[ NNSquad ] Re: Subject: Re: [IP] "Entry level pricing"

["George Ou" <george_ou@lanarchitect.net>]

If you're getting 1/5th the speed, then it sounds like your particular line
quality is so poor that your sync rate is only 768 Kbps.  However, your case
is the exception and not the rule.

See figure 10 on page 26.
http://www.techpolicyinstitute.org/files/wallsten_international_broadband_co
mparisons.pdf

It shows a chart derived from a British Ofcom report in 2007 on advertised
bandwidth versus actual.  The US and Canada are actually the two nations
with the most accurate advertising.  Japan has the fastest advertised speeds
on average but they fall far below on delivered.  The result is that
customer satisfaction in Japan is only 40% while US and Canada are 86% and
80% respectively (based on what I can eyeball off figure 10).

So while there are always going to be exceptions where a customer may not be
syncing correctly due to line speed, that is not generally the case.  AT&T
these days will refuse to sell a certain bandwidth to you if they believe
the line quality won't deliver those speeds.  Now it's possible that their
database isn't always accurate and there are false positives where someone
could get a certain speed and AT&T won't sell that speed and false negatives
where they do sell you a faster speed but you can't actually get it.

Consumers should always verify their own speeds at places like speedtest.net
and if they don't get 80+ percent of the advertised sync rate on
speedtest.net or some nearby test server (due to a lower sync rate), they
should demand their money back.  Some DSL modems are also better than
others, and sometimes you may have to power down the modem for 5 minutes and
power back on to get the DSLAM to reset its sync rates on your port.

Now I do think ISPs bear a lot of the blame for the lack of a transparency
standard, and they also need standards for delivered versus advertised.  It
would seem to me that ISPs should proactively monitor sync rates for
customers and either fix the cable quality or drop the customer's speed plan
and give them the lower rate instead of expecting consumers to figure this
out for themselves.  That would go a long ways towards gaining better
customer satisfaction.



George Ou

-----Original Message-----
From: Russell Smiley [mailto:im.russell.smiley@gmail.com] 
Sent: Sunday, October 04, 2009 6:43 AM
To: nnsquad@nnsquad.org
Cc: George Ou
Subject: Re: [ NNSquad ] Re: Subject: Re: [IP] "Entry level pricing"

George Ou wrote:
> "I like to watch Congressional hearings live.  Over AT&T's allegedly 
> 3Mbps DSL downlink they come in choppily, with frequent pauses and 
> voice-video mismatch."
> 
> I think you are grossly exaggerating here.  Even YouTube 2.25 Mbps 
> 720P streams come in perfectly for me on my 3 Mbps sync rate 
> connection which works out to about 2.5 Mbps raw data speeds.  Most of 
> the government streams are well under 1 Mbps and they're usually 
> buffered.  If you are experiencing choppy video, it's probably 
> something else because I stream stuff from the FCC hearings all the time
with no problems.

I think this is where we come back to the operators network management and
advertising claims.

I am not exaggerating - I was sold a "3Mbps" DSL connection and YouTube
video (or any other type of video for that matter) was consistently choppy,
buffering and other such impairments. The question of whether it really was
a 3Mbps connection comes down to expert knowledge that the vast majority of
consumers (even businesses) wouldn't have the remotest idea of how to begin
testing. I was also unable to use vnc (in low bandwidth connection mode, of
course) for remote desktop work due to latencies of several seconds - and in
this case the problem was identified as not my employers network.

I realised after some time that I had never once observed anything close to
the rated maximum speed on my 3Mbps DSL connection that I was paying
$50/month for (in Canada). In fact I was consistently getting 1/5 of that
speed. I have since downgraded to a 256kbps link for $25/month and have
largely observed almost no difference in performance.

Based on your testimony, George, apparently you have an ISP of the highest
integrity who deliver close to what they said they would. This seems to be
inconsistent with the experience of the  majority of people on this list. My
suspicion is that this maps to general consumers as well, except that they
have no means of characterising what they are delivered by their ISP - nor
even consumer savvy to know that they _should_ test the performance.

Russell.