We, the technologically literate, are not the target audience. Client-side filtering requires awareness of client-side filtering. How many Reddit users want that ? Facebook users ? Oh, maybe they want to fine-tune their scoring rules library or tweak they bayesian filters - so much fun ! And lacking social intelligence, client-side filtering fails against organized harassment. By the way, who is liable for infractions to the local law ?
Enter the party, meet regulars, meet new faces, enjoy a drink and don't even think about safety. Sure, you can manage it yourself for a birthday party at your home but at larger scales even the wild rave party thrown in an abandoned hangar out there will have some form of institutionalized enforcement, even if they are not corporate security.
Yes, that means that successful large public places tend towards bland - the same reason why successful edgy underground dives are small affairs and their character can't survive growth.
There's no reason a private/paid service addon to a client can't work for filtering. Been thinking similarly for Mastadon, in that you can pretty easily have a gated experience, and get top ranking for Android/iOS client use with a monthly service that includes filtering with relatively sane defaults, but toggle-able filtering/reporting.
You can use fdroid, webapps or side-load the "full" unfiltered client but the experience for the normies would be that paid app experience. And I say paid app to avoid the pitfall that is advertising fed results.
> Client-side filtering requires awareness of client-side filtering. How many Reddit users want that ? Facebook users ?
Facebook[1] and Reddit[2] both have a block user feature. It's a very rudimentary version of the killfile that usenet clients have. If users didn't want features like that, then they wouldn't be available on either platform.
> client-side filtering fails against organized harassment.
Usenet killfiles were advanced enough that you could easily filter out messages even from organized campaigns. For example, filtering it based on a number of conditions such as group header, from header, a unique header, certain keywords in the body, etc.
> We, the technologically literate, are not the target audience.
Highly disagree. It's the early adopters of ANY platform, hardware, or whatever that drive real organic demand and adoption. And those early adopters ARE exactly that: technologically literate.
Alienating that group is how you get relegated to the "Digg v4" area of history.
> How many Reddit users want that ? Facebook users ? Oh, maybe they want to fine-tune their scoring rules library or tweak they bayesian filters - so much fun !
How many Facebook or Reddit users even HAVE the capability of client side anything? The big threads are that Reddit's killing most API access unless you pay a pound of flesh. And Facebook does enough horrible things in their coding that makes it impossible to even screen-scrape. The text "Sponsored" shows up as shittery like "essprnodo".
For those gated islands, you'd need their permission to really do any sort of client side anything.
Places like Mastodon make blocking trivial. Ive shared that link, and no, you don't need to know about Bayseian weights or whatever new mod scheme you want to make. Its there, the initial functions are easy-peasy, and the API hooks are there to add whatever you want.
> And lacking social intelligence, client-side filtering fails against organized harassment.
And that already is a major problem on the corporate side. Reddit calls it "brigading". HN has chill-down scripts to defuse flamewars. Twitter, Facebook, and Reddit all have mass-reporting as harassment. There's even current 4Chan threads to organize just that. Sure seems like an unsolved problem, not just "unsolved for federated systems".
> By the way, who is liable for infractions to the local law ?
Irrelevant. Someone in another area could read content that is illegal to them. That's *their* responsibility, not mine. As long as I follow the laws in my jurisdiction and where my server is located, I'm fine. And 17 USC 512 covers me as long as I do good-faith DMCA process.
> <example of comparing physical real world with raves/alcohol/drugs with online "dangers">
And in your latter half, your other problem is that you're equating online "safety" with in-person bodily safety with alcohol and drugs. Those 2 aren't even remotely the same, and its laughable that you'd try to compare them.
If someone is coming on to you in a sexual manner, and your "No" isnt effectively heard or acknowledged, you have a big problem. If someone uploads a bad image (say, animal torture, for something abhorrent), you can block the user, the server, report the user to their/your server admins. You can also choose not to show images. It's terrible, but physical assault is no comparison to a bad image or terrible text.
Enter the party, meet regulars, meet new faces, enjoy a drink and don't even think about safety. Sure, you can manage it yourself for a birthday party at your home but at larger scales even the wild rave party thrown in an abandoned hangar out there will have some form of institutionalized enforcement, even if they are not corporate security.
Yes, that means that successful large public places tend towards bland - the same reason why successful edgy underground dives are small affairs and their character can't survive growth.