[Freedombox-discuss] [Fwd: Re: Freedombox's pursuit of perfection undermines its goals]

[Spectral Emanation]

On Sat, Oct 22, 2011 at 11:43 PM, Weaver <weaver at riseup.net> wrote:
>
> I think we need to avoid tailoring a product to a specific market. Just
> produce a product that enables secure, private, decentralised
> communication.
>

I think we do have a specific market. I've been trying to keep focused
on the use cases that Eben Moglen was describing in his speeches,
which _mostly_ is getting stuff off the remote servers out of the
user's control, and on a box in their house that doesn't take an admin
to run. That was the banner that FreedomBox Foundation was launched
under.

Things that can be ripped out of the Cloud and put on a FreedomBox
right now are: blogs, StatusNet (Twitter equivalent,) wiki and instant
messaging. There's just no installer to make it easy for a regular
person, and the "box" isn't 50$ yet.

Many people on this list are concerned about the fact that centralized
SSL CA's aren't trustworthy, or how to prevent DNS hijacking, or how
to exchange PGP keys. It's not a waste of time to talk about these
things, because they are problems with the security models of all the
things FreedomBox needs.  But the solutions to all these things aren't
readily available to be dropped right on a server.

If the minimal goal of FreedomBox is to stave off government/business
siphoning of nearly everybody's data on tools like Gmail, Google talk,
etc. then just getting the server in someone's house is pretty good
for that. But much(most?) of the conversation on the list seems to be
about the stronger security stuff that doesn't exist yet. If someone
needs "real" security right now, the tools are already available and
they are just going to have to learn how to set up a secure box and
use it. This is all the more reason for FreedomBox to ignore this
"market" for the time being. Easy to use FreedomFighterBox is out of
reach right now; easy to use FreedomBox to get people's crap out of
the cloud IS within reach right now. Stuff necessary for a shipped
FreedomFighterBox is also necessary for FreedomBox in the long term,
but it's hardly necessary in the short term to just get out of the
cloud as much as possible.

In my basic estimation the only significant software "missing" for
launch is the installer; and a way to get around browser warnings
about self-signed certificates, which possibly could be
"good-enough-for-now" solved by one of the two prevalent network
notary systems. Email is clearly too problematic to solve before
launch, and Facebook replacements aren't done yet.

tl;dr: the FreedomBox that won't get freedom fighters killed isn't
doable in the next year, and the FreedomBox that does most of what
_Eben Moglen said it should do_ is.

Three cheers for do-ocracy, but if a project is launched with specific
goals, those are the standards by which success is measured.

Respectfully,
"Spectral"