The domain https://fediverse.observer/ redirects to sedo now.
On October 11th these two commits where made to the repo:
this will be my final release, glad to have contributed since 2011 to some fediverse stats, https://fedidb.com/ is a great resource with more interest
this software is no longer developed
Sad to have lost such a great resource.


True dat.
OOOH, Stamets, right. To explain why I missed it: Stamets not only is an absolute ass, he also strikes me as very untrustworthy. I don’t recall the previous fight between Rimu vs. the entire rest of the world before that LW vs. AN matter (I think it had something to do with bots? later I am adding: YES, I think it might have been that AI-assisted modding thing!), but he seemed to me to be willfully disregarding all of the facts, even as he seemed like he was advocating for the common man but doing a very poor job of it. I 100% do not mind if someone says like “I can’t quite put my finger on it, but I dislike this person (Rimu)” - I can handle vagueness and imprecision - but to be presented with facts and figures and screenshots and evidence and not respond to any of that but double-down all the harder on the points that have already been disproven, strikes me as very poor form. Like even a stopped watch is right twice a day, and even Donald Trump can say a good thing, we don’t need to be tribal to oppose all the “other side” people, just the other side ideas that should be able to stand or fall entirely on their own merit (only resorting to tribal thinking as a tie-breaker or to save time). Also, Stamets was obviously going through something IRL at the time - I believe he said as much - as he left TenForward, the very community that he himself had created. Eventually he created the second Risa, but that was not the reason that he left.
That said, while I would very much distrust any conclusion that he would arrive at (especially around that time-period while he went through whatever IRL), I also would highly doubt that he would willfully make up false facts like screenshots. So I am not calling him a liar, just stating my previous bias to explain why I missed whatever evidence he presented. I had already mentally checked out of his messages by that point. How weird would it be if he was right, and just did not manage to convey that in a manner that I was receptive too, because of that prior matter (AI modding).
Thanks for the links - I have not yet but I will read them. Archived data that has since been changed is fairly damning. This is one of the MAJOR reasons that people defederated from Hexbear, not only b/c it exists primarily to troll other users, but b/c the admins were caught lying to other admins. Whoever did the lying should be removed as an admin - do you know why they have not been? LW tends to be slow, but over time they do seem to care about this stuff, so I wonder why not? Or is Russian disinformation coming into play - some archival sites themselves have been shown to alter past data! I am not that deep into the admin drama.
And really I am not even so much “pushing” that we need to grow the Threadiverse - just stating that if we wanted to, they would likely come from Reddit, and be turned off by all of the enormous complexity, just like most of us here are also turned off by it, even though we who have been here longer know how to navigate around it.
I don’t like that idea - first off it discards data, which may have answers that someone would like to read later on (e.g. how to fix something, or a funny meme picture, or like an analysis of a current event as it is happening, which can be different than the event viewed with hindsight, etc.). Also, it still leaves the “problem” active for multiple years before “solving” it. (and I could go on, like a single post in a 2-year period could prevent that community from going away, plus I recall an occurrence where 196 moved from LW to Blahaj but while >90% of the people moved, a handful did decide to stay - so in that case the community would not be entirely purged, and to state that it is the most popular community with that name would be highly misleading, in that while that once was true, after the migration it is no longer true, so misleading and outdated data).
That said, the issue where on Lemmy communities are sorted by lifetime number of members is absolutely already solved by PieFed/PyLova (I almost want to reverse the ordering of those two, but I suppose PieFed was historically first:-), since it uses MAUs. However many tens of thousands of users a community might once have had 3 years ago does not matter in the slightest if the last post was from over a year ago - MAUs are by far the best way to sort communities by current activity levels, even (especially?) if that would prioritize LW comms over ML ones.
Oh yes, they absolutely hate Reddit over there, on r/RedditAlternatives. That said, they despise Lemmy even more. Our reputation is really, shockingly poor, outside of this space.
You can read more about Reddit demographics, but a few key highlights are:
Also, in https://www.reddit.com/r/dataisbeautiful/comments/1ib7py7/oc_reddit_user_demographics_vs_other_social_media/ you can see that more than any other social media platform (Facebook, Instagram, Tiktok, X/Twitter), Reddit leans left. Though as we agree, that means right.
And yes, Reddit is more atheist than it is religious, I believe you are correct there. But also, it is more religious than here, by far. Xhitter even more so, and especially Instagram & Facebook even more still. Just by virtue of being American, many Redditors are going to have a more religious background (as too I would think are most Hispanic cultures?). Remember the bias that the squeaky wheel gets noticed more - atheists are very ah… uh… “outspoken”. This does not (in itself) equate to any kind of either moral or numerical superiority. Anyway, if the Threadiverse wants to grow our subscribers, those people would have to come from somewhere.
Like here is a good back-and-forth conversation about the topic: https://www.reddit.com/r/RedditAlternatives/comments/1xk7z2v/comment/19rxs9tkgla4h/.
So I am not saying that the Flotilla does not have good reasons to defed from LW - damn, admins lying to admins sounds like a great one to me. “the reality is that a fragmented model cannot work” is the most common refrain I read on there, whenever I look. It’s too complicated, it’s too hard to find content (solved by PieFed, but that too is a complication, like already it’s not “Lemmy” anymore but rather part of the “and the sort”, like what is this PieFed vs. PyLova vs. Mbin concept and why should I care?, I could well imagine them asking), and basically every single topic is highly politically charged (again - no value judgement there, events irl are thus charged!!:-D though Americans are used to sleeping through such, and it’s a shocking awakening conversion process) and trolls run rampant. YOU and I have blocked them, but to a new account, they are not blocked by the vast majority of instance admins. Ironically LW does this a lot, whereas db0 does much less. To each according to their need I suppose…
Here’s a comment that says the above outright
Another highly relevant comment:
Of course this is the cache-22 - we cannot get new members until we first get new members. Which again, have to come from somewhere. Except we keep chasing them away rather than retaining them - as evidenced by the fact that we used to have 65K members, or at least we maintained a 40-45K steady state for a long time, until people simply gave up.
And there’s just so much more like that. Here’s an interesting comment:
And… I am out of space to continue to write anymore text, having hit the limi
No idea. I just know his summary matched my recollection of events.
I had one of those experience with Rimu, about data analysis. I laid out why he couldn’t make the claims he was making and he doubled-down and stuck his head in the sand. I try not to interact with his comments or posts now (no need to start drama where none exists).
No idea. I just linked his Star Trek updates, because I found them on while searching for the doctored evidence.
To be clear, Stamets is not the one making the claim the evidence is doctored (he just had a summary of interactions before that point in time), Loco_Mex@sh.itjust.works was posting the evidence.
I think that’s actually pretty common. In my experience, most people only take information from “trusted” sources. If you already distrust someone, it likely doesn’t matter what they say or how they say it.
Did not know that. Not overly surprising though.
No idea. Looking through all those past posts and comments, I recall seeing that the LW admins wouldn’t make decisions on removing power-abusive mods and tried to get others to do it. So, I wouldn’t hold out hope they ever did anything.
My recollection of that archival site was that it was using visiting users to DDOS-ing select websites. I don’t know of any falsifying of data/pages/pictures/etc.
Gotcha.
Fair points and it does seem like Piefed/Pylova have solved it with active users. I was using some Lemmy instances to search for the doctored evidence, so that was my frame of reference for the “search” functionality.
I know the feeling. I try to keep all my posts and comments neutral so nothing comes off biased, but I do try to add Pylova into any of my posts or comments about Threadiverse, so it has more exposure.
Haha. Not overly surprising, or counter to my understanding of the situation.
Thanks for the data source, some good stuff in there.
Interestingly, it looks like most of the data links eventually source to Statista, I can’t see their numbers without paying, but their graphs appear to visually show Reddit users decreasing in 2026, over 2025. It would be great to get the raw data to look at it in more detail, but oh well.
I think someone tried to make a christian instance. I blocked it quickly, but it may still exist.
Yep, I think this is back to ‘our on-boarding process is a failure and needs major improvements’ realm.
From what I’ve seen, most growth is in spikes (large increases and then significant decreases). So, relatively speaking, only dropping 30-40% from an all time max is actually really good (Mastodon is currently -73.0%). I think Loops is a really good example of what bad looks like (~29k -> ~4k, -86.3%). My generalization, from what I’ve seen, is 0 -> -25% is growing, -25 -> -80% is ‘stable’, and > -80% is ‘dying’. Not to say something ‘dying’ can’t come back to live, just that they are not in a good place and they haven’t been for a while.