Some truth about Comcast - WikiLeaks style
sjs at princeton.edu
Mon Dec 20 10:26:11 CST 2010
On Dec 20, 2010, at 3:45 AM, JC Dill wrote:
> On 19/12/10 10:55 PM, George Bonser wrote:
> This is not the result of many different providers, it's the result of one provider stringing many lines to supply service. I'm guessing this was before they figured out how to run trunk lines and then split out the calls from the trunk into individual lines closer to the end user's location - or how to bundle lines together, etc. So each wire we see in that photo is a *single* wire running from the central office to one subscriber (or party line).
> That photo isn't due to a situation where there were numerous different providers, it's due to ONE provider with a monopoly, doing a half-assed job.
It should be noted that running an individual line from the central office to the subscriber can be a good thing when done in a sensible fashion. Amsterdam's Fiber-to-the-Home project called Citynet is an excellent example of this. The city ran a fiber line to each subscriber, which facilitates competitive open access to each line (and makes for maximum long-term bandwidth per subscriber).
"However, the first decision the Citynet project made was more fundamental: should it deploy a passive optical network (PON) architecture or what Wagter calls “home run” fibre, which is a point-to-point topology. PONs share fibre and equipment near the head-end of the network, which does result in some cost savings. But infrastructure sharing does not allow unbundling (allowing other service providers to put their equipment into the local exchange). PONs have a 1:32 splitter in the street cabinet, which means that those 32 customers get locked into the same service provider — and that didn’t fit with the city’s plan to have an open-access network. (Regulators have proposed bitstream access as a solution to this problem, but it’s more complicated to implement.)"
More information about the NANOG