Re: [GNU/consensus] [SocialSwarm-D] PS: Social Network 3.0 - Preparing a 30C3 workshop

[Melvin Carvalho]

On 30 April 2013 15:21, Klaus Schleisiek <address@hidden> wrote:

Am 30.04.2013 12:32, schrieb Melvin Carvalho:

>
> Could you elaborate on what properties "self determined data storage" should have?

You determine yourself, where your data is going to be stored and in what form
(encrypted or not). Which will probably mean in the future: You have to pay for
it pretty much the same way as you pay for your internet access today. See:
unhosted.org.

Ah I see!  Yes I've been aware of this project for some time, tho it's many about client side web apps.

I'd encourage to also look at Tim's work, on which unhosted is partly influenced:

http://www.w3.org/DesignIssues/CloudStorage.html

What I think you need is:

1. CRUD operations on your data

2. Ability to store in the format you choose (either an existing standard or encrypted)

3. Ability to use the identity you choose

4.1 Ability to store your data in a data store you choose (data as a commodity)

4.2 Ability to store data in multiple locations, I may store social data on my friend's server, but financial data in my secure vault

Now you will be promised all 5 of these by every data store out there from Microsoft to Dropbox to FLOSS

People will implement parts of this, but still be wary of lockin through the back door e.g.

With (1) you often have ability to change but not add.

With (2) you are often forced into a format, often restrictive, encryption is rare, but you can encrypt on the client to an extent

With (3) this is the main danger.  You are normally cajoled into the Trusted Third Party system, sometimes covertly by making you use email. 

With (4.1) often it's choose from our list of approved providers, some are included some excluded

With (4.2) It's rare to see this, most force you into one location for everything

The project I know of that is closest to passing all 5 tests so far is on of Tim's students side projects at MIT http://data.fm/

We need to get serious about data, so I'm very glad you posted this! :)

 

--
~bit