Re: [Internetgovtech] Involvement from other constituencies

[Andrew Sullivan <[e-mail]>]

On Sat, Mar 01, 2014 at 09:27:13AM -0500, John Curran wrote:

> We have to face the fact that the remarkable success of the Internet means 
> that it very difficult for a given government or community to _not_ use IETF 
> protocols and still achieve global interoperability.  

That's not merely a "fact", it's a tautology.  If you want global
interoperability ("with the Internet" is implied in your original),
then you have to use Internet protocols.

> This in turn only reinforces the need for protocol development
> processes which are open to all interested parties, and transparency
> in decision making.

I agree, but I think it's important to recognise that many of the
people who appear to want the pony named "using the Internet without
using IETF protocols" actually don't want a process that is open to
all interested, or transparency in decision making.  Many of them
instead wish the decisions to be taken by treaty bodies in closed
rooms.  I think it is not surprising that people are suspicious of
importing "political considerations" into the protocol-making
discussion, because that looks like a Trojan horse for "takeover by
treaty organizations".

> When we need to consider factors other than purely technical merit, it may 
> result in social, political, or economic tradeoffs being integrated into the 
> protocols, 

I strongly disagree with this approach to describing the question.  If
we need to include certain social, political, or economic tradeoffs,
then those are part of the use cases that the protocol is supposed to
be addressing.  Therefore, this is not some case of considering
factors other than purely technical merit: there's no such thing as
"pure technical merit" without the operational conditions expected of
the protocol.  This is engineering we're supposed to be doing, not
mathematics.

What we sometimes have to confront, however, is indeed the question of
whether this or that tradeoff ought to be on the list.  That is, to my
way of thinking, the difficulty we're having, but it's not different
in kind to the problem developers have always had in understanding
what the product and sales people want.  Getting the requirements
right is hard, and it gets harder the further one gets from being the
end user.  In the early period of the Internet, approximately all the
users _were_ the developers, so figuring out the requirements was
probably easier in that you could talk to people and find out their
problems.  It gets harder the less representative of actual users our
population gets. 

Best regards,

A

-- 
Andrew Sullivan
[e-mail]