On Wed, 31 Aug 2005, William Chops Westfield wrote:

> Maybe.  I don't BELIEVE that's the case.  With packet switched technology,
> you can build a backbone infrastructure that is reliable (by most
> possible definitions) in spite of individual components that are not
> reliable (that's what TCP/IP was all about, though I doubt that most
> modern networks are vert true to the ideal...) (And alas, this does
> not stop would be customers from demanding the individual components
> be extremely reliable as well, driving up the price.  But that's good
> for our bottom line, IIRC, so I shouldn't complain too much!)
>
> There is a lot of interest in latency/QoS for packet networks (again
> driving up prices), but it's not clear to me to what extent that is
> truly important, especially when it seems to be much easier to step
> up bandwidth of packet links compared to TDM-style links (consider
> ISDN vs DSL, for instance.  Or token ring vs ethernet (guaranteed
> latency used to be a big selling point for 4mbit and 16mbit token
> ring networks.  Ethernet was "too unpredictable."  And then came
> 100Mbit ethernet..))

Imho the present state of the art home networking is a dsl customer 
using voip to another dsl customer using voip. This leads to delays of 
the order of magnitude of one second per side all told (handset to 
handset). This is way too long. Out of this, the ping time alone is 
about 180msec best and 300msec average. Talking to someone with 1 second 
delay precludes dialogue and requires iron discipline from the speaker 
(you can hear your own voice as sidetone echo coming from the remote 
with ~2 seconds of delay - this is extremely annoying and causes most 
speakers to shut up when talking).

The 100Mbit ether is very nice but one should not forget that the first 
segment of long haul for most people is 256kbit DSL uplink on a good day 
with no network problems.

Peter

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist