>>18893Yeah, its a really worthwhile project. Note that when it comes to other projects like Mastodon that share ActivityPub federation compatibility, its not like they're all "full" nodes of each other with all features or even most features of each - they're all individual projects (ie mastodon/pleroma is for microblogging, Friendica and diaspora are macrosocial, pixelfed is Instagram-like, Peertube is YT style media hosting and livestreaming etc). So yeah, you can be on a PeerTube instance and when you post a new video have a message go out that goes to supported Mastodon instances as a toot (aka a tweet in the mastodon network), but they're not exactly universally account+feature compatible in all cases. Thus, having a separate Peertube account on a server somewhere (one you run yourself, or sign up on an existing) is often a good idea as opposed to trying to use it through an existing Mastodon one, especially one with what seems to be a restrictive whitelist.
Though yeah, certainly give it a look. Peertube itself is pretty great, with Owncast a nice alternative for more livestream focus (though PT supports livestreaming too).
>>18897Dealing with those actively involved in spam, aggressive scraping of an API or the like is fine, that's what those tools are in my mind designed to solve, not making preemptive, often tenuously based, ideologically minded judgments that restrict your users as well.
>data mining or scrapingThis is the sort of thing that really is best handled in the code of the project itself and APIs to access it, with built in permissions. Sure, there can be work arounds but its a lot easier to spot someone and/or a horde of bots trying to grep it all "manually", thus evidence of bad behavior. Some projects have better support than others.
>spamSpam is an easy one to police by bad actor's accounts or if bad enough a whole instance, but again its a reaction to actual ongoing behavior and is different from what I was talking about
>Differing content, rulesets, defederationThere's nothing wrong with having differing levels of federation with different instances. Its natural to prefer those with similar content and rulesets for things like having it in a "neighborhood/universe" timeline or recommendations or whatnot. Last I checked with most projects there is at least a reasonably granular permission system, the problem I describe as poor administration combines using the highest level ban which impacts all your users in needless situations. For many fediverse platforms of which I am aware, an admin can utilize other settings to do all legitimate forms of protection of one's instance, all without impeding user choice. You can make sure that a "bad" node's content is not mirrored on your server, that its content never shows up in your public/universe timeline, any sort of suggestions or friendly instance's users content list etc… for all intents and purposes, it doesn't exist with regards to your server…except it doesnt' prohibit your users from "going over there" to seek out content, or be friends/follow/add someone with an account on that server, or anything else that should be up to the user .
Putting aside admin rationale for a given decision's legitimacy or lack thereof, if you're trying to implement an open federated network, especially one that is trying to grow and strike a blow against centralized, proprietary mega platforms, than the minimum imposition on user sovereignty and functionality is ideal. Its entirely possible for this to coexist with each node's admin enforcing local rules and favoring more open federation with similar and compatible ones, it just takes (ideally improved technical solutions as well, but in absence of that) admins who are not on a crusade to purge or isolate "problematic" content, instances, or individual, force that view onto their users, damn the effect on the network as a whole.
>>18900>That kinda necessitates an approximate synchronization of rules to be sustainable right?What do you mean? I'm not sure the other party involved here in the synchronization. There's no need for the other instance on which the user(s) in question is registered to implement similar rules or 'agree" to them within their purview.
>Defederation works both ways tho. If those sites didn't defederate over such things they'd actively made fedi as a whole worse, necessitating a defederation anyway. Do you really want to interact with those guys?The argument of "making the fedi as a whole worse" I don't think applies because there's so much potential for differences of viewpoint or context, especially when it comes to ideology and the like (as opposed to spamming and other behavior that's more but not entirely universal- for instance, Subcontinental social media engages in a vast amount of "good morning" posting and other posting that most in the West could see as spamming ).
Ultimately, its an issue of looking not only at one's own instance, but also at the conceptual basis of the federated platform as a whole. The point of federation, rather than just decentralization, is interoperability. People being able to utilize an account from a given node to contact, view, and otherwise connect with all compatible nodes. Any decision that limits this must be done judiciously, for good reason and in an even handed manner to put the minimal restriction on the federation benefits. Furthermore its also worth looking at the dichotomy of who's making such decisions as well as who it affects. Especially for social media-related projects decentralization is for many reasons favored over distributed models for many reasons bu aside from discoverability, the other major factor is that most users will NOT run their own instance. Having the option to do so is great for decentralized/federated projects, allowing those who are technically adept and have the resources (fiscal, knowledge, time, interest etc) can create their own be it just for themselves and friends or a public facing node, but the vast majority of users don't fall into this group. They'll sign up on whatever nodes that have rules they can abide by and that offers community and features they find attractive (userbase, technical features, community themes or focal points etc…). This means that the decision of the admin can potentially affect many and must not be done haphazardly or with a crusader's self-centered zeal. Its not just a matter of if you the admin want to interact with "those guys" but especially if you take the maximal option you're making that decision for all your users as well. Note that your can protect your own server just as well without being the equivalent of a parent saying "I don't want you to hang out with the people who live in the other side of town, they're all degenerates over there. ", so it isn't as though its some necessity to keep you own server from being overrun by whatever boogeyman you feel is most objectionable.
>That's not really comparable, the gated fedi experience is less pleasant than normal fedi so people usually leave to more open sites. Importing and exporting means you can even move to an instance that isn't federated. Also self hosting is a thing.
Besides, what you described has been the case for a decades now. But sites that don't go to that extreme succeed over the ones that do.
Unfortunately, things don't quite work out that way if such behavior moves beyond the absolute fringes. Sure, a couple of people keeping extremely restrictive nodes on the periphery isn't going to affect anything, but when it grows into a practice that leads to fragmentation that threatens the entire network. Regarding importing / exporting, this is a good function and I am glad it is available (though its actual technical detail varies depending on the particular fediverse project) but it is meant to be a relatively infrequent "moving house" feature and doesn't really serve to counter fragmentation/bubbling by bad administration issue as described because you're trading how people view your "old neighborhood" for your "new one". Hell, I've even seen attempts (at least with Mastodon ) to refuse to allow users exporting from "objectionable" servers to sign up on yours by way of looking the "forwarding address" from your old account! This is not a weakness of import/export, it just isn't meant to deal with this issue in general, but it still has a lot of great utility for other situations.
If we really want to to move away from existing, centralized proprietary social media to alternatives like fediverse projects, then these things need to be taken into account beyond just an individual node's behavior. This will mean a lot of individuals, including those who are less technical in interest and skills will join fediverse project servers; something like importing/exporting may be confusing and/or at most a last resort. Imagine you're a content creator and you're on standard proprietary social media like Twitter and YT; you know that every user on those services are going to be able to see your tweets and videos if you haven't gone out of your way to block them specifically. Now ideally in the Fediverse, this is also the case, with the understanding that it takes a bit longer for things to "spread out" depending on the particular project, nodes each with their own local, mirrored, and universe/neighborhood connections to others; a willing if minimal compromise for an open protocol. However, when the behavior I describe gets involved, that is no longer the case. Any user signing up on one instance is essentially picking prefab "allies and enemies" based on the admin's bias, so a certain part of the network (known to you or otherwise) will just be cut off. This isn't great for many reasons, when certain nodes may make a shitlist for shallow reasoning at best, dooming hundreds, thousands, or more users to not being able to interact with those from a "bad" place (on the most severe block tiers, as opposed to some others that have been discussed). Saying that locked down nodes with bad admin behavior will be sorted out is a bit like expecting "vote with your wallet" to solve major systemic problems. Even with relatively niche userbases and certainly as one attempts a federated open alternative platform to centralized big corp ones, a lot of people don't have the knowledge in the first place, don't care, or don't prioritize it high enough to leave on principle especially if it doesn't affect them directly in a tangible way…yet. That doesn't make it "okay" or an authoritative, knowledgeable polling of the user base on the issue. While good admins and instances can certainly help, it doesn't necessarily cancel out the bad behavior of bad ones which can have a wider reach; and that's all before you have ideological bubbles made into technical ones!