Is this claim true, or not?

[Charles Steinkuehler]

Jonathan Hutchins wrote:
>> -----Original Message-----
>> From: Charles Steinkuehler [mailto:charles at steinkuehler.net]
> 
>> Yes, the "last-mile" bandwidth for cable-modem connections will slow 
>> down based on how many of your neighbors are currently 
>> donwloading ...
> 
> What, you sell DSL or something?  
> 
> 1) If the capacity for the local loop exceeds <current active
> downloads>X<current maximum speed>, you won't see any effect at all.
> 
> 2) EVERY SYSTEM HAS TO CONVERGE SOMEWHERE.  WHERE IT CONVERGES, EITHER THE
> CAPACITY IS THERE OR IT ISN'T.
> 
> 3) Every RoadRunner user on this list has reported that downlink capacity
> remains very good - almost always better than the 1,500K they promise (or
> promised me at least).
> 
> 4) Many DSL users report degraded performance due to distance from their
> local switch limiting the actual service they get.

No, I don't sell DSL, and I went on to say a well managed cable-modem 
system seems to win over a well managed DSL system in all markets I've 
got first-hand knowledge of (Topeka, San Antonio, San Francisco, and 
Lawrence).  Poorly managed cable-modem systems, however, seem to allow a 
handful of peak users to saturate a node, to the detriment of their 
neighbors, while the fairly low "cap" on DSL links (typically 1.5 MBit/s 
max) keeps a poorly run DSL provider from having the same problems (or 
at least mitigating the affect somewhat).

I do use a "wimpy" SDSL connection, because even at my distance limited 
384KBit/s rate, upload rates are faster than those for the local 
cable-modem service (important for my VPN link), and I get a /26 block 
of static IP's to go with it.  If (more likely when) the company I work 
for goes tits up, I'll probably switch back to cable-modem service, 
rather than ADSL...the SDSL link is too pricy to keep w/o a business 
paying for the service with pre-tax $$$.

The devil's in the details...

-- 
Charles Steinkuehler
charles at steinkuehler.net