Re: [Social-discuss] Who owns the data once transferred?

[scorbett]

> The above license would be dangerous for user privacy in this medium.
> I'm wondering, what sort of method do you all think would be ideal for
> protecting data? In the end, as we all know, there is nothing anybody
> can do to prevent data copying and sharing, unless we try to implement
> some nasty DRM system, which I'm steadfast against. Decentralization
> comes at a cost, no? :P

I think this notion is damaging to the idea that GNU Social should be
purely P2P; if by sharing data users are pushing it directly to others'
machines, they essentially have offered that data up for modification,
storage, etc. by their peers, unless we use something messy like you said.

This is why (from how I remember it), Mattl's initial vision of the
project was that the network would consist of nodes which would host and
serve data for a small number of users. Not only would this allow users
slightly more control over their data (since it's not being published
anywhere but that one server), but it gets rid of a few problems such as
inconsistent availability of a user's profile and a lack of a single,
canonical version of a user's data.

However, users in this scenario will still be at the mercy of whoever is
running their server. Ideally users would only use a server run by someone
they knew and trusted, but there'd still be no *guarantee* of control over
their data.

I think this evinces a problem holding back a clear specification of the
project. I'm pretty lost as to what form GNU Social should take. We've
seen a lot of good examples of technologies that could be applied, as well
as other networks that have accomplished similar goals; however, nothing
has really been decided on for GNU Social. Are we creating something along
the lines of existing projects like Elgg or Crabgrass? Are we creating
software to unify arbitrary social networking services? There have been a
lot of different views on this subject presented on the list, and I feel
that we need to start pinning down specifics, otherwise we'll never be
able to get to hacking. Can someone help dissipate my confusion?

Thanks everybody,
--sean corbett