nanog mailing list archives

Re: What vexes VoIP users?


From: Jay Ashworth <jra () baylink com>
Date: Mon, 28 Feb 2011 18:15:27 -0500 (EST)

----- Original Message -----
From: "Joe Greco" <jgreco () ns sol net>

With end-to-end digital, you can have reliable call supervision and
status, OOB Caller-ID delivery, crystal clear call quality, probably
the ability to handle multiple calls intelligently, no hook race
conditions, etc.

When you throw that one stupid and pointless analog hop in there, you
are suddenly limited and broken in so many ways.

Sure.

But I don't think it's the analog hop that people are really concerned
about *per se*... it's the fact that the traditional analog last-mile 
*connects you to a "real" CO*, with a "real" battery room, that's 
engineered -- in most cases, to cold-war standards, *through a loop with
very low complexity*.

If you have DC continuity and good balance to ground on a copper pair, 
you are *done*; no intermediate gear, no batteries, no config files,
nothing.

All I need at the residence is a 500 set, and the complexity of *those*
is super low, too.

The real, underlying problem is that people take insufficient notice
of all the complexity pinch points that they're engineering into loops
in exchange for the extra controllability they get because everything's
digital end to end.

When I'm bringing 31 T-spans into my call center, that extra complexity
is easily justifiable.

For grandma's phone?  Not so much.

And it doesn't *matter* whether it's riding on a cable internet link
the complexity of which is already amortized: you're now *adopting* that
complexity onto the voice service... the semantics of which (used to
be) very well understood and not at all complex at all.

From the user perception standpoint, I think, it's a tipping point
thing... just like Madison WI.

Cheers,
-- jr 'that was *not* an invitation' a


Current thread: