[governance] IGP Alert: "Net Neutrality as Global Principle for Internet Governance"

[Karl Auerbach]

Suresh Ramasubramanian wrote:
>> editorial control over what is carried (and deciding what is spam and
>> filtering it clearly is editorial control.)
> 
> You might want to review 47 USC 230 which does provide safe harbor for such
> editorial control, at least in the USA.

That section of the US code is very complex and relies on some rather 
tight definitions that may or may not apply to any given specific 
situation.  Moreover, section (d) does impose certain obligations of 
notification in the context of an agreement with a "customer".  In other 
words, this is a part of the law that providers need to study deeply.

Moreover the necessity for that part of the USC tends to amplify the 
broader general underlying rule that "he who censors bears the risk of 
over-censorship".

>> Put these two dimensions into a 2x2 matrix and we see that in one of
>> the boxes there are providers that do exercise editorial control but have
>> no contractual relationship with the receiver.  Those providers are, to my
> 
> Er, like?  There are some providers who have contractual relationships with
> the ISP that hires them to provide their servers such a service (brightmail,
> etc)

Yes, that's the kind of chain of contractual relationships through which 
permission from the ultimate recipient ought to be passed.  For example, 
I have hired Postini to filter my incoming email.  They have my 
contractual permission to do so based on their expertise and opinion 
about what is junk.  And at one of my office sites I use a local access 
provider that normally blocks TCP port 25 - but I have an agreement with 
'em not to block that port with an obligation on my part to be vigilant 
to avoid becoming an open relay or originator of junk.

But most of us don't have relationships with core providers.  Their link 
to us is indirect via our edge providers.  So the permissions and costs 
need to be expressed in a language well suited for such indirect 
relationships - prices.

And as a practical matter, core providers are not good places for 
traffic to be winnowed for naughty bits - the circuits of a packet 
switching fabric inside a carrier grade router are hardly the place to 
do semantic evaluation of application layer content.

>> A provider proper mode to induce users to give such permission is to
>> raise the costs charged to those users who withhold permission and thus
> 
> Not really. Most providers find that the costs of offering an unfiltered
> account are actually far higher than offering a filtered account. Not just
> bandwidth / disk etc - that is small change. Support costs from users who
> get spammed, phished etc and then complain about it?  Do the numbers.

I think you read me backwards.

I believe that those who want non-filtered will often end up paying more 
- if for no other reason then they are causing more bits to be carried. 
  Perhaps some providers don't do a good job of communicating to their 
customers that if they want unfiltered then those sers can't complain 
(or can't complain without paying $$ to file a complaint) if they 
receive junk.

I'm not at all saying that providers should absorb costs created by user 
demands.  Rather I'm saying that users should be given information that 
clearly defines what they are buying.  Some providers may chose to offer 
a menu of choices (and prices), some may simply refuse to sell service 
to customers who want things that a provider doesn't want to sell (for 
whatever reason, other than a few limited reasons, such as 
discrimination on the basis or race or sex or ...)

		--karl--
____________________________________________________________
You received this message as a subscriber on the list:
     governance at lists.cpsr.org
To be removed from the list, send any message to:
     governance-unsubscribe at lists.cpsr.org

For all list information and functions, see:
     http://lists.cpsr.org/lists/info/governance