[Freedombox-discuss] Establishing Communication between Freedomboxes

[Lukas Nagl]

On 2011-07-07 22:29, Ted Smith wrote:
> On Sat, 2011-07-02 at 16:14 +0200, Lukas Nagl wrote:
>> Hi,
>>
>> although the topic is raised in various subjects, I think that there is
>> a need for a specific place to talk about how different freedomboxes
>> will be able to establish communication with each other.
>>
>> This question doesn't concern itself with the contents of communication,
>> but rather with how a communication can be first established, and
>> re-established in the future (e.g. John wants to be able to communicate
>> with Sue, and once the initial handshake was successful [if any], be
>> able to communicate with her again in the future).
>>
>> There are various ways to do this. Some centralized, some decentralized,
>> some require people meeting each other and some don't.
>>
>> What possibilities do you see, and what do you propose?
>>
>> -- Lukas
>
> It's important to hold off on solutions when discussing a difficult
> problem. Human psychology is constructed so that when faced with a
> difficult problem, we are very quick to propose a solution (the harder
> the problem, the less time we think about it). The consensus that the
> psychology community has come to is that it is better to fully discuss a
> problem before proposing solutions.
>
> Reading this original email, I am confused as to what a "connection" to
> "establish communication" is, since as far as I am aware the FreedomBox
> has no software written with it in mind as of yet. I can only assume
> that everyone reading it has filled in the blank with what they think
> those words mean, but if we all do that we fall victim to the
> transparency fallacy, where we all assume everyone can see what we're
> thinking. Before we can converge on a solution, we have to converge on a
> problem.
> So, what specific use case are we all thinking of here? Lukas, could you
> start us off by describing a use case you were thinking of?

The initial problem is a very general one and, as far as I understand, 
already got separated into two problems in the course of the discussion:
1. Discovery .. the problem of finding communication partners
2. Authentication .. the problem of making sure the communication 
partners are who they seem to be

The short example in my initial email, which was my attempt to provide a 
use case, is very general as well. I didn't want to constrain the amount 
of different ways of tackling the problem.

I'll try to provide a use case from user perspective with details from a 
more technical perspective:
1. Alice wants to be able to send Bob a message.
1.1. Alice wants to be able to discover Bob.
1.2. Alice wants to be able to authenticate Bob.

I want to make sure it is understood that this is not a concern of what 
is communicated (mails, messages, friend invites,..), but rather of 
creating communication channels between FreedomBoxes.

I hope I understood your request correctly, and my answer clarifies the 
context of this discussion.

-- Lukas