[c-nsp] Transporting TDM voice DS3s over Ethernet

[Robert E. Seastrom]

Clayton Zekelman <clayton at mnsi.net> writes:

> OK, Let me get this straight:
>
> 1) Cappucio accuses me of illicit drug use.

Asking if someone has been smoking { pot, crack, etc } is a fanciful
way of suggesting that someone's reasoning is off in the weeds or that
they completely failed to grasp the problem.  Not an unreasonable
response considering that you stated this is "basically large scale
VOIP" (which typically involves echocan and call routing/termination
and such) when actually the problem is much much simpler, basically
just bit-banging.

> 2) He then "corrects" by telling me a DS3 has 672 DS0's (no kidding!) 
> not 1344, when the original requirement was for TWO DS3's (1344 
> DS0's) of traffic.

Why do you care?

> 3) He then suggests the original poster build something with an FPGA 
> (gee, I'll get my Verilog design tools out...), as if that is going 
> to help the guy who has a real need for a real solution, not a 
> science fair experiment in a development lab.

"You could" is a common idiom that in a bygone age would have been
expressed as "One could".  It expresses possibility, but is not to be
taken as a recommendation as to an actual course of action.  "It could
all be done with an FPGA and a small general purpose CPU" is a
reasonable assessment of what it would take to encapsulate a 4760 bit
DS3 multiframe into a UDP packet and toss it out on the wire.
Probably why the stuff from RAD is so reasonably priced.

> And I'm the one accused of responding in a "waspish" way?

Res ipsa loquitur.

                                        ---Rob