ietf
[Top] [All Lists]

Re: As if you don't have enough to read..

2015-03-14 12:40:34
On Sat, Mar 14, 2015 at 11:38 AM, Henning G Schulzrinne 
<hgs(_at_)cs(_dot_)columbia(_dot_)edu
wrote:

The US regulatory process provides both individuals and organizations
opportunities for input, both through written and oral contributions.
Submitting a comment is easier than writing an I-D, as no XML is required,
but there are deadlines ("comment periods") just as for I-Ds.

From my experience, FCC staff are generally more than willing to meet
individuals willing to provide technical information, particularly if they
have done their homework and are not just shilling as for-hire experts. If
you look at the citations in any major FCC item, you'll see plenty of
references to individual faculty presentations. For various reasons, the
faculty visiting the FCC tend to be law professors, rather than
engineering, but that's mostly self-selection.


The risk of making a submission is being asked to help fix the problem...

And the fundamental issue here is that the Internet architecture does not
have any concept of payment for transmission. It is assumed that everything
is 'free'. While this solves a very large number of problems it also
creates problems.

The telephone system has a very clear payment model in which the initiator
of a call is responsible for payment over the wired network and the
receiver bears the cost on wireless. That model simply does not apply to
the Internet and attempts to apply it produce nonsense.

Settlements and interchange agreements are tricky. Especially so when the
data streams are very one sided. But at the end of the day the consumer is
paying their broadband ISP for the ability to connect to NetFlix and
YouTube.


However, engineers should also realize that the high-order-bits of
decisions are not merely technical. They tend to involve legal and economic
efficiency arguments that are rarely clear-cut - they often involve
predictions on how a decision will affect investment, competition, public
safety and human behavior, among others.


Engineers are capable of understanding political and economic arguments!
they just understand them according to a totally different ideological
frame to the one the rest of the world lives in.



It's important to get the technical details right, but it's at least as
important to recognize that this is not just (or even mainly) about the
proper definition of IP addresses or "best effort(s)".


The proper definition of IP addresses is probably irrelevant. More critical
is the understanding of the way that interchange agreements work.

I think there is also something of a problem with the advice some folk
give. I was in a meeting where a very eminent IETF-er ranted about a
'nonsense' requirement that a part of the Federal Govt would like. A week
later I had a solution worked out that did not cause any of the
catastrophes claimed.

It is important to recognize that the 1980s Internet architecture was a
pragmatic approach in which the answer to a lot of hard problems was 'we
are not going to consider those right now'. Well now we have got to the
point where they matter.