Vicinus

The accounts are across many instances …

Good point. If they have a shared identifier (IP or something) than I could see the block list working. If not, and it’s probably not, then yeah, I don’t think the block list would be able to handle that spam.

I wonder if there is any approach that could. Different origins, different entrances, and different targets, that’s a difficult one. It’s not my knowledge base, so maybe something exists that could handle it? I’d be curious to hear about it, probably pretty cool.

Maybe machine learning on interaction rates (sign in to post timeline)? I’m assuming it’s automated though…

Appreciate the feedback :)

The difference between blocking and purging is that a blocked account is still residing on the instance (even more so for user blocks instead of admin blocks). As the data is still there, it opens possibility for legal trouble. Full purges by instance moderation action is different.

The scope of my proposal is not to help curate and filtering content for instances, but to enable widespread coordinated action against content which was already created and is harmful without any room for interpretation.

Makes sense. I suppose my described lists could also be made into purge lists.

That is a whole other issue. Scrapers and request spammers, not content spamming accounts, can be combated with already existing technology at the reverse proxy server. There are big crowdsourced IP banlists which are continuously updated.

Not my knowledge base, so good to know that already exists.

This would be the easiest solution. But I think it will lead away from individualism of instances, as all decisions are automatically applied, instead of only a manually reviewed subset.

Your idea with applying the mod action after a threshold of other instances did it too is something to consider. It will keep individualism but also creates “avalanches” once enough instances in the network agreed on a decision, which might be even better for propagating swift action against general harmful content.

Definitely a balance needs to be struck, between quick action and preventing abuse.

I think Pylova (a soft-fork of Piefed) is at the forefront of moderation tools in the threadiverse.

I think their recent addition of shareable user-subscribable block-lists is a good direction to go. I think this can be expanded to instance-level subscribable block-lists. There could be a variety of lists (CSAM, scrappers, scammers, agentics, etc.) and each instance chooses what list(s) to subscribe to or build their own for others to subscribe and contribute to. The lists would be easily fork-able so that any instance can take control of their own moderation, if they disagreed with an addition/subtraction to a list.

For some lists (like scrappers), I think automatic addition could be set up to quickly “inoculate” the rest of the threadiverse against them. Perhaps one approach could be any IP that clicks on a non-user clickable link or gets stuck(?) in Anubis could be added.

I think subscribable block-lists for instance admins could be a great way to protect the threadiverse (perhaps expanded to the fediverse as a whole or just any website in general) against any problematic actors quickly and at the first encounter (alternatively: after (x) # or % of instance admins add them to a list).

Apparently some of these lists may already exist in some form (@Christopher@lemmy.grey.fail: “I was able to implement community maintained blocklists.”). So, my thought would be to make sure these lists are a standard format (if they aren’t already) that instances can auto retrieve and contribute to.

Anyone: let me know if you have any thoughts or see any issues with the idea (constructive criticism only, please).

A lot of the problems with Lemmy boil down ultimately to overall lack of moderation.

Yeah, I would say we are definitely at a “work-in-progress” level, with a ways to go.

The recent thread in the Fediverse community praising the devs of many Fediverse platforms is a great example: it became highly toxic, ultimately being locked.

I saw the post originally. Going back it looks like FiniteBanjo went to pick a fight. I’m not trying make any judgement on their comments, just that they appear, to me, to wanted to start an argument (“Well, not all of them. Some of them are a stain on this earth.”). Not surprised it was locked at that point.

… with the admins themselves literally flinging actual death threats against admins of other instances.

My recollection and understanding of that situation was that it was an inferred threat, that, when confronted, was immediately rescinded and that admin stepped down due to their actions. So, I think, it’s a bit disingenuous to use that as a negative against them. To me, they responded appropriately when confronted with how their message was taken (clarification that’s not how it was intended and then stepping down from their position of authority).

I believe soon after the .world admins faked “evidence” that the anarchist admins were really fascists trying to sow discord or something (I don’t recall the exact allegation, but I remember the evidence being shown and confirmed to be fake). My recollection is Rimu backed the .world admins even after their evidence had been shown to be faked. I want to say that is when Rimu introduced the instance “silencing” feature, to reduce anarchist influence after his and the .world admins’ attempts to expel the anarchists with the faked evidence failed.

So there will never be a solution to this step that involves all of us getting onto the same page.

Decentralization. There doesn’t need to be, in my eyes.

I do decry the right of places such as Chapotraphouse to exist, but I do think that they need to be labeled properly, with a warning label …

… the instance picker (the last time I checked) was very confusing and came nowhere close to explaining what individual users could expect when signing up to an instance.

… as people who jumped through all of the prior hurdles flee this toxic shithole …

On-boarding is a huge gap in my mind. From my view, we basically throw everyone in the deep-end and hope they learn to swim and aren’t too concerned with fixing the issue, because we all learned to swim, so why can’t they.

I could see an on-boarding that quickly explains the structure (like 2 sentences about decentralization/email-analogy and identifying a users name and instance. Then go into a quick questionnaire (for determining best host instance for them) about what kind of posts they want to see (one of those could be a quoted .ml account, to see if the new user wants to see “triad” content). From there some, basic filtering/block list yes/nos (block-list for spam accounts: yes/no, agentic accounts: yes/no, bot accounts: yes/no, etc.) and interested topics (software, cats, art, etc.). I think that would go a long way in helping get an easier and more welcoming environment (may only be fully possible via Pylova features at the moment (block-lists)).

Unfortunately, PieFed seems to be continually actively working against this goal now, making the options more opaque rather than transparent - e.g. any PieFed community on an instance that does not explicitly disable the downvote feature (and which ones have done that? how is anyone supposed to know!?) is shadow-censoring contributions offered from many different software platforms (Lemmy, Mbin, nodeBB, Friendica, Mastodon, etc.), without disclosure of that fact.

That’s a good point I hadn’t thought of before. I’m not sure how an instance could keep a user informed without keeping a larger database of community and instance settings. Alternatively, sending that info for each load or interaction the user makes. The larger database sounds much more reasonable, but still daunting and not easily implemented and probably with a significant compute cost.

… whose devs look to me to be tired on working to support new features, especially for PieFed.

This isn’t surprising to me. Piefed has 20% of the users Lemmy has (and that increase to 20% is relatively recent) and developer generally want to get paid for their work (especially if they don’t have a main corporate job).

PyLova might be the single best hope the entire Threadiverse has to grow. We’ll see what happens!

Yeah, definitely looks to be the brightest hope we have right now. Hopefully, they can keep up a sustainable pace and don’t burn out.

I do need to find better language though, so that I don’t annoy people when I say that it is sometimes hard to have hope that this place will grow.

You may want to try using “feeling language”. I’ve found it helps a lot online and IRL.

Ex: “I feel like that …”, “My understanding of the situation is …”, “To me that …”, etc.

Bc really, growth isn’t the point?

Though it is necessary, even to remain afloat.

Agreed, for all the reasons you mentioned earlier, plus I think it would be nicer if the activity, administrative, and development load was spread out a bit wider (makes it easier on everyone/less burnout, and provides redundancies).

That said, the community that you pointed to might be in a weird lockdown state, since I noticed that it has zero active mods. See the sidebar on the home instance and then the modlog offers the further knowledge that the only prior mod has been banned from that instance. If you want to revive it, you could appeal to the instance owner and become the new kid of it?

I was going to take your suggestion, because I doubt it is much mod responsibility, but it looks like programming.dev was having federation issues when I tried to post and the post never made it. Remade it just now and it went through, no issues.

The people that have thrived here so far are those who approach it from an IT hobbiest standpoint and are willing to roll up their sleeves and get to work building it. But most people are not willing or capable to do likewise, which turns many away. The tools are still being built, and definitely lack polish atm.

Agreed. I feel it’s currently very much a make your own space and if people come by great, but don’t expect or rely on it happening.

But we are building them nonetheless (well, Mia is:-P, though we can help with suggestions!), steadily over time. So I am less concerned about current number of MAUs than in seeing the challenges be overcome, such as building these tools. Although that requires funding which in turn requires some level of MAUs, thus I am ecstatic about these subscribable blocklists to help keep people from fleeing this place!! 🤩

Agreed :)

Oh, do you mean the private voting issue?

If it helps, I had previously responded to some of the more controversial points that users had raised a year ago.

Honestly, you could probably lay out a dozen controversial decisions Rimu has made and I wouldn’t be able to pick which one I’m vaguely recalling.

I admit to being very wrong here - I thought, and even said publicly - that surely Rimu would reverse course and walk back that decision, as he had so very many others previously.

I thought so too.

He acts like Huffman, treating people as landed gentry on his platform. Which is… the truth!! Hence people have left it, and even more are choosing to simply go to it. Nonetheless, I truly do wish the best to the project… I just no longer have any hope that it will continue, most certainly not like it had in the past.

It’s his, but at the same time, being open source, it could very quickly not be (ex. Pylova). As long as there is a solid conversion path, Rimu could one day wake up and find .social is the only Piefed instance.

GET CREATIVE?!?

Rimu not only took away voting, and did so in a way that breaks the ActivityPub model - leaving users to scratch their heads wondering what error had occurred where the arrow turned green but the count did not change - but in response to why people were not allowed to vote anymore, he said to simply move on and do other things besides voting.

Yeah, his communication skills need significant work, which wouldn’t be so bad, if he was open-minded, but it really seems, as you’re saying, it has to be his way.

… the hidden, third option is simply to not engage with communities located on piefed instances anymore. And bam, just like that, communities such as !nonpolitical_comics@piefed.social had their contributions drop like a rock (I skewed those stats a bit b/c I seemed to forget for a moment and submitted 6 posts before I remembered, as I tried out different 3rd-party apps to see what those experiences were like).

Absolutely. This is what I love about open source and federation. It may take time, but the ‘system’ routes around ‘damage’. Looks to me like Pylova may be the route around the Piefed/Rimu damage.

REALLY! That’s amazing! I would guess a migration over from Reddit. I hope they feel welcomed.

I guess from Reddit too. It’s been pretty consistently 100+ new users a day. Given instance “stickiness”, I’d say ~60% stick around.

I’ve seen a couple posts from newbies, the threads look really positive, there still is the allow/block triad conversation/fighting, but it was very minor from what I saw.

About downvotes, I misrecalled: Mia said that it was added to the to-do list to be added for the next release, although that was 3 months ago, and yet I have not seen it added yet.

If you haven’t seen her response, it’s appears to be a user toggle-able setting, which may turn into an instance setting.

There are a lot of “unfriendly” aspects such as this to the Fediverse. I see some people fleeing now because their instance went belly-up - off-hand I can think of mander.xyz and slrpnk.net and… at least one more that I cannot now recall the name of (their admin team simply listed a single account beginning with “The “, which at least at one point had not posted for years), all of which are having issues. Of course that is simply par for the course for the Fediverse model!

Yeah, it’s unfortunate, but I think it could also be an opportunity to figure out problems/frictions and solutions. I think for the worry instances go down, a few things would inspire confidence, for me. Off-site backup redundancy (data), a set up non-profit with financial transparency (finances), and a “succession plan” (admins: if admin people want to leave they can but there is a plan in place to hand things off and replace them).

But less necessary are those bot spam posts made ostensibly to make the place look more welcoming and active and alive - at first glance, but then those accounts (almost?) never respond to anyone’s comment, which sours people and leaves a negative impression of this place as being a ghost town.

Agreed. I think clearly labeling that this a bot post, could be good to feature. I still like seeing articles, even if I don’t want to comment on them. Maybe having them separated out as RSS feeds would be better though. I think something is happening with regards to RSS and Piefed, but I didn’t look into the specifics.

But… it is simply not worth the effort, when Reddit and FaceBook and Insta and X and Bluesky have REAL (high amounts of) content, so investing all that effort to learn and block what they do not want will yield significantly greater dividends… they seem to think, and by the mantra that “the customer is always right”, they are correct.

One thing I kept hearing from the new users this week, is how calm it is here. Like there is much less antagonism, insults, and rage bait and much more thoughtful discussion and nice interactions. So, I think, the threadiverse will hold some of the new converts until they learn how to curate their experience better.

Thereby if we want to attrack them more, we need to offer such blocking by default, making unregistered bot accounts either outright illegal or at least opt-in rather than allowing them to be opt-out.

Now I would put my hopes into the shareable blocklists to do such.

Yeah, I think “informed default” (like .zip), where the blocks are setup by default, but new users are informed about them and there is an easy way to disable them, if the new user chooses to, is a good path forward.

About your idea: I disagree. An admin that is hosting their own instance is different than someone who is simply using an existing instance. Such a person should have access to a “connectivity health” monitoring tool yes, but I don’t see why it would need to be on a per-post basis.

I mean as opposed to having every single post do this across the entire Threadiverse, at all times.

Anyway, I feel like an admin tool would be a better fit for that situation than changing the way that all posts federated all of the time - if I understood you correctly there.

It seems I may not have conveyed the idea well. So to try again, my understanding is that an instance sends to the appropriate “hosting instance” a post or comment. My thought is, if it doesn’t already happen, have the “hosting instance” send an ‘acknowledgement’ back to the sending instance. Then, if the comment has been ‘acknowledged’, the sending instance would put an indicator on the comment (check mark or something) to show “yes, my user, your activity successfully arrived at it’s destination.”

In my mind, this wouldn’t be every instance checking every other instance’s user’s activity to make sure it federated properly. This would be an instance confirming things for only its own users.

For instance if I make a post to !Fediverse@lemmy.world, even as a mere user (not an instance admin) I can go over to that instance at https://lemmy.world/c/fediverse/ and look for my post there (similarly for comments). I have routinely done that many times in the past, especially from StarTrek.Website that had such horrible network connectivity at the time that I joined it, prompting me to seek out Discuss.Online that has such fantastic uptime due to spectacularly capable admins (including Jgrim the author of Sublinks).

Previously though, I would likewise post something and then getting zero up- (or down-) votes, go seek out why, and discover that my post had not federated. Worse, when it eventually did get federated, it did not appear to others as “new”, and instead looked like it had been submitted several days in the past, at which point people were not alerted to it, and it became buried amidst all the flood of prior posts that had already been engaged with, even though those had not been. It was quite sad, how bad the connectivity of the Threadiverse was 3-4 years ago:-(. These days it is a LOT better!

I’m think the feature I’m suggesting would do away with this “federation uncertainty friction”. I rarely go to check other instances to see if my comment or post has federated or not. I’m usually on mobile and this is not an easy maneuver, so I often have a question if my post/comment federated or if I’m only talking to myself. Not a great feeling and definitely reduces the urge to contribute, for me.

I just checked the post I made to the !servo@programming.dev community yesterday, did not federate (https://programming.dev/c/Servo is missing my post). Glad I spent time on that post… :(

I think the “acknowledgement” feature could also help elevate the issue entirely. I’m not sure how to ‘force’ a re-submission of my post, but if an instance didn’t get an ‘acknowledgement’ the instance could check that the '“hosting instance” is still receiving and then try resend the post/comment after a certain time interval.

Probably would have missed it. Appreciate the tag :)

How votes are viewed is a user setting, letting anyone choose if they want combined, separate, or a percentage. Is this something you would also like settable as a global instance default choice?

Honestly, as long as I can easily enable it, it doesn’t matter to me if everyone has it by default (only viewed your instance without logging in). Might be worth a user poll, but I have no idea how wanted that feature is.

FYI I altered my [FR] on Codeburg to reflect you’d already implementing the feature.

We mostly use Grafna

Link doesn’t work, but I think I know what it’ll look like.

Given both your responses, it seems I didn’t conveyed the idea well to OpenStars. My understanding is that an instance sends to the appropriate “hosting instance” a post or comment. My thought is, if it doesn’t already happen, have the “hosting instance” send an ‘acknowledgement’ back to the sending instance. Then, if the comment has been ‘acknowledged’, the sending instance would put an indicator on the comment (check mark or something) to show “yes, my user, your activity successfully arrived at it’s destination.”

In my mind, this wouldn’t be every instance checking every other instance’s user’s activity to make sure it federated properly. This would be an instance confirming things for only its own users.

Hopefully that makes sense. I graphical up-time graph is probably nice to have too. I was going to say those are available on fediverse measurement sites, but with a quick look, it seems like that was a Fediverse-Observer (RIP) only feature.

Also I occasionally search for PyLova and I love reading this comment chain when I do, thanks for contributing so many ideas.

Happy to help and thanks for being so responsive, open-minded, and just all around great :)

Edit: when making my comment to OpenStars, I checked a post I made to the !servo@programming.dev community yesterday, which did not federate (https://programming.dev/c/Servo is missing my post).

I think the “acknowledgement” feature could also help elevate the issue entirely. I’m not sure how to ‘force’ a re-submission of my post, but if an instance didn’t get an ‘acknowledgement’ the instance could check that the ‘"hosting instance” is still receiving and then try resend the post/comment after a certain time interval.

I am a bit behind in catching up - thanks for sharing all that, both the topics with the PyLova devs and making sure I was kept in the loop as well:-).

Yeah, no worries :)

I don’t recall the prior issue, unless it was the one with user blocking being silent. On that one I was swayed by the argument that it simply had to be that way due to limitations within Lemmy.

I don’t think that was it but maybe?

… all of which Lemmy will ignore, as is its right and choice to do so.

Agreed.

But the vote quota crosses a line where iirc it doesn’t even tell other PieFed instances - nor Mbin nor nodeBB nor Mastodon etc. - what is going on. It’s horribly designed, or at least it was upon release and if it has changed since then I have not heard of such. Some features take literally years to be implemented, but this one is slammed down everyone’s throats against all objections to the contrary without anyone seeming to have been heard?

Absolutely. I was ready to leave Piefed over it.

Sure, whatever, Rimu can do as he pleases, but it makes it really hard to hold out any hope for the future of this project when people keep losing investment in it (now Snoopy having become the latest casualty, seemingly? at least it appears that way).

Yeah, Snoopy leaving is interesting, in that I’m not sure how much new blood Rimu is going to attract in the future (he’s burned a lot of bridges). The Codeburg commits lists Rimu, Walter Schwenger (wsj018), and wltb as the only developers over the last 3 months. Looks like there were quite a few more other people pushing commits 4 months back. Looks like the Pylova split happened 5 months ago and a couple of the light commiters followed Mia to Pylova.

The Innuendo Studios YouTube series The Alt-Right Playbook does a superb job at explaining how someone can fall into authoritarian thinking.

Sounds interesting, I’ll have to check it out.

Particularly interesting are those in the outer circle of the onion-like layers that are shocked, Shocked I tell you, SHOCKED 😲 😳, when they find out that the people who had been screeching all along at the top of their lungs who they are, suddenly start doing the very things that they always said that they were going to do if they only had the power to make it happen.

It would be kind of funny, if the effects weren’t terrible :(

PieFed is still growing despite that though, just not by nearly so much or robustly …

Yeah, looks like there is some significant growth in the Threadiverse this week (~+750 MAU). Roughly 600 for Lemmy, 170 for Piefed, and -10 for Mbin. Edit: looks like I was running the analysis on the same processed data instead of processing each new day’s data before the analysis (only ran two of the scripts to check the daily numbers instead of all 3 :/). So, actually +250 MAU, rather than +750 MAU.

I’ve brought up the topic of separating downvotes from upvotes with Rimu years ago, it seems very unlikely to happen. Mainly because he does not want it himself, therefore does not want to expend much effort to put it into the code (understandable), and also it is legitimately a difficult UI issue that he prefers to leave to others.

I vaguely recall this, so might have seen your post/comment at the time.

PyLova has already done it though …

They did? I just checked their instances and it has the same consolidated total votes UI. It would be great if they already did it, maybe it’s an option and not the default?

On a different note, I ran across a post who deleted there account, because almost no one was voting on their posts. I think it was a federation issue, as in the community hosting instance never received the posts and they only stayed local.

I was thinking maybe a check-mark or something to let a user know their post or comment was successfully received by the target/host instance. I don’t know if hosting instances tell sending instances they successfully retrieved data or not, so could be a significant increase in traffic.

Maybe not that significant though. Piefed only had ~1700 posts and comments over the last week, Lemmy top 20 most populated instances were ~12500 (“Activity” Data). Taking the most “active” instance (lemmy.world), they had ~4700 activities, assume worst case they are all for non-Lemmy.world communities. That would be 670 extra packets per day (28 per hour). That doesn’t seem too terrible. Thoughts?

My understanding is that Piefed does group the posts together, but that no apps currently implement that feature.

From Munkisquisher@lemmy.nz, it sounds like the Connect app does stack the posts, but that may be through their own feature and not Piefed’s (no idea). I just looked for a site or post that compared app features and couldn’t find one that listed grouping cross-posts.

My recollection is that Lemmy 1.0 may have cross-post grouping, but I don’t see it listed as a feature on 1.0 beta testing post.

I’ve asked for #1 already (but not in Codeberg issue format) so this is good confirmation for them that such is still desired.

Looks like they like the idea and it’ll be in v5.1. Looks like they already completed it.

… also instance admins are not going to rerun it each time they update to the latest software, so their “surprise” came about due to the lack of #1, and bc Rimu tried to sneak it in buried amidst a rather large list of other features as if it were a tiny unimportant matter (even though he knew that it was a huge deal and extremely controversial …

Fair point.

… essentially Kickstarted PyLova to be created …

Silver linings :)

… “hard-coded” features existed buried in the PieFed codebase …

Yeah, I remember that ‘argument’. That was annoying to deal with.

I can well imagine others being opt-out or hard-coded at first and then upon push-back being converted into opt-in.

There was another feature/set of features that were opt-out and poorly communicated before the vote-limiter, maybe a year ago. I don’t remember what it was though. I recall it went over just as poorly as the rate limiter and made me view Rimu in a different light.

Speaking of, the last I knew the shadow-censoring daily vote quotas are not opt-in, but rather turned on by default, being instead opt-out.

That’s my understanding as well. I had to ask my admins to remove it.

Only if looking at their actions rather than words 😜.

Haha. Yeah. I do think their call for a vanguard kind of gives away the game (wiki: “According to Lenin, the purpose of the vanguard party is to establish a dictatorship of the proletariat”). I think using the word “dictatorship” at all is unlikely to go well or attract the people you want that would be okay with giving up power, once they have it.

Okay but by FAR the biggest change there in PieFed subscriber numbers was due solely to fedinsfw.app, which in turn resulted from the demise of lemmynsfw.

And it begs the question for me: how many MAUs did Lemmynsfw have, compared to fedinsfw.app? Like, yes PieFed gained ~7k MAUs, but how many more did Lemmy lose? I doubt they were equal, which again illustrates the overall pattern of decline across the Threadiverse instances.

Absolutely. According to Fedindex, it looks like Lemmynsfw may have only had 3.6k MAU when it crashed. FediNSFW is now at ~3.6k MAU. So, weirdly it came out to be an even exchange.

PieFed stats are still fairly similar to what they were before, right?

Looks like Fedindex is putting Fedinsfw creation on 02-18-2026. Fediverse-Observer data, puts Piefed at 4816 MAU on 02-17-2026. Piefed is at 6742 MAU and Fedinsfw only at 3583 MAU, leaves non-Fedinsfw Piefed at 3159 MAU today.

Last time Piefed was below 2k MAU (1990 MAU) was 2026-01-20. So, up ~59% in 10-ish months, not bad, percentage wise.

Personally I think that a more nuanced take might be to look at the increasing number of instances total - even if they are only tiny ones, that’s still something? Otoh, pound for pound, Lemmy already has both those AND the huge buckets of users, so perhaps it’s not worth digging into such, bc in essence such matters are already baked into the comparison between platforms anyway.

Looks like Lemmy has had 63 new instances since 2026-06-08, 20 still active (Piefed: 18 new, 8 still active). So better percentage fro Piefed, better volume for Lemmy.

From 2026-07-28 to 2026-10-27, Lemmy alive servers went from 321 -> 294 (Piefed: 66 -> 59). So, with the increased total threadiverse MAUs, it looks like some consolidation is happening.

And even if we do bring in the porn instance into the comparison, as the largest PieFed instance, it is still far smaller than the largest Lemmy instance, by quite a large margin. Also, people may literally have multiple accounts - one for porn and another for other matters, for separation.

I only hope that Rimu considers why Mbin became so irrelevant, especially after the huge start given it by the demise of Kbin: Mbin does multiple things yes, but it does neither of them well.

With you there. ‘Whole ass 1 thing, rather than half-assing 2 things’.

PieFed has numerous layout concerns too, such as making it exceedingly difficult to distinguish between downvotes vs. upvotes in the overall count (+151 & -100 is enormously different than just +51 & -0, or +71 & -20 - that is outright information that PieFed obscures and in many cases makes nearly impossible to get readily).

This has legitimately been a huge annoyance for me on web (and a reason I prefer Voyager). I made FR about this too.

I’m quite curious how they are going to implement it.

Very exciting indeed!! :-)

Looks like it’s already live. Looks like it’s only contained to users. I was hoping this could also be an admin feature to help stop scraping (with automation only 1 instance has to be hit by a scraper before the subscribed instances are ‘inoculated’). Added another FR.

Edit: added an additional sentence to the ml discussion.

Looks like the Threadiverse is ~45k. So, relatively stable and up year-over-year. Peak would have been 2023-07-17 at 76585 MAU (Lemmy + Mbin/Kbin; during the Reddit 3rd party exodus?).

Actually that probably will not begin until v1.0 …

Good to know.

If they even know that setting exists, and are willing to tweak the code to make that happen.

The PyLova devs were actively seeking suggestions so perhaps it is indeed worth pushing some additional thoughts their way - it’s a win-win there!:-)

On that front, I opened another Pylova Issue for instance settings (user facing and one for admin installs). Hopefully that will bear fruit.

The vast majority of Rimu’s features that have been complained about are offered as opt-in …

My recollection is that they are now opt-in. They didn’t start that way and didn’t become that way until backlash. I think he’s learned his lesson at this point, to make everything opt-in.

… this centralized authority for community names - which is not chosen to be a third-party resource - is opt-out.

Yeah, I mean Marx-Leninism (ml) is an authoritarian ideology. Seems par for the course.

… as a whole they have been exceedingly slow to add more than merely rudimentary support for PieFed specific features. The low number of people that actually use PieFed likely plays a role there, as too surely the decreasing subscriber counts in the Threadiverse as a whole world as well.

Yeah, that’s my read as well. Before a spike the Threadiverse was ~40k and Piefed was ~2k (5%). Now Piefed is 7k (17.5%) of that same ~40k, not a huge user-base/market and Piefed being the smaller part of that.

Mbin mentions a lot more contributors at https://github.com/mbinOrg/mbin#contributors, although I didn’t look to see how active they all are especially recently …

My “2 part time developers” (looks like it’s actually 1.5 developers, 1 full, 1 half-time) is from a recent post.

… even if every single one jumped over to PieFed it would be not all that significant a change compared to its already 6-7K.

My feeling is it’s less about MAU numbers (best case: +10%) and more about becoming ‘the only game in town’. I think good integration of Mastodon, could significantly increase Piefed MAU over Lemmy and would make Mbin redundant. Making a good Lemmy to Piefed conversion script after that, could knock Lemmy into a death-spiral as well. Piefed would then be the only fediverse Reddit-alternative.

I also suspect that Rimu’s vision of Mastodon integration will go much deeper, like Mbin suffers from splitting its purpose to be 2-in-1, whereas PieFed would become simply one interface that trafficked in both types of content. And perhaps more besides - Friendica, Pixelfed, Loops, etc. That would be grand …

Agreed. I think I was sold on the original ‘dream’ of the fediverse all being connected and it was a bit disappointing when I learned that wasn’t really the case.

… who say that they are very turned away from its default web-browser accessible UI. Which if the goal was to incorporate Mastodon posts makes more sense to deprioritize for now anyway, to avoid having to redesign it again later.

It wouldn’t surprise me, if that were the case.

Speaking of, I hope the subscribable blocklists will allow integration of instances, communities, and individual accounts - possibly also even domains (aka webpage URLs, like various news sources). That way someone could subscribe to just one and get it all. It’s not the worst thing if that cannot be done, but if it’s easy then…? Yet another thing to ask them, because they would know better what’s easy or not:-).

Yeah, I’m quite curious how they are going to implement it. I suspect, if it needs to be iterated on, it won’t be that onerous to change how it can be utilized.

I think what you are referring to is not content so much as that new communities needed - in the past - to have been added by someone before others could find them in the community listing. This was not restricted to admins though & people like Blaze had an army of alts and did much of that, as did I on my own instances as well. This was necessary before the first post could come in from each new community, needing to be performed on a per-community basis.

I’m aware of that aspect, and I may be conflating it with on-boarding new instances.

Although it gives Lemmy.ml specifically veto power in that if a community is banned or the instance it is on is defederated from Lemmy.ml, then the automatic pulling in of the community name is no longer done.

Also, the admins can designate a different instance than Lemmy.ml, though Lemmy.ml is set as the default, so this procedure is opt-out.

The Lemmy software now utilizes a centralized authority, TIL.

Rimu seems less interested in the Threadiverse these days and more in interoperability with Mastodon, despite how most of us seem to not even want that, and Mbin already exists.

I don’t want to prescribe motives to his actions, but I feel like some of his positions and efforts are to “dominate” the threadiverse and knock out the competition.

In this case, Mbin. From my understanding, Mbin is pretty far behind in development (lots of bugs and missing features), only have 2 part time developers and they are trying to supplement that with AI/LLMs. I suspect the LLMs will help them for a bit, but the project will die when the code is only manageable by LLMs and prices for vibe-coding go up significantly.

When that happens, all the Mbin users could “easily” migrate to Piefed. Though, I imagine, if Piefed gets Mastodon implementation working properly, that may cause a migration then too (possibly death spiraling Mbin, if the migration is large enough).

Those PyLova devs are hella impressive!!!

… it makes it all the more impressive to see how responsive the PyLova is to the needs of the people on the Threadiverse.

Yeah, they are :)

I don’t think that most instances do delete old content. Lemmy.world self-reportedly has content from decades in the past from nodeBB but I think their database got corrupted and I don’t trust that. However, here is an example on my own instance https://discuss.online/post/5497584 showing the retaining of a post to a community on Lemmy.ml from 8 years ago.

Yeah, now that you mention it, I’m not sure how wide spread deletion is. I believe I recall one of Rimu’s blog posts mentioned it in some depth, so I assumed at least some instances are taking advantage of it (especially with the price of hardware).

Unfortunately there’s almost no way to know short of directly asking the admin team, who may not even be aware themselves what settings are in place on their own machines.

… yet I haven’t seen even PyLova expose whatever these settings are to the public. Even though someone might make a choice as to which instance they want to make an account on (or even start up their own) if they could find the answer.

That’s a good point. Maybe worth another Pylova feature request?

I think even subscribable blocklists would only be the first part of an overall answer. Like UBlock Origin, it’s going to require people to manufacture and maintain those lists.

Hmm. I’m imagining someone will just make a block list as their only blocking measure and anytime they need to block someone, they’ll just add them to the list, and people will subscribe to that list or multiple lists.

… are quickly defederated from by the rest of the Threadiverse, but such requires effort on behalf of the instance admin. I thought, though perhaps I am wrong.

I have no source for this, but I think, now, first content/post/comment from a new instance has to be approved by an instance admin (could be wrong and misremembering though). If not, might be worth seeing if that is a option admins would want added. Thoughts?

… or to begin blocking people, one by one …

All of which is a lot of effort, especially since it must continue to happen with each fresh new alt made to subvert bans. Each new joiner must then be subjected this again and again, like a hazing ritual that we who have been through such already now offer little to no help to new joiners to avoid. Except for your codeberg request that will help to change that - hooray!

On that note, it seems Pylova has finished writing up the block subscription code (yesterday) and it’ll be in the next update (unknown date).

… if ML were to ever go down, that content would continue to remain, foreve, just like happened with Lemm.ee) …

It could remain forever, but I think most instances (or at least Piefed instance) are set up to delete old content after “x” amount of time. Alternatively, admins could just manually delete the old content, if they wanted.

… A place’s reputation should proceed it, not be entirely ignored and force each new joiner to discover it all over again on their own, as what happens with ML.

Agreed. I think .zip’s approach where it has the default blocks and explains them is a good approach. If the subscribable blocks also make it into the on-boarding, I think there will be a large swath of new .zip users who almost never encounter the triad issues (assuming Pylova subscribable block list spreads back to Piefed and beyond). If that on boarding spreads throughout the Threadiverse, the triad could be dealt with, without major confrontations, which I think most admins don’t want to deal with.

Note that if ML ever were to be defederated, the absolute most toxic trolls would simply switch to using their alts on some other instance such as Lemmy.World, which they have announced they will do in advance, several years ago. They even indicated that they would make apps that automate the process of trying to send messages to someone, cycling through an array of alts until one is found that is not banned on that instance (or by that recipient?)

My understanding is the trolls exist because .ml admins allow it. If they try using .world, I expect the .world admins would ban. If they ban evade, I imagine a new safeguard would be added to counter them.

On the whole, I think our reputation is closer to nostr and 4chan than it is to Bluesky.

I don’t go to other social media, so my only knowledge of our reputation is from people repeating things here.

Hopefully “our reputation” can jump to a different comparison scale (I wouldn’t want to be compared to any of those options).

Institutionally at least, anything goes - including the worst CSAM.

Okay, I’m going to push back on this. At no point have I ever heard CSAM being accepted or allowed. The only time I’ve ever heard of it coming up, everyone reacted negatively. The poster was blocked, all the posts were deleted immediately, and new safeguards were put into prevent it happening again.

That team is so energetic and helpful, it’s a pleasure to see it! 😁

Agreed :)

Edit: missed a section.

My understanding is federated “payment” systems are difficult. I think there is active discussion on the topic, but I’m not sure where code/development is at, or if it’s even started.

I think flohmarkt (“federated decentral classified ad software”) is the closest we have right now. It has low (~10-30) Monthly Active Users (MAU), but is in active development.

This sounds horrific to me! The question in my mind is not how many ghosts to allow over here, but how to generate authentic interactions between real humans. We don’t need every single niche covered, we just need people to enjoy their time spent here.

I do not want to have to scroll down reading through tens to hundreds to thousands of comments, none of which represent real humans who would respond if I replied to them - that is the precise opposite of what I want.

I do not think that the overall lack of content is the problem here - it’s the overall lack of content that is worthwhile to read, …

I didn’t think I conveyed the idea well. It would not be an automated or bot process (Reddit comments wouldn’t be added as Threadiverse comments). The good/useful comments from Reddit would be utilized by a poster to add to their own threadiverse post to help add more relevant content (Example that got 57 comments).

A further 90% improvement can be attained by blocking Lemmy.ml …

While that may be the fastest solution, the triad haven’t been doing great in at least the last few months, even without major intervention (Post with Data):
Lemmy.ml:

• Difference from 3 Months ago: -103 MAU (-5.5%)
• Difference from 1 week ago: -23 MAU (-1.3%)
• Posts and Comments (per week): 727 (+106 vs historic)

hexbear.net:

• Difference from 3 Months ago: -78 MAU (-5.7%)
• Difference from 1 week ago: -18 MAU (-1.4%)
• Posts and Comments (per week): 1659 (-3545 vs historic)

Lemmygrad.ml:

• Difference from 3 Months ago: 22 MAU (+3.9%)
• Difference from 1 week ago: 0 MAU (0%)
• Posts and Comments (per week): 311 (-1 vs historic)

Average of the other 16 Instances (of Top 20, excludes lemmy.world who’s similarly crashing):

• Difference from 3 Months ago: -16 MAU (-1.4%)
• Difference from 1 week ago: +3 MAU (+0.3%)
• Posts and Comments (per week): 387 (+22 vs historic)

It looks like hexbear.net is suffocating by itself. Maybe some users move from hexbear.net to lemmy.ml or the declining hexbear.net activity is forcing lemmy.ml users to be more active (I have no evidence of either, but for argument’s sake say they do), even then lemmy.ml is shrinking too.

In this very conversation thread there was a particular onerous troll who kept following our conversation and downvoting everything I said …

I did wonder what happened to that person.

New joiners, it seems obvious from the comments they leave (usually on other platforms after already leaving), don’t bother to put in the work to curate their experiences and instead simply flee …

Seems like some efforts toward better on-boarding could return good results (could include default blocks, etc.). I have some other analytics I need to get working again first, but after, I think I’ll take a stab at looking at decoupling and identifying key instance on-boarding differences that effect retention.

Edit: moved comma.

Edit 2: Looks like Pylova likes the subscribable block list idea: https://codeberg.org/quokkau/PyLova/issues/30

That is sad to hear that mander hasn’t come back yet - I guess it’s still waiting to see if something can be resolved, since the backups weren’t recent.

But it is even more sad to hear that Rimu’s community migration tool isn’t working, on other instances without his direct involvement.

Agreed.

Some of what you said makes sense to me though, like a community “owner” must be located on the instance where it is housed - if that were not the case then the system would be vulnerable to all sorts of spam where any bot could spin up communities on any instance, so some level of accountability must be maintained.

Hmm. I see what your saying. I was thinking more someone on the instance would be required to make the community, but could hand it off to a user on another instance. I think you are right though, a user on the instance should be forced to be “owner”, as it doesn’t take much work to create a new account and they can always add their other instance username as a mod and run the community via that (never really needing to log into username of the instance the community is on again). Only issue I see is, people DMing the “owner” username that is not monitored, but I suppose that could just be stated in the sidebar and/or user profile.

I suspect the official PieFed app will be Blorp, just from the way that Rimu talked about it iirc.

I can’t recall any specific source tied to that thought, but that sounds right. Guess we’ll see.

… it’s far easier to pay Reddit and then get Reddit coins back in exchange.

Haha, sadly very true.

… an enormous fraction of our own traffic here are reposting bots, often from Reddit, …

I just saw a post that Reddit changed something to block reposting bots again and Lemmit is tired of trying to figure out a workaround and is going to shut down next year when the domain subscription ends. So, potentially fewer reposts in the future.

As a bit of an aside, JohnnyEnzyme @piefed.social just told me about a posting technique that worked well for him utilizing past Reddit posts with their best comments to create interaction in the threadiverse (his comment). So, we may lose repost bots (no great loss in my mind), but if people start taking more from Reddit and curating it personally, I think that could be a great exchange.

Edit: As an update I added a feature request to Pylova’s Codeburg about the subscribable block lists. We’ll see if they’re interested.

Topic Feeds are already user-customizable and shareable, so why not block lists?

I never made that connection, but yeah, this could be much easier to implement than I was thinking.

Admiral Patrick’s toxicity filter went behind people’s backs and did such opaquely, but when handled transparently and with full informed consent that would be an entirely different story. And then people could choose like no filter, minimal or maximal filtering, or perhaps some intermediary steps between those.

Yeah, I could see this being a few easy setting/feature addition.

It’s been like 3 years. I for one have lost hope …

Understandable.

Some issues remain, like instance blocking does so without informing the recipient, who gets essentially shadow-blocked by the recipient. But that is not something fixable by PieFed alone, and would require cooperation with Lemmy to receive the signal and pass it on to the affected users in whatever way they chose.

My understanding is that this was set up that way to prevent harassment. I’m not knowledge on the subject so could be wrong.

No actually there is - Rimu converted many Lemm.ee communities over to PieFed.social (I forget if they lost access to modify their older content, or perhaps if they messaged him before lemm.ee went fully down from their old account then he could take that as proof of ownership and grant such access), and Quokka has converted fully from Lemmy to PieFed and then from PieFed to PyLova, so aside from Rimu are the preeminent experts in such conversions. With Rimu’s community migration tools and Quokka’s experience, communities on Lemmy are not held hostage, anymore.

My understanding was that it “worked”. As in it was haphazard and a very manual effort to get what did happen to work.

Not the same thing, but die4ever@retrolemmy.com and I tried to save the past posts of ecospheres@mander.xyz to piefed.zip after mander.xyz went down. It didn’t work, I (on piefed.zip) could see the old mander.xyz posts, but no one else on any other instances could. I also had a problem with adding mods to the new piefed.zip community (lemmy mods couldn’t be added, non-piefed.zip mods couldn’t be made owner).

So, while I expect it can be done, I don’t think it’s an easy, straightforward, or automated process to move software or even communities. I know my example was not a 1-to-1 typical move, so if you have any counter-examples please let me know (I can utilize them for the !fedigrow@lemmy.zip wiki).

Otoh, Rimu himself is personally quite opinionated on certain matters, and rarely but decidedly goes against community push-back to enact his own agendas.

I am quite aware of Rimu’s nature (he and I had discussions in the past).

Rimu lost a lot of trust in his handling of that situation.

Agreed and I would say he’s lost a lot of trust many times at this point (I used to be one of his biggest cheerleaders).

Of course there is the PyLova fork, but it is extremely new, so I would wager that people are taking stock of it and seeing how it will shake out.

I suspect so. I mentioned it to the piefed.zip admins, so they are at least aware of it. But I suspect there won’t be any piefed -> pylova conversions anytime soon, unless Rimu screws up big (again).

… all the more so since most 3rd-party apps only nominally support PieFed much less PyLova.

I believe their is an official Piefed app coming (https://fedidb.com/software/compare?software=lemmy%2Cmbin%2Cpiefed). Assuming it’s open source, presumably the other partially Piefed compatible apps could utilize the “Piefed Official App” 's code to improve compatibility within their own apps. I’m certainly going to suggest it to the voyager Dev when the “official app” comes out.

The technology has to get better if we want to bring people here. Which I suppose it is, but at a pace that is exceedingly frustrating.

Agreed, but not sure we should expect more. FOSS is generally slow moving, due to it being small, part-time teams. I’m in a place where I can help support FOSS financially, but I know that is not an option for everyone, plus it’s a new consideration/viewpoint for people coming from corporate/proprietary software (where they are the product).