[okfn-help] How should the OKF talk?

[William Waites]

* [2011-07-08 15:46:01 -0400] Michael Chelen <michael.chelen at gmail.com> écrit:

] Much research can be done freely, before adding additional
] infrastructure. For example, testing if SIP and Mumble clients meet a
] minimum standard for ease of use? Server setup and operation can also
] be tried out in advance of scaling for widespread usage.

Certainly. I'm not necessarily the best person to evaluate ease of
use, and I'm not in a position at the moment to make a detailed
investigation (though that may change later in the summer).

I can tell you from experience (I wrote the original NAT-traversal
code for Asterisk's SIP implementation for example) that because of
the way SIP works, separate UDP ports for signalling and media, and
the fact that many commodity routers have horribly broken NAT
implementations that the good functioning of SIP is highly network
dependent. I can also tell you that if we were to run our own
telephony system, Amazon would not be the place to do it because of
their own private addresses / NAT arrangement which complicates things
further - it could be made to work in some circumstances but would be
brittle.

Also there is the IAX2 protocol, which is an Asterisk invention and is
supported by the likes of FreeSwitch has much better NAT traversal
properties because it multiplexes signalling and media onto a single
connection and does not bother trying to know the "real" IP addresses
involved. In this respect it is similar to Skype. But since it is not
a proper IETF standard it is not as widely implemented in clients.

What I know little about, since I have not worked with VoIP systems
much since about 2005 is the current state of the art of client
software, and this is precisely where the ease of use concerns are
most obvious.

If we are to look at these things, an important but not overriding
consideration is the ability to connect the system to real
telephones. I don't know how easy that would be with something like
mumble.

Cheers,
-w
-- 
William Waites                <mailto:ww at styx.org>
http://river.styx.org/ww/        <sip:ww at styx.org>
F4B3 39BF E775 CF42 0BAB  3DF0 BE40 A6DF B06F FD45