more on comments? Does faster broadband really matter?

[David Farber <dave () farber net>]

Begin forwarded message:

From: Brett Glass <brett () lariat net>
Date: December 27, 2005 6:40:38 PM EST
To: dave () farber net, ip () v2 listbox com
Subject: Re: [IP] comments? Does faster broadband really matter?

Jeremie Reimer wrote:

> Internet blogger Om Malik has written an interesting piece on the
> new, faster broadband connections that are now becoming available to
> US consumers. His premise is that the faster speeds are not that
> important, because they don't translate into a significantly better
> experience for the end user.
>
> The gist of his argument is that most online activities, like
> standard websurfing, are not significantly sped up by high-bandwidth
> connections, and the few that are, such as downloading, are not
> typically time-sensitive anyway:

Mr. Malik is correct. As an ISP, we see that once speeds get above
a megabit per second, bandwidth is not the limiting factor in most
users' Internet experiences. The bottlenecks that most affect users
who are browsing the Web is the speed of their computer systems --
their ability to assimilate and process a Web page quickly -- and
the server's response time. Often, it's the ads that slow a page
down; the browser is ready to compose the actual content, but can't
do so until the ads come in.

Legitimate downloads are similar. Even if downloads are not throttled
at the source (a rarity) or come out of a high speed cache, most
consumers' computers still aren't able to accept the data at more than
a few megabits per second. Our wireless network can deliver the goods
at up to 11 Mbps, but it's a rare user whose computer can accept it
at that speed. Our tests show that the most users' maximum pull rate
is 1 Mbps during a download.

Usually, when our users experience bandwidth exhaustion, it results
from malware (viruses, worms, spyware, zombies, etc.), too many users
on a limited size pipe (e.g. a whole office on a 512 Mbps DSL line),
bandwidth hogging on a shared connection, or very inefficient practices
(such as sending an uncompressed, high resolution scan of a memo instead
of a document file -- ugh! -- to a long mailing list of users). P2P is
also a frequent cause of bandwidth exhaustion; naive users often don't
realize that after they've finished downloading, their computer is still
consuming their bandwidth sending illegal copies of what they've  
retrieved
to the entire world.

We take great pains to engineer our network to give our users the data
rates they've paid for. So, when they report slowness, it is virtually
always a reflection of what the user is doing rather than of what we're
doing.

The only exception is when we get interference on our unlicensed  
wireless
links (which we must run unlicensed because we can't get licensed  
spectrum
for love or money). But that's a whole 'nother topic.

--Brett Glass, LARIAT.NET





-------------------------------------
You are subscribed as lists-ip () insecure org
To manage your subscription, go to
 http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/