News for hackers. Stuff that matters.
Re:Pleroma 2.0: a free, federated social networking server b... by seemslegit on Sunday March 08, 2020 @04:35PM (#22520426)
Me hearing about the fediverse for the first time:

"Man this thing looks really promising, just think of the potential of a social network unhindered by the dark machinations of the googtwitface !"

Me after ten minutes on the public feed of mastodon.social:

"So do the furries post for the sex workers or is it the other way around or what ?"

Re:Pleroma 2.0: a free, federated social networking server b... by stevenicr on Sunday March 08, 2020 @06:59PM (#22521270)
I've been thinking about these kinds of things for some time and starting to talk about it.

I think we need an open source set of 'bouncers / blockers' - that people can use, perhaps best as a chrome / ffx extension? That blocks content from various portals or individual posts or posters that you may not want to see.

Different people want to censor different things, and some want to censor the same things less and some more.. so by have an open source set of blocker bots people can discussion the blocking lists.. people can pick and choose what they see / not see..

That's on a basic beginning - I could see them also getting into 'unsubsbrcibe to stuff from server A and mastadon B' can take it further..

Since cloudflare came out as censor friendly - I've given up hope the big players will be all 'open info utopia' ever again. I think the best we can hope for is to get individual blocking bots and open up the web more. Otherwise we will have more censoring across the board and will succumb to whichever lowest common denominator of whoever gets to be in charge and make new rules.

This way some people can choose to filter sex workers and not furries and some can choose to censor furries but not sex.. Of course I am sure many people will choose some broad auto-block and stick with that, making their internet more like fbook's daycare.. but having transparent options will certainly have many people turning off the 'safesearch' and explore the whole world, or larger parts of it.

You are right in that we may need more tools for individual users to take control in this expanded universe.. but things like 'peter's ad block list' kind of things can be more beneficial than depending on MostPopular Mastadon Server being pushed to choose what to block for the masses and what to let through and the whack-o-mole that tends to create.

If they are open and clone-able / editable - some might want peter-and-pauls-block-more lists.. and some may like furfriendlyButnotPron lists.. and I hope many more options for choose your own censor.. rather than having them chosen for us as most have succumbed to.

My current random thoughts on the subject - hope others chime in to make it better and more of a possibility instead of just napkin notes.

Re:Pleroma 2.0: a free, federated social networking server b... by philpem on Monday March 09, 2020 @05:01AM (#22523990)
I low-key love that blocking furries is the example here.

I used to be a part of that community and it's got a hell of a poor signal to noise ratio and a major broken-step problem. There are some good ones, but all the others ever seem to want to do is fight.

I'm amazed they can arrange to get out of bed in the morning.

Re:Pleroma 2.0: a free, federated social networking server b... by nirui on Monday March 09, 2020 @03:19AM (#22523560)
I was having my 2 cents thought on the topic as well, my conclusion is: A shared blocking list (Or a blockchain system, hehe) can only be used as the last resort.

Systems like such already been developed, for example there is one for Twitter made by third-party called "Twitter Block Chain" (https://github.com/satsukitv/twitter-block-chain).

Yes, it can block "unwanted" posters and bots when a correct list is given, but it could also sometime block innocent accounts when somebody is been mistakenly added to the list (Or that somebody just simply changed their behavior for the better).

For such system to work, the maintainer of the list must be trustworthy, which is the hardest part and almost always falls apart.

Add to the injury, it can also be very hard for others to detect mistakes sneaked into the list, as the list may contains more than 10,000 items, makes it impossible to be completely verified manually.

And what to do if a mistake is indeed discovered? Maintain an unblock list to automatically request the user to unblock? Now user has to maintain their own "Yes I manually blocked this guy" list to counter act the automatic one.

My suggestion is: Yes, you can use a shared block list if there is no other way (The Twitter case). But since Mastadon is open sourced, maybe let's try something else first while still can?

Please give this a little thought: Can a system be designed in such way that makes it really hard for their user to accidentally come across some unwanted posts (without having to block the poster)?

Re:Pleroma 2.0: a free, federated social networking server b... by endgame on Monday March 09, 2020 @06:09AM (#22524216)
Don't forget that blocklist admins can be jerks, too. If I run a blocklist that gets popular, I could make people I don't like disappear without anyone really noticing.
Re:Pleroma 2.0: a free, federated social networking server b... by stevenicr on Monday March 09, 2020 @11:06PM (#22532404)
Good point! This is one reason why I a want a few portals with the bots and showing discussions about them, easy option to fork / clone / change.

Would be nice to have something check for updated bots and discussions about them and put a flag in the browser bar maybe every two weeks -

Some bots will get better over time - but some people who liked the old one may not appreciate the changes and want to re-hatch the old one with a merged list minus X and Y but leave Z.

Re:Pleroma 2.0: a free, federated social networking server b... by stevenicr on Monday March 09, 2020 @11:33AM (#22526191)
I had not heard of the twitter block list - thanks for the info.

My thought on the blocker bots things in this thought process is that a few portals could spring up with the various bots listed and open discussions / comments about them will be available - and it should be a one-click to clone and rename / change.. so that people are encouraged to create new ones.

Having open source blocking and discussion with one click clone and change will hopefully create lots of transparency, and give end users for options for finer grained control -

The open discussions should have shape the various bots - something I have not figured is a process to auto-take and another bots updates and remove all instances of x and y terms.. and make Paulines-Copy-bot auto update when Pete's now block lists are added.

I too would prefer not to block people and servers and instead just hide by topics as a first line of defense.

Like mashable on fbook - I'd like to keep them in my feed, and block the stupid things they post about trump.. if I don't get a way to do this, I'll have to lose mashable completely.

I think the discussions on the bots would be educational for people as well. I myself have read articles that discuss some of the edge cases of bans for things like the ISPs in the UK and the health sites and such.. but I don't think most people see that kind of news.. and some may benefit from seeing discussion about other things that are unintentionally, or intentionally blocked.. and make decisions on the info and controversy instead of taking the one size fits most censorship each country seems to have.

Re:Pleroma 2.0: a free, federated social networking server b... by seemslegit on Sunday March 08, 2020 @08:59PM (#22521951)
I don't have anything against either furries or sex workers being there, just that between this and half of the public feed being in languages other than English the "what is this all about ?" experience is rather deterring.

Two basic things they could do is add some categorization and have users opt-in to topics for their public feed and limit the posts shown to English plus the languages that can be deduced from the user-agent data - it is after all one of the few legitimate purposes it can be used for.

Re:Pleroma 2.0: a free, federated social networking server b... by yborg on Monday March 09, 2020 @03:30AM (#22523611)
Mastodon client can filter on language. Hashtags are a thing, most clients can search/filter on them.

The pond is tiny, so certain non-mainstream communities that were driven out of other "real" platforms (i.e. platforms that are trying to sell you something) are a much higher percentage of the feed. But it's more or less just like Twitter, loudmouths tend to dominate, and 95% of the feed is garbage.

Instance homing is really the biggest problem Fedi has right now, instances are hard to characterize without camping them for some time, they're mostly small hence the loudmouth problem, and stability is an issue, they still come and go frequently. One of the problems with anarchy is that it's hard to manage.

Re:Pleroma 2.0: a free, federated social networking server b... by kick on Sunday March 08, 2020 @09:56PM (#22522237)
Why not just find people you like on it first and then use it? Joining a social network you know nobody on seems weird.

Also, the Public Timeline is your local instance's. My local instance has no users that speak different languages than I do, so the timeline is without content I don't understand.

The Whole Known Network goes way too fast to be of use, but it too is mostly filled with English speakers (albeit mostly non-desirable sorts).

Re:Pleroma 2.0: a free, federated social networking server b... by lainsoykaf on Monday March 09, 2020 @06:40AM (#22524321)
Mastodon.social is sadly one of the worst instances to join. Onboarding / Selecting a good instance is the hardest problem for the fediverse.
Re:Pleroma 2.0: a free, federated social networking server b... by grishka on Monday March 09, 2020 @12:45PM (#22526852)
The problem isn't even that. There's no global account catalog you could search. Thus, you have no way of knowing who of the people you know is on there already. Asking everyone if they know about Mastodon and which instance they're on isn't the answer. I have an idea about making a DHT out of instance servers, but at this point it's purely theoretical. If it does work though, you'll be able to easily import your existing friend lists from centralized services. Could use either their APIs or — better — the GDPR-mandated export feature.
Re:Pleroma 2.0: a free, federated social networking server b... by FalconSensei on Monday March 09, 2020 @01:58PM (#22527783)
> Asking everyone if they know about Mastodon and which instance they're on isn't the answer.

Ironically, the only way to do that now is posting on twitter, asking people to reply with their mastodon accounts

Re:Pleroma 2.0: a free, federated social networking server b... by sneak on Monday March 09, 2020 @07:02AM (#22524414)
Now think about all the places where that doesn’t happen, and think of all the censorship that is being imposed on harmless true self-expression.

That’s a feature, not a bug. Being able to post without fear of getting ToS’d and losing your entire audience is one of the major benefits of decentralized systems.

Now we just need to figure out how to solve for overzealous admins. I think the default for AP should become “bring your own domain”, even as a leaf user, so that you can seamlessly migrate instances if, for example, your web host or local admin takes issue with something you say.

Re:Pleroma 2.0: a free, federated social networking server b... by rapnie on Monday March 09, 2020 @02:31AM (#22523410)
And when choosing your following wisely your timeline will just contain what you want it to (e.g. not a furry in sight). This is not so different to Twitter.
Re:Pleroma 2.0: a free, federated social networking server b... by seemslegit on Monday March 09, 2020 @07:01AM (#22524401)
This isn't a problem when choosing who to follow, just with the first impression someone gets of the public feed.
Re:Pleroma 2.0: a free, federated social networking server b... by hkt on Monday March 09, 2020 @03:00AM (#22523508)
Same here. After the subsequent ten minutes I concluded that moderating a distributed social network is likely to be impossible.

I'd be pretty open to a non-federated social network provided it was a co-op or a foundation running it along a wiki style funding model (eg annual mass begging).

Re:Pleroma 2.0: a free, federated social networking server b... by sneak on Monday March 09, 2020 @06:59AM (#22524389)
Nobody who you aren’t following can put things into your feed. That’s all the moderation you need; you need not be concerned with the content of the timelines of those you do not follow.
Re:Pleroma 2.0: a free, federated social networking server b... by hkt on Monday March 09, 2020 @03:09PM (#22528626)
Except there are often default feeds of public posts from the community. So, actually..
Re:Pleroma 2.0: a free, federated social networking server b... by seemslegit on Monday March 09, 2020 @07:28AM (#22524536)
Moderation of one's own feed is not the problem here, the problem is that a community like this is likely to remain perennial if the first impression of it is wierding out the average would-be user.
Re:Pleroma 2.0: a free, federated social networking server b... by FalconSensei on Monday March 09, 2020 @03:16PM (#22528688)
As someone already mentioned, mastodon.social is not the best entry point. Which is ironic, because it should be. But even if you find a cool instance, it's hard to find people there to follow or conversations you want to join, because you can only search for hashtags. #and #people #dont #write #like #this