I like kbin but I'm hoping to find a fast way to filter out all of the German subs. I have absolutely no issue with them, I just can't understand the content so it's useless to me. It seems to take up about half of my feed
I like kbin but I'm hoping to find a fast way to filter out all of the German subs. I have absolutely no issue with them, I just can't understand the content so it's useless to me. It seems to take up about half of my feed
Because feddit.de is a main Lemmy instance and there are some popular instances over there. I'd love to block feddit.de out right but they have the main starwarsmemes and some other good practical magazines, so I'm stuck unsubbing from them on a one-by-one basis.
If you're on kbin.social you can select subscriptions as the default in your user settings. That way only the magazines you subscribe to show in your feed.
You guys just love the Pole(-ish) position?
You can click the cancel icon besides the subscribe button in any magazine to stop them being shown to you
I don't see either a subscribe button or a cancel icon. When i click on a community name it brings me to a page that's just a feed of the links in that community.
I play this game where I try to understand the ich_iel memes. I usually get close. Some of them are hilarious.
Plus it goes the other way. If someone only knows German, the English will be a much-worse flood.
Probably needs a PR to add a per-user language filtering setting. Probably a good idea to permit multiple options for multilingual users. I don't believe that it exists today, though clearly the metadata required to do the filtering is available.
I’m actually hoping that at some point we may have an automatic translation for content not in our chosen language. The way Mastodon does. I want to understand what the threads and comments in other languages are saying! I feel like it broadens understanding.
I just block them
I 99% agree with you, but I remember Reddit 10 years ago. For me, this feels like what I would have imagined Reddit felt like a couple years before I joined, 13 years ago.
Would that make them Lemmigrants?
I get that this is about Active Users but should we/the fediverse be concerned about the sheer number of bot accounts being created on brand new, questionable instances?
To tell if we should be concerned about bots is a difficult task. To make a decision, here's some things we should consider:
This is a fantastic response.
Fist those who comply.
typo, translation issue, or...[shivers]
Regardless; I, for one, welcome our robot overlords.
Some bots can even be helpful, like if they were to repost content that we like to know about. But yeah, it can also put things on unequal footing like person A doesn't like person B so makes 50 bot accounts that downvote them everywhere they go and no matter what they say. Then again, a group of 50 people could also accomplish that without bots, or like 5 people each with 10 alt accounts (but like, otherwise normal & active). Also, someone could spin up their own personal instance and join the federation and infiltrate the entire network that way (I mean in a way that even an instance admin could do nothing about, b/c they are the admin for it). As the federation grows, I expect to see full-on brigading, and infiltrations, and yes even Russian trolls.
Then again, there's an important difference: Reddit has 2000 employees and can barely hold that website together - most of them have got to be like public relations, advertising, accountants, HR, interns and the like - whereas the Lemmy/kbin codebase is open source, so we can expect to see contributions from people who care and are knowledgeable, much as Reddit had mods who did the same, tirelessly devoting hours of their weeks (often per day even) to improving the place. Now that Rexit happened / is happening, expect to see the pace of improvements to increase. :-)
There are tons of ways - e.g. is there an account that has existed for a whole day and the only thing it has ever done is downvote posts? That's a bot. Perhaps similarly for up-voters, although lurkers exist so perhaps only question those with a captcha rather than straight-up remove. If such measures remove the easiest-to-detect bot spammers, then it would be too exhausting for one person to have like 1000 bot accounts - they'd have to do nothing but make posts all day long just to keep them alive! - thus it would limit their influence.
Well I'm still trying to wrap my head around the idea of federation but initially I imagine an instance could just not federate with an instance that doesn't use captcha/email verification during signup but my confusion is wondering if like there is a transitive trust in the fediverse. Like Lemmy.1 trusts Lemmy.2 and Lemmy.1 does not trust lemmy.3 but Lemmy.2 DOES trust lemmy.3... can bots from Lemmy.3 post on Lemmy.2 which will then be federated to Lemmy.1? Sorry, that seems super confusing as I type it out but it's the best way I can describe it.
If your question in this scenario is can a defederated Lemmy.3's posts reach Lemmy.1 if they're posted to a neutral intermediary (Lemmy.2), then no, instances that are defederated from each other don't trade information. Doesn't matter how far the post/comment spreads if your instance is ignoring it based on point of origin.
If both users subbed to a mutually federating instance's community, they would both be able to interact with Lemmy.2 users, but they wouldn't even be able to see each other.
As for bots......I fear there logically is no good way to tackle this because of the nature of the fediverse itself, short of blacklisting every baby instance and manually approving users. At the very minimum, asking for verified email and never auto-suggesting usernames, which is far from airtight but could slow them down.
Ideally, one could write their own bot to evaluate users for brigading behavior, copied posts and comments, etc., but that seems a little sci-fi when some instances barely have a working search bar right now
If your question in this scenario is can a defederated Lemmy.3's posts reach Lemmy.1 if they're posted to a neutral intermediary (Lemmy.2)
You worded it so much better than I did. That's what I was wondering, thanks for clearing it up for me.
How many are bots though? Instances should be required to have captcha when making accounts to curb bots.
Also seems like Dansup is finding a lot of spam accounts on lemmy instances too. He manages the fedidb of all the fediverse software and platforms with various metrics
thanks for using Leebra!
go to feed...
I'm pretty content with KBin. As time goes on the content level will increase and hopefully remain at a level which makes it easy to curate my feed and reduce noise. Truth be told Reddit has been getting worse for a long time and being here reflects that. This feels a lot like what Reddit felt like 10 years ago.
save