Re: BitTorrent swarms have a deadly bite on broadband nets
On Tue, October 23, 2007 4:06 pm, Jack Bates wrote:
Errr, 8 pairs per customer? Even 4 is a step backwards. If we're going to do construction at that level, might as well drop in fiber. We're still enjoying the fact that ADSL runs on 1/2 a pair while the customer's phone service is out.
Doing (or getting the incumbent to do, where the last mile is a monopoly) a little bit more of what you already do seems to be an awful lot easier than doing something completely different. Certainly in the (admittedly all European) countries where I've seen it done, getting 4 or 8 copper pairs from a customer site to the exchange is an order of magnitude or more difference in both cost and lead time to doing anything at all with fibre. Every house I've lived in has had 4 pairs already going from the house to the first street cabinet, with just the first pair connected for voice, and getting a second line has always just needed a patch onto a spare pair from the cab to the exchange. Obviously if *everyone* wants 4 or 8 pairs to their house, there's going to need to be a lot more copper between the exchange and the street cabs. It's not clear that *everyone* wants upstream though, and 2M to 5M on a single pair (depending on distance / quality) is quite possible if you wanted to think in terms of ubiquitous symmetric service. I take it that getting spare / new copper in the US is more painful? Regards, Tim.
Tim Franklin wrote:
Doing (or getting the incumbent to do, where the last mile is a monopoly) a little bit more of what you already do seems to be an awful lot easier than doing something completely different. Certainly in the (admittedly all European) countries where I've seen it done, getting 4 or 8 copper pairs from a customer site to the exchange is an order of magnitude or more difference in both cost and lead time to doing anything at all with fibre.
Sorry, I am the incumbent. ;) I was just thinking of the copper necessary to do such a task on a massive scale. It's definitely not in the ground or on a pole at this point in time. One reason DSL was so desireable for many small ILECs was the recovery of copper from dual phone lines caused by dialup.
Every house I've lived in has had 4 pairs already going from the house to the first street cabinet, with just the first pair connected for voice, and getting a second line has always just needed a patch onto a spare pair from the cab to the exchange. Obviously if *everyone* wants 4 or 8 pairs to their house, there's going to need to be a lot more copper between the exchange and the street cabs. It's not clear that *everyone* wants upstream though, and 2M to 5M on a single pair (depending on distance / quality) is quite possible if you wanted to think in terms of ubiquitous symmetric service.
In newer homes, most of the ILECs I work with use 6 pair drops. This is relatively new, though. There are times that it was just 2 pair. It's amazing how things change back and forth over 50+ years.
I take it that getting spare / new copper in the US is more painful?
Depends on locale and quantities. I know of several rural ILECs which are currently undergoing a 3 year process to recover copper. This involves replacing bad peds and boots, cutting off the bad copper in them, and either pulling up some slack or splicing in some fresh copper (splicing 600 pair in a bucket truck is rough on the legs ;). We're shocked DSL has worked at all in some of this plant. The cost to recover and repair what we have is far less than throwing anything else into the ground, but no one considered needing as much copper as it would take to bump everyone from DSL to a 4 pair system. I won't even discuss RBOC mentality when it comes to rural plant (including the entire state of Oklahoma). Jack Bates
participants (2)
-
Jack Bates -
Tim Franklin