@ilpianista There are a few different approaches to this. And yes, I need to write A Thing.
You've also got a few distinct problems even in addressing the issue and making headway with someone on it.
There's ignorance on the issue.
There's /willful/ ignorance.
There's the question of "well, what would change your mind?"
There's the historical context.
There's scale.
There's power, and information's role in power dynamics.
#surveillance
#surveillancecapitalism
#privacy
#tootstorm
1/
@ilpianista There's a whole 'nother set of lines of argument about metadata, scale, and data as toxic waste. This is getting long, and others have argued this far more eloquently than I.
There's using metadata to find Paul Revere:
https://kieranhealy.org/blog/archives/2013/06/09/using-metadata-to-find-paul-revere/
There's the fact that the /useful/ part of most surveillance and stakeouts /is/ metadata: who is talking to (or seeing, or receiving materials from) whom, and when:
Bruce Schneier: Metadata = Surveillance:
https://www.schneier.com/essays/archives/2014/03/metadata_surveillanc.html
/29
@dredmorbius @ilpianista Here's one interesting point, by the way: *everybody* should route everybody else's traffic as much as possible (in a network like #tor or #gnunet, for example), not just for better decentralization and anonymity, but also to to make it more difficult to tell *when* a message has been sent/received and *how big* it was.
And that requires rather cheap and fast Internet connection… Unsurprisingly the laws are intended to make it quite the opposite! #ISPs
@amiloradovsky @ilpianista How do you deal with bandwidth, reciprocity, and/or abuse, under such a scenario?
@dredmorbius @ilpianista Ideally, the download/upload speed should be fixed, bandwidth throttled, if needed, or fluctuate randomly.
If we assume the traffic to be e2e encrypted, there's no way to tell what is in there. So we can just hope our node is used for good…
I don't know how to avoid the free rider problem now.
#GNUnet is actively researching all these questions at much deeper level than I can do here.
And, according to Zak Rogoff, there is some incompatibility of the desired properties.
@dredmorbius @ilpianista I'd say, such a networks are still mostly at the "active research" phase (as was the #Internet in the 70's), which is done on intersection of computer science and economics (both equally important).
I'm not sure if the principles of #BitTorrent are applicable here.
And, yes, #DDoS and censorship through overwhelming are some of the issues, if I remember correctly.
@amiloradovsky @ilpianista A key challenge is that if you're /both/ distributing /and/ encrypting traffic, you've got virtually no management left.
There are some systems that attempt to work around this. I've looked up these in the past, one is FAUST, and another is a fair, anonymous, queue management (I think) system. Let's see if I can track these down....
@amiloradovsky @ilpianista It's possible to throttle or QoS shared bandwidth, yes, but there's no protection on keeping a single source (or worse, a distributed source: DDoS) from saturating bandwidth.
There might be some means of setting up reciprocal arrangements or budgets, but those seem to get complicated fairly quickly.