Maybe this is spam designed to make you (and others) leave matrix?

In XMPP public rooms we definitely had a very motivated hater spamming gore images (and worse) everywhere they could. The point was precisely to make everyone run away.

The way this crap is moderated on big platforms may be more effective, but at what cost? ("facebook moderation ptsd" on your favorite search engine, but I bet you know the story already)

I've just been asked by @matrix to delete this post. They had the audacity to blame *me* for giving publicity to the problems they have.

If you can't run a service safely, you shouldn't run it at all.

Seeing posts like this, I simply had disabled invites on my server before it could spread to me, so I've not seen anything.

I'll keep the server for now, while migrating everything to ejabberd. It's got WIP Matrix support, so if I really have to, I'll be able to join a Matrix room, but everything else will be XMPP. There IDs can be hidden in a chat, making the spread of spam much more difficult, just because IDs cannot be scraped as easily.

Really don't wanna see THAT.

I think we need to do something about the gap.

We're stronger together as one large community with an open protocol instead of two divided once.

Together we could take over the world!1!!1!! :p

Oh and the privacy features!

XMPP can hide IDs in group chats, making it really hard to scrape IDs and spam.

I wish that simple feature would come to Matrix.

And setting a nick when joining and completely hiding the global nick.

Being able to hide the profile picture in public chats and only share it with contacts.

I don't want to show my face to the whole internet, nor do I want everyone on there to be able to contact me easily.

If those improvements are made -> fanboy.

Please let me know whether this reply is:

(1) helpful or
(2) not helpful

Currently on a train and cannot really focus, so it's a bit stream of conciousness.

Also, I agree "it sucks" is unhelpful.

But also feel like feedback is often just ignored, unfortunately. Hope the community interactions will improve, love the ejabberd community as a positive example.

My issues are mostly not even the spam (I just disabled invites), but mostly sloooow performance. Otherwise I'm fine with it all.

Sometimes Fractal neeeds 30 minutes to sync up when opening, it's written in Rust and should be fast enough.

I suspect Element would be slower.

From your linked post:

[Bluesky]"Like Mastodon, this social platform is also decentralized and open source. In terms of use and features, it’s similar to X and free to use."

It's not really decentralised, even if it markets itself like that. In fact, it is more Twitter pre-Musk than Mastodon.

Not only the linked article does not provide any means to "opt out", clearly stating that opt-out is not an option, but also no way to "object" for users in EU could be found.

5 out of 10, would not recommend.

…let’s rephrase this. If you are using Meta’s products (WhatsApp included) your data *is* going into the AI woodchipper. Full stop. Don’t think otherwise. Opting out/in is just appeasing politicians. At least 3 reasons for this; it is extraordinarily difficult to prove your data is in there or remove it once digested, the US isn’t going to prosecute them, and they can pay (or delay/etc) any EU fines/prosecution.

“Opt out” is exactly as effective as “Do not track”. That’s it.

Yes please!

Slop PRs haven't been a big issue yet on my repo. What I've had has come from automated third party accounts that I could easily block. But now the call is coming from inside the house!

I support that idea, but also github under microsoft is too centralized anyway.

This would be a good moment to consider alternatives and moving. If the platform is supposed to help productivity and it just doesn't do that...

I'm sure you're aware of that, I'm just advertising the idea, because network effects.

Would requiring a specific mandatory wording to be included for personally submitted issues and PR's be a temporary work-around?

It should then be possible to filter out all submissions without this wording included to be dealt with as seems appropriate.

The fact people use AI to generate Issues and PR/MR's is for me grounds to block that dev from all my projects.

if you cannot find the time to sit down and have 1 original thought then you should not be writing software.

according to the docs, the agent must first be enabled: docs.github.com/en/enterprise-…

So I think it's just a matter of not enabling it. Also, someone must explicitly request it to do something and it goes through a PR approval process as usual, following branch rules etc.

That said, I don't think I'd enable it myself. Copilot already makes a big mess with relatively simple tasks. Automating mess doesn't feel right to me.

Enabling Copilot coding agent - GitHub Enterprise Cloud Docs

You can allow Copilot to work on code and raise pull requests.

^{GitHub Docs}

Sensitive content Reveal/hide