3 ms·
If the network can be adversely affected by a "muh righz" device then the network's threat model is shoddy. Taking away freedom to prop up a badly engineered pr
by EvanAnderson 6y ago
If the network can be adversely affected by a "muh righz" device then the network's threat model is shoddy. Taking away freedom to prop up a badly engineered product isn't fixing the bad engineering.
The Internet is a good example. The threat model has been far too trusting, historically. We're paying for that in a variety of different ways. Burning it all down and starting over is impossible, so we're stuck in a mess. Maybe we can do better in the future.
- oneplane 6y agoIndeed. I would perhaps formulate it slightly differently but it is what it is. This is also something that feeds the 'it used to be better back in the day' feeling, because some aspects might actually have been better because too many possible threat actors back then wouldn't take internet seriously and as such weren't an actual threat. So it wasn't safer, it was just less-attacked. As a result where was less pressure to make hardened clients and servers, and as a result of that, it meant that things like digital signatures were extremely optional (and computationally too expensive to include for the sake of it). On the other hand, it's also the openness that brought its success, and may very well cause its downfall. (that said, nobody has been able to come up with a worthy replace ment so far) Having no single owner makes it better in that regard, but also worse.