[Freedombox-discuss] Freedombox threat model

[intrigeri]

Hi,

Daniel Kahn Gillmor wrote (29 Jun 2011 17:46:00 GMT) :
> Other reasons to not use tor for certain workloads/environments:
[...]
>  * you have tight latency constraints (not just throughput).

Right.

>    For example, the subjective experience of telephone calls is
>    severely degraded by even a 150ms lag, even if the throughput is
>    minimal.  Tor's extra hops add to the latency of any circuit.

About this specific example, I've tested VoIP between two Tor hidden
services (which means the end-to-end connection has twice as many hops
than "normal" Tor), both clients sitting behind poor DSL connections;
it results in a constant ~1s lag and no noticeable throughput problem.

Regarding subjective user experience, I would not say mine was
"severely degraded". As a matter of fact this configuration transforms
conversation a bit; it is slightly slowed down, mostly because you
need to wait 1s to be sure the person you're talking with has finished
his/her sentence before you jump to the microphone to speak yourself.
I will go as far as saying this tiny slow-down can indeed often make
the conversation greater and more dense in a certain sense, because
every part of it has had more "empty" time to mature and is thus more
thoughtful.

Bye,
--
  intrigeri <intrigeri at boum.org>
  | GnuPG key @ https://gaffer.ptitcanardnoir.org/intrigeri/intrigeri.asc
  | OTR fingerprint @ https://gaffer.ptitcanardnoir.org/intrigeri/otr.asc
  | Did you exchange a walk on part in the war
  | for a lead role in the cage?