I made a blog post discussing my biggest issues with Lemmy and why I am kind of done with it as a software.
👏👏👏👏👏
Well said. I don’t disagree with a single point you made, and some of it echos concerns I’ve had since day 1. And extra points for calling out
.ml
as lemmygrad-lite. I think I’ve called it exactly that as well.The only thing I really have to add is on the topic of toxicity. Like you, I’m an instance admin and have a bird’s eye view of a lot of behavior patterns. I’ve recently started wondering how many people are here because they’re too toxic for regular social media rather than because they want to be here. I won’t guess an actual number, but I would say it’s not insignificant.
I’m firmly the latter case: I want to be here, I want this to succeed, and I’m trying to put in the work toward that result. And I’ve interacted with lots and lots of people in the same boat. But, like you, I’m also growing disillusioned for many of the same reasons.
On the bright side, I’ve gotten much less rusty as a developer after having to write scrips and tools to fill in the massive gaps in moderation features.
I’ve recently started wondering how many people are here because they’re too toxic for regular social media rather than because they want to be here.
This has largely been my operating assumption as well since day one when I came over during the Reddit API lockdown. I was fairly active on a NSFW alt up until recently and I’ve actually seen dozens of comments from new users mentioning that the only reason they were here on Lemmy was because they were banned from Reddit and had no other viable options. They were always an asshole to the posters and the reality is that with a lower population of users is that there aren’t enough other voices to drown out these people yet and you end up with a feedback loop of toxicity.
I’ve recently started wondering how many people are here because they’re too toxic for regular social media rather than because they want to be here.
Dude yes, I’ve been thinking the same thing. I worry that users curious to leave reddit are going to go to a big instance, see concentrated worst-parts-of-reddit, and decide it’s not for them.
In theory, decentralization enables freedom from the average user being forced to put up with toxicity. But we don’t really have that (yet) until the ratio of jerk to non-jerk improves.
it demands that I log in, to view the post??
I migrated over to writefreely, here’s the new link. https://my.jewy.blog/my-love-hate-relationship-with-lemmy
Cert expired yesterday (6/4/24) and needs to be renewed.
deleted by creator
There are a lot of good points here, I appreciate the time you put into it.
As an end user of both Lemmy and Mastodon, it’s always an eye opener to see how developers greet user requests and suggestions with curt or snarky replies. Even “Why don’t you open an issue on our source tracker” will often effectively shut down suggestions from less tech savvy newcomers.
My own concerns are more on my own level, though. It resonates with me when you write —
The Fediverse has its own existing cultures that thrive here. And when you enter a space that already exists you need to be mindful of that to prevent issues from occurring.
I’ve seen a few user migration waves, and I think your description of (some) Lemmy users who just want a drop-in Reddit replacement is on point. Mastodon has had its share of Twitterati who surged in trying to recreate their previous circles and tone. Obviously, it’s a generalisation but we do need to face the problem.
The transition from a walled garden environment like Reddit or Twitter — moderated by professionals or enthusiasts, and algorithmically curated — to a federated space with carefully cultivated etiquettes will never be like simply picking up a conversation in another UI.
I’d be interested how a project like Sublinks would/could accommodate the existing fediverse cultures, and hopefully bridge the cognitive gap that seems to exist between threadiverse and fediverse?
Even “Why don’t you open an issue on our source tracker” will often effectively shut down suggestions from less tech savvy newcomers.
How should developers handle feature requests? Keep in mind there is a need for the whole team to see the suggestion and it’s also good to have a place to gather feedback and further discuss.
Community managers - sometimes just talking about your issue with someone will help tremendously in figuring out how to put it and they often can just do it for you. That said, Lemmy devs do not value work being put in the issue tracker - they have admitted to not reading it. People who cannot contribute code are just entirely ignored and have no power in the project’s direction.
I suspect the small size of the dev team and the general nature of an OSS project means there aren’t swarms of people around volunteering to be community managers.
Small projects your sway with the project is directly proportional to your ability to submit pull requests. It’s just a sad fact that it’s easier to say “I wish we had feature X” vs. “Here is a pull request that implements feature X”.
At least with OSS you are getting what you paid for (nothing!), vs commercial companies where you pay for the software and they STILL ignore you.
I mean, I essentially proposed to do this myself in private conversations with Dessalines but there was no willingness for a shared roadmap so it felt pretty pointless.
Many users on Lemmy seem actively hostile to the idea of decentralization in a way that feels self defeating. They don’t want a better alternative to Reddit, they just want Reddit 2.0 and attempts to sway them towards something better feels like pulling teeth.
I keep seeing this, and I don’t really understand. Lemmy is a link aggregator that allows users to organize those links into categories/communities/etc, and lets people comment on the links and have discussions about them. From an end-user perspective, that’s exactly what Reddit is. So I’m genuinely curious what’s meant when people say they don’t want Reddit 2.0 from a technical perspective. From a social perspective, the toxicity, brigading, shitposting, etc are definitely not desirable. But with shit moderation tools, those sort of things don’t get sorted, and federation just magnifies all of those problems. Though I think disabling voting definitely helps discourage shitposting and low-effort responses.
But I genuinely do think a lot of problems really come down to the fundamentals of federation. And given how many downsides there are to it, I’m not convinced it’s actually a benefit at all.
Do you mean disabling downvotes? That’s how it is on Blahaj. It definitely makes a difference to the amount of toxicity I think.
On Blahaj reports are the only way to express disapproval of content. So you could for example spread fascist dogwhistles about not liking politics, and if Ada doesn’t understand the dogwhistle then your content doesn’t get removed. That gives cryptofascists free reign
I’ve seen quotes directly lifted from fascist works upvoted by hundreds on Beehaw. The problem with only-positive user feedback is that as long as it seems like a positive statement that others support people will often grant it further support without thinking about what is actually being said.
Or at least that’s what I hope was happening.
If you can point me to any comments like these, I’d love to remove them fwiw.
Would that actually violate the guidelines? This was around all the time of the defederation drama, so ages ago in internet time, and I recall looking at the guidelines and thinking “Well this isn’t a bad faith argument, and it’s not technically hateful unless you know where it leads.”
Bad faith or not - Fascists are not accepted.
Domain blocks are always publicly visible.
Mod logs are always publicly visible in the public mod log.
What? It is crucial for the users, not a bug.
It can become a source of targeted harassment, as it has on the rest of the fediverse before.
Censorship and targeted sliencing of users are the source of bad moderation. To top it off, the mod can target and harrass the banned user and we wouldn’t know bc of censorship if allowed.
targeted harassment
-
It is about the anonymity of moderator, not about modlog
-
Quit this job as moderator if you can’t take it
if the social prescription to harassment of moderators is “quit because you’re a baby” then you’re going to have many fewer pleasant spaces on the Fediverse in which to exist—because yeah, a lot of people will just quit. i am agnostic on the public modlog overall, but this is an obvious concern with it that i’m not convinced can just be dismissed idly. i obviously have better things to do than a thankless, payless job in which harassment would be dismissed like that.
It especially leads to harassment of vulnerable people. There’s many aspects of moderation that is done here that if implemented in other fediverse software would become a vector for Kiwifarms level harassment.
-
This is the first time I heard of Sublinks, and honestly after a quick look through here in Lemmy I get the impression that main devs of Lemmy and Sublinks can work together to improve what it is currently the best option (Lemmy).
I honestly think it is way too early to have a Lemmy “replacement” even if it is all running in the Fediverse, I just think it is a split of efforts, granted, I don’t know all the background that runs behind and it seems like Sublinks dev does it like a hobby too.
Regarding moving communities to Sublinks, yeah, it is up to instance maintainers, but that is a no for me, heck, I already had to recreate my stuff from the dead FMHY account I had (there was no account migration at the time), it seems like adding more decentralization to me, and we already had that with multiple repeated communities ¯\_(ツ)_/¯
Finally, don’t get me wrong, options and alternatives are always welcome in my book, but as I said before, it feels like way too early for me.
The thing is, sublinks lead developer tried to work with development team of lemmy. It was like pulling teeth constantly and his experience was overall negative.
This is completely false. jgrim has never opened up a single issue or PR in lemmy, and we certainly wouldn’t refuse any of either. These things are easily verifiable by anyone who wants to look, its all out in the open.
Does anyone know of any Sublinks instances? The main page for it speaks in present tense, but I haven’t found any active instances. (aside from the demo, of course)
I apologize for my stupidity:
this is me
It hasnt been released yet, still working towards parity (but getting there soon)
The first instance using it will likely be sublinks.art and some other instances will be switching over from lemmy when it hits parity like programming.dev and literature.cafe