Re: On privacy and cloud services

That is precisely what I was thinking. I would think there would have to be
a couple of layers of images(data structure images not jpegs) though; there
will be things that are accessed on a regular basis, and other stuff, that
might take a little bit to retrieve.

Nomenclature for certain data types is necessary, so this accomplishes
semantic web needs.


On Mon, Jun 17, 2013 at 4:48 AM, Michał 'rysiek' Woźniak <rysiek@fwioo.pl>wrote:

> Dnia niedziela, 16 czerwca 2013 o 20:42:33 Darrell Prince` napisał(a):
> > Self hosted also leads risk of data loss; few people have data mirrors
> set
> > up. No data loss has to be a key feature, as well. Would it be possible
> to
> > keep people's data in encrypted chunks on other servers and laptops ?
>
> I believe distributed, federated, self-hosted, peer-to-peer social networks
> (a'la Sneer and RetroShare) should be built around the concepts of
> distributed
> data stores, so that even when one node is not accessible, data from it is
> still acessible (to those with access, of course) as it's been distributed
> among other nodes.
>
> Think: BitTorrent with secure, public-private key cryptography layer. no
> explicit backups needed, as the stuff is always somewhere In The Net. Of
> course, what would be needed is to have it work automagically, and not on a
> "popularity contest" basis like BitTorrent.
>
> Each user of the network could decide how much disk space it can "lend" to
> the
> network for that purpose (and I guess that would also qualify how much data
> this user can put into the network; quid pro quo). This aspect would be
> similar to SETI@Home, only the shared resource wouldn't be RAM and CPU,
> but
> disk space.
>
> --
> Pozdrawiam
> Michał "rysiek" Woźniak
>
> Fundacja Wolnego i Otwartego Oprogramowania
>