NNSquad - Network Neutrality Squad

NNSquad Home Page

NNSquad Mailing List Information

 


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[ NNSquad ] Re: Avoiding unfair ISP bandwidth manipulations


One criticism I often see leveled against Vuze and FP and their supporters
is that they make no alternative suggestions.  But they have, and I suggest
that people go read their published blogs and FCC filings (petitions,
complaints, and comments).  There are plenty of ideas -- some good, some bad
-- there.

 > Question: Would this problem be mitigated if all IP-based traffic,
 > other perhaps than basic not-on-demand, non-PPV TV, were subject to
 > the same bandwidth caps and other limitations?  That is, if an ISP
 > were cajoled or required to treat its own offerings that competed
 > directly with external services as being subject to the same monthly
 > bandwidth caps, throughput throttling, etc., what would be the
 > effects?

Good impacts:

1.  Customer-preset or application-based QoS might actually begin to be
embraced by the ISP, since there is nobody behind the scenes interfering
with it.  In so much as they did not work, the ISP no longer becomes the
scape goat and developers could find and fix QoS bugs with more confidence.
Meanwhile the QoS standards themselves can better mature.

2.  Returning the congested network to a neutral state might uncover new or
increased problems with erratically-sized, erratically-spaced packets in the
presence of large, steady packets.  The responsibility for improving these
problems will now lie with the application publishers taking the specific
steps necessary to improve reliability and robustness.  

3.  Instead of dropping them, NAT devices and network stacks might begin to
receive and respond to network and routing signals being sent via ICMP and
IP.  Currently, the cheapo routers simply drop them on the floor.

4.  QoS-aware consumer devices will increase in popularity and drop in
price.

5.  The times and amounts of congestion (overselling and/or poor conditions
even further upstream) would be revealed to applications and end users as
they could not be masked by deflecting all of the impact onto one or two
disfavored applications.

Bad impacts:

1.  For a plan like this plan to succeed, ISPs would have to operate
honestly, correctly enforcing the self-discipline necessary -- or being
forced by audits and punishments -- to put their own self-interests under
their own cap.  One way presumes dishonesty and the other has been
demonstrated not to work.  And is PPV really where to draw the line?  Why
aren't the HBO channels under the cap?  What about ESPN?  Or the rest of the
basic channels?  Currently, if you don't subscribe to CableTV at all, you
pay $10-$15 more for the same share of bandwidth that your TV-watching
neighbors do.  If fair sharing of bandwidth is the goal... anyway, this
should be workable, but it seems like a complicated mess.

2.  Bandwidth caps and other limitations means that someone, somewhere, is
not being served.  At what point is the cap unreasonably low?  At what point
does a cap not only prevent excessive use, but also stifles expression or
innovation?  Comcast's cap, albeit invisible, is currently one of the most
generous in the industry.  To avoid this, the caps, limitations, and the
customer's accurate usage as compared to those limits need to be more
transparent.  The customer also needs to have options (buy more, shift usage
away from prime-time hours).  Wouldn't it be better if caps and limits were
reserved only as short-term stop-gap measures to get us through the
construction period of building the network we apparently want?

3.  The majority of Internet users won't have any idea why the network is
suddenly so congested that everything crawls and VOIP sucks.  And even if
they figure out that their ISP is oversold, the lack of competition gives
them zero or one avenues to do anything about it.  



And now, other useful perspectives:

DIAL-UP INTERNET ACCESS

Figure out how many different dial-up Internet Access options do you
currently have at your address over a PSTN connection at home.  How many
companies can you find?  How many different ways does each package its
product (e.g. flat-rate, by time, or both)?  Stop after 10 minutes of
research, even though you won't nearly be done, you'll know enough...

Does a market with so many choices need a great deal of Network Neutrality
regulation?  What can we learn from the dial-up market to improve issues we
see in the CATV-Internet market?  Telco VOIP/FIOS?  


ELECTRICAL POWER

An electric company offers its customers two different billing options, not
including a few additional options for those who want renewable power or to
generate some of their own power.  You can read these choices --
http://www.portlandgeneral.com/home/products/power_options/default.asp -- 
I choose this example because there are commonalities -- despite the
choices, the feed line and the power quality to the customer never changes.
Despite the chosen plan, the consumer's on-demand access to the pool of
energy is the same. The customer can always choose to do whatever, whenever
(under one plan, the customer can save money off the basic rate by moving
consumption to non-prime hours, or pay more than the basic rate if they do
not).  

If an idea like that were applied to Broadband, is that neutral?  Is it
"Reasonable Network Management?"