balsoft

Yes, that can happen sometimes, but I find that there are plenty of cheap options with unlocked bootloaders if you look for them.

Nah, cheap phones often have their bootloader unlocked/unlockable. Really happy with my POCO M5 running modified AOSP. Also, unlike every expensive phone nowadays, it has 3.5mm jack, SD card slot, and exceptional battery life for hiking/trekking (it survives 5-6 days as just a camera+map phone with all power saving on, in comparison people with flagships typically only last 2-3 days with the same usage and power-saving techniques).

You can almost always replace the battery, even when the manufacturer doesn't want you to. As for flashlights, they typically come with easily user-replaceable ones, often even sold separately. Worst-case, you can get a AA or AAA flashlight and use rechargeable AA/AAAs.

I've paid quite a lot for my second headlamp for hiking, but I am really happy with the purchase as it's very light (35 g) compared to my first cheapo one (~120 g), while being the same 200 lm max. It doesn't sound like much, but it's enough for me to not even notice it, while the heavy one was getting annoying after a while.

I'd say it depends. For safety-critical stuff maybe, but for a headlamp or something I prefer rechargeable as I can easily recharge it from a power bank or a portable solar panel if needed. If you run out of a disposable battery for whatever reason, you're screwed.

I'm not sure you should "cheap out" on headphones per se. The really cheap ones are usually horrible, both in terms of sound quality, usability and comfort (well, except for wired Apple ones, allegedly, though they never fit me right). It's just that it makes no sense to go for really expensive ones, unless you're really into audio and love hearing the tiny sound reproduction differences between them, or enjoying the different tech etc. The middle ground of $50-$100 for in-ears and $100-300 for over-ears will often offer you good/great/excellent sound quality and the same usability&comfort as more expensive ones.

Problem is not how weak or strong the encryption is

Here it's definitely part of discussion. The context was

It’s encrypted anonymous communication capabilities.

It's barely anonymous, and poorly encrypted. The latter is the reason Durov is in custody while Signal devs are scott free. He could easily turn illegal stuff over to French authorities, but doesn't.

The bigger problem is that people somehow assume this a huge threat, while all previous cases didn’t involve anything like that.

There have absolutely been cases where a backdoor/weakness/lack of encryption used to catch criminals before: https://en.wikipedia.org/wiki/Operation_Trojan_Shield https://en.wikipedia.org/wiki/Ennetcom https://en.wikipedia.org/wiki/EncroChat . I distinctly remember that there were also arrests of opposition activists in Russia based on personal messages in VKontakte, but can't find the news right now.

real criminals do their stuff everywhere (especially on telegram) for years, staying safe.

Some are staying safe, others are being caught precisely because of this.

Problem is not how weak or strong the encryption is, but that once you are under oppression and do opposition activities, you’re going to learn by yourself how to deal with it.

Using better encryption schemes is definitely part of that.

Toy may call it TLS but it’s a custom protocol.

Sure, it's mtproto. The security it provides for non-encrypted chats (which are the absolute majority of chats) is not any different from just having TLS for transport. It's potentially even worse as it's not as well-audited.

Data is not kept unencrypted on their servers, according to their docs.

That just means that they store both your data in some encrypted way and the key. They can still read it trivially. You don't even have to know the protocol to understand why: you can add new devices without having any other device online, and read all non-secret chats. It might also just mean disk encryption, in which case it's plain-text in RAM while the server is running.

FYI a great way to share your logs (if you have internet on the machine but no GUI) is to use a pastebin like 0x0.st. E.g. do journalctl | curl -F'file=@-' https://0x0.st and post the link here.

I'd try the kernel version used in 23.11 to see if that fixes it. Add boot.kernelPackages = pkgs.linuxKernel.packages.linux_6_1; to your config (make sure to add it before the closing }) and sudo nixos-rebuild switch . You can also try older versions, like linux_5_19. If it doesn't fix the problem it might be that the ACPI error is a red herring and the problem is something else entirely, in which case it'd be more difficult to diagnose, and I'd recommend just staying on 23.11 for now and only taking the new packages that you need from 24.05. There's a great post on how to do this here: https://discourse.nixos.org/t/installing-multiple-packages-from-unstable-channel-in-configuration-nix/19271/2 (probably also in the docs somewhere but I couldn't find it easily).

telegram is not safe enough, but for some reason it is preferred by drug dealers. The lives of these people literally depend on the messenger.

Most people are technically illiterate. On top of that, most criminals are idiots (otherwise they'd have calculated risk/reward ratio rather than only looking at the reward side of things). The reason it is used by drug dealers is (1) Telegram is convenient, (2) Telegram is not moderated so they accumulate there due to "moderation selection", (3) Law enforcement didn't care enough to do anything about that. Now that (3) has changed and (2) is on shaky grounds, I expect a lot of them will move elsewhere.

Signal chats can read messenger developers with basic hack (add one more person in chat and collect messages on disk)

How exactly do you think that would work? To add a new recipient the client needs to explicitly encrypt messages with a key available to that recipient. What command in the Signal protocol would trigger that action without first establishing trust in the recipient? (FYI when adding a new device, there is a key-exchange and verification process, which requires access to some other device with keys already on it).