Re: Federation protocols

2013-05-31 17:56 skrev Darrell Prince`:
> By the same token, I want to OWN my data. I don't want 20,000 profiles 
> and
> passwords. I want someone or something to ask me for permission to 
> view
> certain information, and say yes or no, and have it not be permanent. 
> I
> want retinal scan and access to everything I ever worked on, and a 
> good
> filter for it.

I would actually like us to go away from such a limited view on the 
storage and control of data. I hope I misinterpreted your request 
somewhat, or at least that this is not a widespread view on data 
control. But your phrasing of "have it not be permanent" begs of me to 
clarify the impossibility of guaranteed temporary internet data storage.

One probably doesn't really want to _own_ one's data, as in control 
every instance of it (including destroying one's "temporary" copies). 
That's the same mistake the music and film industry are making in their 
futile battle against filesharing.

The interesting aspect is only related to the _initial_ release of data 
- to whom you are giving it. After that, any restriction on sharing or 
usage must be a matter of respect (why we're "friends" in the first 
place), national law (whether sharing personal data is allowed) and 
such. Because when _you_ publish something to _me_ - _I_ am fully in 
control over the bits and bytes that are stored on my machine.

These things cannot be regulated in the protocol, if we have a truly 
federated system. Whatever data is published and has reached my feed is 
up to me - not the protocol - to decide how (not) to store and use. This 
is simply because the protocol can't configure _your_ restrictions on 
_my_ server. The mass data analysis and privacy issues are far from 
fully solved in a defined protocol. Rather, it relies upon user 
interaction, by people only sharing data to those they Trust that data 
with.

(...not saying there shouldn't be a "delete" action defined...)

-- 
Mikael Nordfeldth
http://blog.mmn-o.se/
Xmpp/mail: mmn@hethane.se

Received on Saturday, 1 June 2013 06:23:25 UTC