I wish people would quit assuming my interest in decentralization is based on privacy concerns. There are lots of things that are much higher priorities:
• Censorship resistance.
• Resilience. An entire library of content/social graph isn't destroyed because of a change in corporate focus.
• Evolution. Protocols can change more when not under a single administrative domain of control, and in ways that meet the interests of user communities rather than productization committees.
• Flexibility. Since things are decentralized anyway, it's much harder for the system to ossify in a way that relies on one particular family of devices.
• Namespacing. Being able to 'reserve your brand' across all social graphs is an antifeature.
And by censorship resistance I'm also interested in resistance to the Content Cartel's censorship regime and their perversion of the government into a rent-collecting institution that exists to preserve an archaic business model based on the assumptions of hard-to-reproduce artifacts.
"The OS is also guilty, because of the amount of freedom given to devs, and the lack of control in the store...."
"Apple are closed and hostile to repair but..."
@Azure I can definitely see a lot of advantages to decentralized networks and I generally prefer them, but we also need to tackle the disadvantages like not being particularly resilient to brigading and harassment being an issue. The latter might be solved by having a database of people known to have engaged in targeted harassment or have sent threats to others that an instance admin could optionally look through and prevent bad actors from communicating with their instance.
@Azure I think as far as brigading goes its a matter of instance admins having policies in place for people who might be particularly draining to the network (such as celebrities or politicians) and ensuring their are plenty of moderators with a database of each instance and the amount of moderators they have on hand compared to regular users
@bluetechgirl I agree 100%. Though the details of exactly how to do that are something I'm not sure on. I like the idea of distributed allow/deny lists people can maintain and share, subscribing to those from folks they trust. I've heard proposals as well on ideas for establishing a cost (perhaps computational) that people can demand an untrusted stranger wanting to message them must pay to get a first message to them. (That becomes harder nowadays in the age of botnets, though.)
@Azure I think having a system that prefers people that have been on the such a database for longer as well as verified admins might offset manipulation caused by bots a little. IDK for sure how it would work in the long run, but ensuring places like this remain properly moderated is very important. So how this should work definitely needs to be a priority.
@bluetechgirl I agree. The other thing you want to make sure of is that it's friendly for new users who don't know anyone on the system but want to meet them. And, you know, people making new friends generally. There's a tension there, since making it easy to read the writing of and talk to total strangers also makes it harder to block out harassers since they can cheaply make new identities. Some ActivityPub developers are prioritizing exactly this, among other things, with ideas like giving people multiple inboxes some of which might be public and some of which might not be with different restrictions on each. (e.g. you could simply have a hard limit on the number of messages from strangers you'll accept in a given time period.)
@Azure Yeah, you don't want to block normal communication in the process. I guess handling moderation long term is just something decentralized networks will always face and have to address creatively. I think a blacklist combined with enough moderators to maintain it and protections against bots will be better than what we have now.
@Azure We should also have more options to be able to let the user decide how much harassment is a risk. Some people will have more need for anti-harassment measures than others. So having each user set their own limit to how much messages from strangers they want, how much verification is needed to talk to them as well as a supplementary instance blacklist that works in addition to the instances the admin already blocks will be beneficial to everyone.
The Vulpine Club is a friendly and welcoming community of foxes and their associates, friends, and fans! =^^=