https://github.com/LemmyNet/lemmy/issues/2977

According to the lemmy devs, deleting your account also overwrites and purges your comments and posts. This deletion is being federated - but other servers may choose to ignore it, be buggy, down or whatever, and therefore not fully deleting your content there.

Since lemmy is decentralized, that’s as close to a full deletion as you’ll get, and to quote your rationale, that’ll be the only bridge you’ll be able to burn.

Side note: Why are several of you so aggressive on a simple question? If they want to leave the platform and delete their content, why shouldn’t they, especially since that is absolutely non-trivial on a decentralized platform?

Neat, thank you for the hint - only recently got a ccwgtv, and I’m still not used to the fact that you’re able to install apps like this 😂

Same for me on Chromecast, it’s pretty much unwatchable overnight, started probably about a week ago. I’m to lazy to set up adblocking for my full network, and I was fine watching the skippable 15s ads, and the occasional longer ones, but currently I mostly get at least a full minute of ads on my TV.

Before you talked about the Fediverse as a whole, now from a single user perspective.

IMO it affects the Fediverse as a whole by abusing it. The whole idea is an open network, where instances can federate with each other to bilaterally share information and create a seemingly single platform. This is not the case with the planned Threads integration, because they explicitly plan to feed on the content, but hiding sharing their own content behind an (for most of their userbase) obscure opt-in.

From a single user perspective it doesn’t affect you directly. But it affects the platform you are part of with malicious intent.

I am not against Threads joining the Fediverse, and I do actually think it would be great for the growth of the Fediverse if actual big players join, and if it brings content that I personally do not like to see, I can use the tools available (e.g. blocking user/communities/instances) to hide it. But only if they plan on joining as a “regular instance” like any other - but Meta does not intent doing so, since they have chosen the opt-in with obvious intent of simply gaining additional content on their walled platform for their own gain.

The problem is not them reading data, but that Threads will take Fediverse content, and display it on Threads. In the opposite direction, Fediverse will only see the select few user content that do actually opt-in, and let’s be honest here, most users won’t know what the Fediverse is, except for again the few people that are on both platforms. This is absolutely not “playing nice” as you’ve put it before, and purely parasitic and, again, purely a greed decision by Meta. I don’t really know why you are shilling so hard trying to excuse absolutely unexcusable behavior.

https://www.threads.net/@mosseri/post/C051TLnud8U

There’s your playing nice. They want to feed on ActivityPub data, while only contributing in a per-user opt-in selection. It’s a joke, and both their motivation and what they are doing is absolutely fucked. It’s another cog in their data ingestion machine that they can keep fucking around with, again.

I really can not comprehend why anyone would give this advance of trust to Meta, when all signs are showing you to bail.

Meta has React, RocksDB and pytorch, and a few other “niche” frameworks and tools. “Half of the internet […] run[ning] on open source code and infrastructure that Meta built and maintains” is a big, big exaggeration. Also maintainance is done by the OSS community for big parts, and I’m really curious what open source infrastructure Meta is running.

I’m not saying Meta has no relevance in OSS, but I can hardly think of an open source org that does open source purely for its own benefit. React helps them shape the web in the way Meta wants it, their ML stuff is important for their own internal needs (ads, BI, and the whole social networking, etc.), their AR/VR/XR contributions are for the Quest, and KI/LLM since they need it themselves instead of relying/partnering with OpenAI. Meta (the company) absolutely does not stand by the principles of open source, no matter how much you want to sugarcoat it.

It’s not a hate train, it’s being cautious. And do you really think that Meta is open sourcing because of their passion for FOSS and standing by those values? They’ve taken an internal framework they’ve build, open source it so that they can advertise how open and great they are on the page you linked, and after it gains traction (which it will, since it’s used by Meta it must be good /s) they can reduce their own internal efforts to a minimum, since the community will contribute. Open source may be a passion for the developers of Meta, but the company Meta does not give a single flying fuck about FOSS or the Fediverse.

You’re right, containers are not VMs, and I’ve never claimed that. For the matter of basic unix access control for a beginner they are similar enough to treat them as such. It’s enough of a baseline for basic security for a beginners workload imo. For advanced use cases - absolutely do not treat containers as you would VMs.

Imagine your containers as very lightweight mini-VMs. Would you run everything as root in your virtual machines? Containers aren’t really that different to classical VMs from an operations point of view. You have a different attack surface, but it is still there, and running as a non-root user inside the container reduces this attack surface, and should IMHO be the default. Privileged containers and users may be required for specific purposes, but should not be the norm, if possible.