[smcphile]

> I am partial to the idea of someone getting to decide, in advance, who they can see or hear. Letting the mob in, and then muting or banning after the fact can be exploited too easily.

I see that that could be done in a private group setting where everyone knows each other. In a forum for the general public though I have difficulty seeing how that could be done. Do you have any implementation details in mind?

[squibbles]

I don't know of an easy or economical way to prevent or discourage abuse of communications.

For social communications, the main problem is: How do I find people I like or trust? That is a tough problem given constraints.

For network communications (IP), practically any two machines can communicate with each other, which opens up a number of serious security risks.

[a1369209993]

> For network communications (IP), practically any two machines can communicate with each other, which opens up a number of serious security risks.

This isn't quite a correct characterization of the problem. The problem - rather in the style of your previous post - is that a given machine will listen to communications from any other machine, and often react to them. My usual go-to example is that most web browsers will happily execute completely arbitrary code from any random website they happen to stumble across.

There are practical reasons (eg opprotunistic scans for exploitable vulnerabilities) to keep machines isolated, but the root problem is not that your machines can send and recieve communications with arbitrary others, but that they choose to.

> How do I find people I like or trust?

I don't have a general solution to this one even in principle, but I've found it helps if you don't need to trust them. Proper anonymity or pseudonymity means I can have a perfectly comfortable conversation even in the depths of 4chan, and laugh in the faces of anyone who tries to threaten me.