and networks that have entered the market, a number of companies are utilizing proprietary protocols within products in order to provide VoIP functionality right from within web pages. In fact, if one needs VoIP functionality to operate in a client/server environment like the web, what is the reason for using SIP when one controls both the client and the server application? There is really very little reason, especially if the signaling can be converted into something on the backend when calls are directed around an enterprise.

Today, H.323 commands about 80% of the international long distance VoIP minutes, which is an estimated 15.8% or 49.4B minutes of all international voice communication. Even Skype, perhaps the single-largest VoIP network, has not caught up with H.323 in terms of minutes of traffic. (Skype had 13.8B minutes of VoIP traffic in 2006 and held a 4.4% share of the international voice business.)

Where does this leave SIP? People are still working hard to try to make SIP a success story, even after trying for so many years and gaining so little market share. Standards bodies including the ITU SG11 and ETSI TISPAN are trying to position SIP as the NGN protocol and a replacement for the PSTN systems already deployed. But perhaps what is not so well understood in the industry is that the carriers are not likely to provide new, innovative services on top of SIP. What would be the financial motivator? Will users pay for new kinds of services and capabilities? What would those services and capabilities be? Can SIP provide them and would SIP be the right choice? So far, nothing new or innovative has been delivered and as financial reality sets in, you will likely see that little actually changes.

SIP is problematic. It's not simple. Nearly every vendor has a different interpretation of the protocol, leading some to jokingly call it the "Subject to Interpretation Protocol." SIP might be a success one day, but do not expect to get more than basic voice services anytime soon. Perhaps you might be able to get video conferencing functionality, but it will not be any better than what you can get with H.323 today.

The industry should really pause and ask itself: if we are to continue to invest in multimedia communication, should we not try to reach much further than the empty promises of SIP?