> > Now, the idea of FBox has evolved into something which has mesh > networking (even not requiring DNS), has anonymisation as well as > serving as a separate 'Internet'. While there is no doubt that these > ideas are interesting and useful, there should be a progression of > stages through which the FBox will pass through to reach this goal -- > and not try to do everything at once. This has to be clarified. From > the discussions on this ML, to me it seems like at least a network > protocols and privacy/identity distribution WG should be created. > > Once it has been agreed what will be there in the _first_ release of > FBox, setting up WGs will become easier.
I fully agree that things should be done in steps. Furthermore I am quite confused as to why one would want the FBox to be essentially the identity of the user himself, and not simply be a user-managed repository of services for the user, which can and should be easily replaceable/duplicable. To me, the architecture of the software should be one that is heavily abstracted, with everything as a service with API entrypoints, and service configuration being automatic (think zeroconf). Therefore, necessitating thoughtful consideration as to what/how/where that core user-centric config data should reside. I think we need a working group for that kind of skeleton work which would help into bootstrapping services on a stepwise basis. -- henri.tel _______________________________________________ Freedombox-discuss mailing list [email protected] http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/freedombox-discuss
