« Looking for new network service ideas for FWD | Main | May 2004 was a Blur... »
May 31, 2004
A Better Way to SIP?
It looks like the Détente in the protocol wars may be at risk. The battle is poised to start in the Fall of 2004.
In recent months, the advent of home NATs and Firewalls coupled with the success of Skype is putting a lot of pressure on company CTOs to find solutions that “just work.” For those who are deploying SIP based solutions, in practical terms this is a pretty big challenge. In the era of the Session Boarder Controller, the whispers are growing louder each day for an alternative scalable Skype-like protocol that can be used to deliver real-time communication services to the masses.
The SIP protocol has made great strides as a solution to help make real-time communications over the internet happen. And SIP is happening at breakneck speeds around the world. But for a growing group of people trying to make money from SIP today, not fast enough.
Delivering real-time IP Communications isn’t easy. Many home networks are hostile to real-time communications and are not optimized for voice. The need for technical knowledge for those using home networks is a real limiting factor in getting standards based solutions in place. For the most part, this is not an environment that is “plug and play” since home owners are using a variety of products and edge devices. This in turn has become an issue for anyone looking to do large scale rollouts. For some, this is the way to spell “nightmare” from the perspective of network operations and support.
If and/or when IP Version 6 happens, many of these challenges just go away. But how long do we have to wait for IP Version 6 to just happen?
Posted by jeff on May 31, 2004 07:20 AM | Permalink
Additional resources: #140conf events | Watch the Jeff Pulver Show | Jeff's Qik Videos
Comments
I tried a Grandstream's Budgetone and it doesn't support IPv6. I think IPv6 is the way to solute many of SIP's problems, is PulverInnovations selling SIP phones with IPv6 support? Are there softphones with IPv6 support out there? It seems really hard to find.
Posted by: Vincent Brousseau at June 6, 2004 11:01 AM
I wanted to comment about WiSIP and it's potential use for aeronautical and maritime service, specifically the test call on FWD using Boeing's Connexion. It sounds like a fantastic application, however from what I read, Boeing using geosync satellites. I wonder if it's practical given the delay?
Posted by: Jim Stefano at June 3, 2004 11:48 AM
Jeff, I would love to agree with David... but I can't. I am currently working with several service providers to implement end-user connections using IAX2, the Inter-Asterisk Exchange protocol. Much like the Skype protocol, IAX2 uses a single port and cuts throught NATs like warm butter. Currently the only applications/devices that support IAX2 are Asterisk, a number of soft phones, and Digium's new IAXy. More devices, including embedded Asterisk systems and an IAX2 hard phone are in the works. And unlike Skype, IAX2 is open, and an opensource implementation exists.
While I hate to fragment the market, things that work 'ok' now are better than things that may work 'better' in the future. IAX isn't perfect, but it works. Perhaps David will correct my heretical ways at AstriCon this fall...
Steve
Posted by: Steven Sokol at June 3, 2004 12:44 AM
When will IP v6 "Just happen"?
Hmm, considering it will be the single largest rolling software upgrade ever undertaken on the planet across multiple platforms (UNIX, Mac, PC), well....
Posted by: Doug Mohney at June 1, 2004 12:02 PM
David - I agree with you. The long term for VoIP's future is in open-standards protocols, and what is critical over the next 18 months/2 years is to see the SIP/STUN protocols honed to a level where a true plug & play useability can be reached. SIP is indeed gaining momentum, but that momentum must be capitalised on for Joe User on the street to be able to utilise the technology.
Posted by: VoIP User at June 1, 2004 08:31 AM
Jeff, I've seen a bit of this talk about a 'new simpler SIP' and I think it's dangerous and represents a big threat to the progress of VoIP. We don't need more fragmentation and confusion as a result of a dozen different protocols. We need better implementations of the protocols we have, including SIP and STUN. SIP is gaining momentum and we should not get in the way by giving vendors mixed messages. SIP is the way, the one true way. :-)
Posted by: David Beckemeyer at June 1, 2004 01:22 AM