[Freedombox-discuss] Kickstarter's initial goal was reached

[Matt Willsher]

On 28 February 2011 21:02, Thomas Lord <lord at emf.net> wrote:

You do seem to raise points that I like to reply to Thomas :)

> Fantastic but it raises this paradox I've been trying to
> wrap my head around.   Maybe its good to raise this earlier
> rather than later:
>
> We want to (quickly but also in a lasting way) distribute
> and decentralize personal computing and communication.
> We want it, with the help of freeboxfoundation to get really
> big and very accessible.
>
> How do we prevent the freedom box project (both the software side
> and the larger "get it on the shelves" side) from *itself* becoming
> exactly the kind of centralized point of control we want to get
> away from?

> I am worried about problems like these:
>    ~ shouldn't we try to avoid having too many users get
>      their software stacks from exactly the same entity?

Yes. IMHO FB should be the reference implementation - a catalyst for
other devices and integrations.

>    ~ shouldn't we especially be worried about too many users
>      getting automated or semi-automated updates from the
>      same entity?

Yes. If Debian's signing keys get compromised or a key package
maintainer likewise (although see my comment below) all FBs are at
risk.

>    ~ if all of the developers of the core stack are under
>      one umbrella, isn't there a non-ignorable risk of
>      overt or covert political or economic interference with
>      that project?

Yes. It may well be that some of the software is already compromised.
There should be defence in depth with parts of the FB defending others
from itself.
There is a wider issue of trust which might be a difficult one.

> Of course there is a flip side.   A risk of the opposite situation:
>    ~ suppose that freedombox development and production really
>      is distributed and decentralized.   Perhaps we should call
>      this situation "freedomboxES" rather than "freedombox" to
>      emphasize that while boxes of different makes and variant
>      stacks might inter-operate, there is no singular, privileged
>      source defining what a freedombox is or controlling what you
>      get when you buy one.
>
>      Suppose all that - then - how can average people who see these
>      products on a shelf in a store have any sense of which ones to
>      trust?

The FB mesh is self policing and won't communicate with bad nodes. How
does that work? No idea. Is there any work in that direction that can
be done in an automated way? IRL plugs can be pulled and routes can be
dropped. In FB land the end users can't be making that decision.


>
>> In the US, if you want to hire even a few people and pay some travel
>> expenses to get them to the right places... even if those people are
> so
>> passionate about the work that they're willing to work for something
>> like a minimal stipend instead of a "real IT salary", $500k isn't
> really
>> very much money.
>
> It seems low, to me, for the mission.
>
> In the name of decentralization, perhaps a worthy goal is to
> try to very, very quickly get to a situation where many small
> businesses can begin to (a) make money using freedombox technology;
> (b) enough money to pay for their own contributions to development;
> (c) and enough money to, if they choose, pay membership dues
> to the foundation.

I agree whole heartedly with this. Perhaps the goal of FB should be
boot strap these small businesses with the foundations of the end
vision. Provide a saleable 0.1 that delivers something of value to
moderately tech savvy individuals.
Although, techies aren't always salesmen and visa versa. And care
should be taken not to make it too US centric. But I agree with what
you're saying there.

> My bigger concern about the software side, so far, is that
> if we have too many "highest priority" goals it will either:
>
> a) never get done
>
> b) get done as a "proof of concept", unusable, unmarketable,
> and stuck in a perpetual state of "next year is the year it
> finally takes off".

I agree. It has to be a simple start. There are many, many good ideas
on the list and it's really difficult to pick between them. The
framework things will sit on I would like to see as the first goal
however that doesn't sell boxes. I think an easy to use, workable mesh
might be a good start. To wit:

- FB forms a wireless meshes.
- Small business sells boxes and sets up mesh for a community. Initial
problems of finding neighbours may need incentives. In the UK, small
communities in remote areas stand most to gain from mesh networks in
the early phases.
- Bandwidth from broadband links shared by those that have it.
Bandwidth can be limited as required by the users.
- Updates are release early and often allowing the users to gain new
functionality quickly
- If the applications bundle their UI in their 'package' (and I don't
me .dpkg, some other bundle) this can seamlessly be added.

All in all, for up take in democratic first world countries FB needs
to fill a market gap and expand from there.

m.