[Tmpcommonsnet] Free Network Definition vs Pico Peering Agreement

[Marc Juul]

On Mon, Oct 14, 2013 at 4:57 PM, Isaac Wilder <isaac at freenetworkmovement.org
> wrote:

> At the meeting in Berlin, I offered to do a more in-depth analysis of how
> the
> current draft of the Free Network Definition compares to the long-standing
> Pico
> Peering Agreement.
>
> I've been travelling and settling back in at home, so I apologize for the
> delay.
> Here goes.
>
> Relevant documents live here:
> https://commons.thefnf.org/index.php/Free_network_definition
> and here:
> http://www.picopeer.net/PPA-en.html
>
> Before getting into nitty gritties, I want to give my impressions at a
> higher
> altitude. I think the design goals of the documents are fairly disparate.
> The
> PPA, I think was intended to act as a governance document. Almost a middle
> ground between the FND and the NCL. In large part, I think that trying to
> do two
> things at once means that neither one gets done particularly well. Many of
> the
> provisos in the PPA *do* have a place in the NCL, imo. For that reason,
> it's a
> little bit odd comparing it just to the Free Network Definition. Still,
> interesting.
>
> Perhaps the preamble of the PPA says it best: "This document is an attempt
> to
> connect those network islands by providing the minimum baseline template
> for a
> peering agreement between owners of individual network nodes - the Pico
> Peering
> Agreement."
>
> That is not the objective of the FND. The (modified) preamble to the FND
> states:
> "Our intention is to build communications systems that function as a
> commons. We
> call such systems 'free networks' and they are characterized by the
> following
> three freedoms." (Note that we're not trying to enable actual
> interconnection
> with this document, only an understanding of what free networks *are*).
>
> Okay. so.
>
> I think the following items from the PPA:
> -The owner agrees to provide free transit accross their free network.
> -The owner agrees to publish the information necessary for peering to take
> place.
> -The service can be scaled back or withdrawn at any time with no notice
>
> relate to the following from the FND:
> The Freedom to participate in the network, and to allow others to do the
> same.
>
> Major salient difference here is that the PPA offers absolutely no
> guarantee of
> any kind that folks will be able to join/expand the network. Only that if
> they
> do so, they will be granted free transit. This is the major difference
> between
> the documents. FND views Free Networks as networks that anybody can expand.
> (Perhaps we need to say, even in the FND, that folks are free to expand
> and be
> expanded, so long as it doesn't hurt the network? Or is the presumption
> that you
> can't do stuff that's bad for the network? Seems problematic, as in a
> certain
> sense, any expansion places addition strain on exists componenets.
> Probably best
> to leave as is, and clarify in the NCL).
>
>
>
> Next, the follow PPA clauses:
> -The owner agrees not to modify or interfere with data as it passes
> through their
> free network.
> -There is no guaranteed level of service
> -The service is provided "as is", with no warranty or liability of
> whatsoever
> kind
>
> relates to:
> The freedom to communicate using the network for any purpose, without
> interception or interference.
>
> PPA I think gets too deep into network performance territory. I see no
> reason
> why one could not have SLA-like agreements within networks, spelling out
> responsibilities. The PPA also makes no mention of interception. Baseline
> PPA
> agreement would allow anyone on the network to snoop traffic. There is
> something
> to be said for not making guarantees, but in a framework of rights and
> responsibilities and freedoms, making sure that people know its not okay to
> intercept seems important. Adequate caveats can be made for necessary
> maintenance and such, but the baseline should be 'no snooping'.
>
>
>
> and finally,
> -The owner agrees to publish the information necessary for peering to take
> place
> -This information shall be published under a free licence
>
> relate to:
> The freedom to modify and improve the network, including the ability to
> access,
> author and distribute information about how the network functions.
>
> Pretty good accord here, though FND takes a much more active stance.
> Again, it
> easy to see how the three freedoms in the FND, if taken to an extreme,
> would
> lead to bad outcomes. Such is the case, I think, with all freedoms. Does
> it need
> to be said somewhere, maybe in the preamble, that these freedoms extend
> right up
> to the point where the interfere with other peoples' same freedoms?
>
>
>
>
> And then the PPA has a bunch of stuff that doesn't match up with FND
> content.
> This is due to the afformentioned mismatch in scope:
> -The owner is entitled to formulate an 'acceptable use policy'
> -This may or may not contain information about additional services provided
> (apart from basic access)
> -The owner is free to formulate this policy as long as it does not
> contradict
> points 1 to 3 of this agreement (see point 5)
>
>
> These are gritty details that have no place in a definition of free
> networks,
> but probably relate strongly to the NCL. In a sense, I think that the NCL
> would
> outmode some of these considerations by creating a more unified framework.
>
>
> I am obviously biased, but I really do think the Definition/License
> approach
> makes more sense. Main take-away is the question of whether the extent of
> freedoms should be circumscribed in the document/preamble.
>
>
>
> I hope this helps. For me, it strengthened the feeling that we're on to
> something good.
>

Thanks for the comments Isaac. This helped me better understand what you're
trying to do with the Free Network Definition. I definitely think that each
freedom in the FND needs an explanation. Specifically, it is not at all
intuitive what the difference between participation and communication is
(freedom 0 and 1). These at least need to be explained, but it would be
better if we could word them so they don't need an explanation.

The problem seems to be with the word "participate", but I'm having trouble
coming up with a good alternative.

Here's another idea. What if we were to define three levels of
participation? Here's my take on it.

0. Freedom to physically connect to the network and to allow others to do
the same.
1. Freedom to send and receive data and have it be fairly routed through
the network, for any purpose, without interception or interference
2. Freedom to route network traffic, fairly, as part of the network and
allow others to do the same.
3. The freedom to modify, expand and improve the network, including the
ability to access, author, remix and distribute information about how the
network functions

What do you think?

-- 
Marc
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sudoroom.org/pipermail/tmpcommonsnet/attachments/20131014/f4de91fc/attachment.html>