>>> I read that cable modem speeds depend on how many
>>> users are using the lines at the same time, while this
>>> is not the case with DSL. IS this right?
>>NO, this is fundamentally wrong (and pure DSL marketing B.S.).
> No, this is not pure DSL marketing B.S. The problem is that much of
> the existing cable was setup with only broadcast in mind and not
> for Internet access, whereas everyone phone lines are inherently
> setup for bi-directional, point-to-point. This is one of the main
> reasons why the cable to the home needs to be upgraded, so that
> there is no performance bottleneck into the home.
Ummm, nope again.
The cable upgrades are to enable bi-directional access to the home, not to address a "performance bottleneck into the home". DSL's bidirectional point to point is (heavily) concentrated in the CO into a DSLAM where it's loaded onto a circuit bound for the overloaded Internet through an ISP. Overloading ratios for DSL are typically 100:1 or more (meaning 100+ "1.5 mbps" DSL users sharing a single 1.5 mbps T-1). The advantages of dedicated 'point-to-point' DSL is a fallacy. Instead, @Home users share a much larger bandwidth on the neighborhood coax segment, and they've got an architecture specifically designed for high-speed. In comparison, DSL is a joke.
I suggest you re-read the 'xDSL myths' piece on the @Home site (www.home.net), or check out the discussions on comp.dcom.xdsl (where the 100+:1 DSL concentration was discussed in detail recently).
--Altec |
