[Asterisk-Users] Comparision of IAX2, FWD, iaxtel etc etc.

[Jim Van Meggelen]

asterisk-users-bounces at lists.digium.com wrote:
>> IAX2 is a VoIP protocol created by Mark Spencer and released under
>> the GPL. It is an alternative to SIP, MGCP, H.323 and the like. To
>> those familiar with it, it is considered to be technically superior
>> to all other VoIP protocols, and is popularly expected to render all
>> the others obsolete.
> Spoken like a true zealot :)

I looked up the term "zealot", then re-read my statement. From what I
can read, I was describing a) some facts about IAX, and b) a sentiment
popular in the Asterisk community. How that makes me a zealot is
unclear.

> There is a time and a place for the IAX and IAX has benefits
> in certain circumstances. Its not that one protocol is better
> than the other in *all* cases.

LOL! Are you telling me the world is not all black and white? Well,
being a zealot, I have no choice but to reject such thinking as
heretical.

IAX is less of an academic protocol, and more of a practical protocol.
IAX and Asterisk will succeed as Linux did - more and more shops will
use them to solve problems, until somebody takes a survey and realizes
IAX is king. The difference is, the growth of Asterisk (and IAX) will
proceed at a far more rapid pace than Linux. Why do I say this? I have
at least two reasons: 

1) because Asterisk (or something like it) is desperately needed in an
industry that has been stagnant for far too many years and is still 
producing products that don't solve problems.

2) because although SIP has been embraced by the industry, no one has
really got it working well yet; at least not without a lot of complex
engineering. The VoIP media is well aware of this, and IAX is attracting
interest everywhere because it just works.

Now, I'll grant you that SIP has tried to cover just about every kind of
media you might want to connect, but that's not why it has been embraced
by the industry. The key to the success of SIP is that the closed,
proprietary mindset of the telecom giants killed VoIP, and in doing so
killed the whole telecom industry for quite a few years (it still has
not begun to properly recover). SIP is the first well-regarded VoIP
protocol that's open to anyone, and was developed by a network-centric
body (the IETF) rather than a telecom-centric body (the ITU). SIP
doesn't care what the telecom giants think -- it's free for everyone --
therefore the telecom giants can't bugger it up. They can accept it, or
be left behind. Some of them will be left behind (hint: sell Nortel
stock while you can).

> Common Benefits of IAX
> 1) NAT
> both the signaling(call setup/ tear down) and the audio
> stream move using the same udp port, thus it does not have
> the issue of multi-port VoIP protocols like h323, SIP etc to traverse
> NAT devices. 

Rather important in enterprise networking.

> 2) Trunking
> Its is able to 'trunk' the audio streams for multiple iax
> channels with a single group of udp control headers. Thus it
> is more bandwidth efficent when moving a larger number of
> channel beween 2 common points.

Rather important in both enterprise and carrier networking.

> 3) PBX
> Its the native protcol for a * PBX, so it could be argued is easy to
> use for a * PBX... 

Indeed. And since Asterisk is going to render all other PBXs obsolete,
IAX is poised to become very successful indeed. That's not zealotry;
it's prognostication.

> It is open to debate to see how it will scale since the
> signal/audio stream are  in the same udp encapsulation, it
> makes it more difficult to off load the audio from signal (eg
> registrations) 

Any more difficult than peeling SIP apart in order to get it through
NATs? Besides, as soon as that is determined to be a problem, it will be
addressed.

> IAX has virtually no support out side of asterisk community,
> and no current effort to supply an rfc for the IETF.

Perhaps not yet. SIP was pretty well ignored just a few years ago as
well, so were many technologies in their infancy. As for an RFC, I get
the impression that you are arguing that an RFC is an essential
component of industry acceptance. My question is: If that is true, can
you help me to understand how NAT became so successful?

> You might wanta take a look at Marks comments about iax
> http://www.voip-info.org/tiki-print.php?page=IAX+versus+SIP

Very nice link. I enjoyed reading that. Thank you.

Regards,

Jim.