[governance] What is Network Neutrality

[Ralf Bendrath]

Parminder schrieb:

> when Milton in his email dt 7th Jan asked if I was speaking
> against edge catching, I specifically responded in my email dt 8th  that
> I was *not*.
Thanks for the clarification, and sorry if McTim or I have misunderstood
you here.

> I have written at least 4-5 times in the last few days that I am *only*
> speaking against price-differentiated QoS/ speed of transmission/
> delivery of content, which in my view violates NN.
> 
> I also asked you and Ralf specifically if you are for it or against it.
> So I ask it again. (and pl do respond)
I am sorry I can't give a clear-cut answer (and no, Ed Felton is not
particlarly clear here either, e.g. End-to-End and Content
Nondiscrimination is basically the same, because both implies that ISPs
are not entitled to look into the payload of the packets, but instead just
care about the headers.).

There are several ways to read this. I'll try to distinguish them based on
who would be doing the delivery and discrimination of the content.

1. Downstream ISP

1.1. My downstream ISP is throttling specific services / protocols and
treating others better.

This may be ok if I as the customer can choose it - say, if I don't need
p2p protocols or fast pings and get a better price in return.

But one may want to argue against it in principle, because it includes the
danger that these last mile discriminations make the internet less open
for innovation at the application layer.

-> I am not 100% sure which position to pick, though I lean towards the
latter. In any case, this has to be transparent.

1.2. My downstream ISP has a deal with a specific content provider to
fast-track their packets, or is throttling packets from specific other
sources.

-> This is not what I would like. But again, I don't really see this as a
realistic threat anyway if based on money flows from the content providers.
The real threat here is filtering and censorship based on more or less
arbitrary decisions by the ISP or the government or shady private
governance arrangements.

1.3. My downstream ISP does edge-caching or co-location.
-> This is ok if anybody can get a cache there for the same price.
(Yes, I know you're also ok with it, Parminder. This is just for the sake
of completeness)

2. Upstream ISP

2.1. The upstream ISP of the content provider is connecting him through a
faster pipe than other customers.
-> This is ok as long as anybody can get this deal for the same price.

3. Backbone provider

3.1. The backbone provider is treating traffic differentially based on
from which ISP it comes from, depending on the price he pays.

-> I don't know enough about the technology and the economics of the
peering and interconnection agreements behind all this etc. I guess it is
should be acceptable that you can get faster backbone access for a higher
price.

> *Now if this helps further understanding I may tell you that Milton (and
> Lessig) supports 'flavour' or type 2 above as constituting NN, but is
> against type 3 (Milton, you can pl confirm this.)
As far as I understand Lessig's position, he would also agree to NN as
type 3, "Content Nondiscrimination" (i.e. based on the payload of the
packets), but would allow "Source Discrimination" (i.e. special fasttrack
routing deals, edge caching etc.) if it is nonexclusionary.
Milton?

> I can though judge that you got confused by a quick - and, habitually,
> dismissive - reading of the following sentence I wrote
> 
> " Now, if some content providersare able to pay and line their content
> up closer to the user relative to other content, without her exercise of
> such a choice, it obviously constraints her freedom and choice."
> 
> This is not at all about edge caching. Please read it in the specific
> context I have built, of the picture of Internet as a democratic media.
> It is meant somewhat figuratively, as content which is pushed harder at
> the user through higher-price better-QoS than the rest of the content
> she receives. I wrote the above sentence after I drew a picture of the
> Internet where one is assured *equal relationship to* all the content
> present on the Internet and whereby how price-differentiated QoS
> violates this equality of relationship. It is this sense  I meant 
> higher QoS content as being pushed harder at the user, and spoke of it
> figuratively as lined closer to the user, *relegating the rest of the
> content*. It is in this way that the user's free choice is constrained.

Thanks for the clarification. Do I understand correctly that this would
roughly be described by my option 1.2 above? In this case, I still don't
see how it would work and that it is a realistic threat, but I would agree
with you that this is not desirable.


I hope this all makes clear that there are several dimensions to the
discrimination of packets - who is doing it, based on which kind of
agreements and choice, is it done based on traffic sources, on content of
packets, on protocols, etc. So a straightforward answer is at least tricky.

I suggest we try to prioritize the different dimensions along the severity
of their limitations on user freedom and choice. There are surely some
kinds that we probably would not like to see at all (e.g. censorship based
on content or traffic source), while for others, we may at least want to
ask for transparency and user choice (e.g. throttling of specific
protocols), because it seems difficult to find agreement among us for them.

Best, Ralf
____________________________________________________________
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