Will WebRTC Replace SIP?

No.

But WebRTC will definitely marginalize the importance of SIP.

I just had to write this post. Got the “idea” to it from Andrew Prokop, who wrote a post with the same title a while back on NoJitter – Will WebRTC Replace SIP:

For me, the ultimate answer to the question of one over the other comes down to this: Do you like apples more than oranges? While both are fruit, they are very different and each serves a role that the other cannot fill.

While I tend to agree with this statement by Andrew, I think there’s a missing point here somewhere.

SIP does signaling. And also defines how media gets handled.

WebRTC does media.

But WebRTC doesn’t define how signaling is handled. Nor does it care. There are three aspects of WebRTC that end up marginalizing the importance of SIP:

1. WebRTC is all about “dumbing down” communications – making it accessible to a lot more developers than just us VoIP engineers. And if you aren’t a VoIP engineer, why the hell would you want to learn something as complex as SIP? You’ll just go and roll your own solution that fits your own exact niche use case
2. WebRTC is about embedding communications – changing it from a service into a feature of another service. That other service? It already has logic of its own and a type of signaling of its own. It may be a scheduling system, a messaging system, a dating site logic – whatever. What would be easier? Adding a bit of WebRTC signaling into the existing logic or trying to push a full-fledged SIP infrastructure somewhere in there?
3. WebRTC is about killing federations – As Dean Bubley puts it nicely in a recent presentation (slide 14): “Enterprise comms becoming dis-unified” – WebRTC is enticing a silo approach to services. You need comms? Just plug WebRTC in and you’re done. No need to think about interworking with others, connecting or federating with more networks – just pure comms inside your app/service. The funny thing is, that even those trying to federate WebRTC, do so with proprietary protocols (check out OpenPeer and Matrix)

All that, and still SIP will endure.

SIP is the best way for WebRTC today to connect into legacy networks – be it enterprise VoIP networks or PSTN and cellular services out there. Some do it by shoving a SIP stack inside the browser along with WebRTC for the ride, while others do it by placing a gateway between the browser and their SIP infrastructure. Whatever the decision is, there’s room for SIP in the world of WebRTC. So no. WebRTC will not replace or kill SIP.

But will you be using SIP for your next vinyl project? Most probably not.