Re: [Internetgovtech] Involvement from other constituencies

[Brian Trammell <[e-mail]>]

On 1 Mar 2014, at 5:26, Pranesh Prakash <[e-mail]> wrote:

> Russ White <[e-mail]> [2014-02-26 15:12:56]:
>> 
>>> And this, I think, illustrates the main reason why the IETF's "show up on
>> the
>>> list" policy _doesn't_ lead to increased input from policymakers: the key
>>> problem here seems to be to be the language barrier. We deal with
>>> mechanisms, policy people with policy, and even then the vocabularies for
>>> the same concepts are often wildly different.
>>> 
>>> We (generally) try very hard in the IETF to define protocols which are
>>> "situation agnostic"; this comes from the fact that separation of
>> mechanism
>>> (how you do things) and policy (what exactly it is you want to do) is good
>>> engineering practice, not least because what I want to do may not be what
>>> you want to do, and what we both want to do may change over time.
>>> Experience has shown that when we do embed policy, we create protocols
>>> that are less scalable, less capable, less successful.
>> 
>> +1
>> 
>> Thanks for saying this, Brian -- you expressed it perfectly.
> 
> I couldn't disagree more strongly with this :)

Thank you for perfectly illustrating my point about "wildly different vocabularies" :)

> Protocol decisions whether it is the trust system which underlies BGP,

Not a member of the routing tribe so I can't answer questions about the BGP trust model in any sort of detail, but I would characterize BGP's trust model as "technically expedient" and "risk inconsiderate" rather than "political". 

>  the CA system,

My opinions about the mess the Web PKI has become preclude me from speaking objectively about it, but you're right, I can't disagree. I'll note that the IETF inherited most of the decisions there.

> or the attempting to make all public XMPP servers use s2s and c2s encryption by default,

Decisions taken in reaction to the revelations about pervasive surveillance are, agreed, inherently political.

> or many other kinds of decisions, are inherently deeply political.

> The decision to require all SMTP servers to only prepend to the "received" header, has political ramifications.

Indeed. As does the very decision to have a packet-switched network with source _and_ destination addresses at the network layer. I'll point out that the people who make all this stuff work day to day really like both of these decision, because they make it not-impossible to fix things when they break.

> The attempt by so many people to "fix" what they see as the broken system of e-mail (since until further changes are made[1] there is neither metadata integrity (for the most part, discounting that which DKIM covers) nor is there any possibility of concealing metadata) is also deeply political.
> 
> The question of whether to allow for patent-encumbered standards is a deeply political question.[2][3][4]

Heh. I find myself forced to agree with you here.

> It is, imho, futile to pretend that engineers don't make political decisions all the time.

"That engineers don't make decisions with political implications all the time" I can agree with.

Cheers,

Brian

> [1]: http://tools.ietf.org/html/draft-cailleux-secure-headers-04
> [2]: http://www.ietf.org/rfc/rfc3979.txt
> [3]: http://www.fosspatents.com/2013/03/nokia-comments-on-vp8-patent.html
> [4]: http://www.ietf.org/mail-archive/web/ipr-wg/current/msg02436.html
> 
> -- 
> Pranesh Prakash
> Policy Director, Centre for Internet and Society
> T: [phone] | W: http://cis-india.org
> -------------------
> Access to Knowledge Fellow, Information Society Project, Yale Law School
> M: [phone] | W: http://yaleisp.org
> PGP ID: 0x1D5C5F07 | Twitter: https://twitter.com/pranesh_prakash
> 
> _______________________________________________
> Internetgovtech mailing list
> [e-mail]
> https://www.iab.org/mailman/listinfo/internetgovtech