It is 240 W actually. Seems Techpowerup was wrong. See my first comment: https://discuss.tchncs.de/post/59511400/25528592

The source of law here is Directive 2022/2380 (which amends Directive 2014/53), in Article 2 a grace period until 2026-04-28 is defined for the category of laptops. This has now expired, which explains the renewed wave of articles being published.

The directive itself is not that interesting to read, as a lot of it is just empowering the Commission to make a decision on the specifics. The result is in the Commission Delegated Regulation 2023/1717. Although it seems to me like something is missing. I can’t find more though.

A very interesting Q&A from their Commission Notice – Guidance document:

16. Are laptops and other radio equipment that require more than 240 W of charging power exempted from the ‘common charger’ rules?

No. They are not exempted. Radio equipment which is subject to the ‘common charger’ rules must incorporate the harmonised charging solution.

The Commission has updated (in Commission Delegated Regulation (EU) 2023/1717), the references to the standards cited in Annex Ia to the latest version of the European standards. Therefore, due to the amendments introduced by this delegated regulation, radio equipment subject to the ‘common charger’ rules must incorporate the harmonised charging solution up to their maximum charging power or up to 240W if their maximum charging power is above 240W (as opposed to 100W in the previous versions of the standards concerned).

The Commission will continue to update the technical specifications set out in Annex Ia, in order to reflect scientific and technological progress or market developments provided that such developments meet the objectives of the common charging solution.

But then also

20. Are proprietary charging receptacles allowed in addition to a USB-C receptacle?

Yes. The RED only requires radio equipment subject to the ‘common charger’ rules to be equipped with the USB-C receptacle. The use of other receptacles is therefore not prohibited as long as the covered radio equipment is also equipped with a harmonised charging (USB-C) receptacle.

That means those hefty laptops going up to 350 W or whatever, now need to accept 240 W over USB PD, but they may still include additional proprietary charging solutions that are rated higher.

Also I don’t think the 100 W limit that some outlets report is actually in force since 2023/1717 has replaced the references to ‘EN IEC 62680-1-3:2021’ by those to ‘EN IEC 62680-1-3:2022’

Reading on, yes they make that explicit further down:

27. Is a radio equipment allowed to charge above 240 W when using an additional charging protocol?

Yes. If the radio equipment proprietary charging solution requires more than 240 W (e.g. 300 W), the concerned radio equipment must also support USB PD up to 240W.

The Commission has updated, via Commission Delegated Regulation (EU) 2023/1717, the references to the standards cited in Annex Ia to the latest version of the European standards. The updated version of the standards will apply as of the date of applicability of the relevant rules introduced to the RED by the Common Charger Directive, i.e. for handheld mobile phones, tablets, digital cameras, headphones, headsets, handheld videogame consoles, portable speakers, e-readers, keyboards, mice, portable navigation systems and earbuds, as of 28 December 2024 and, for laptops, as of 28 April 2026. This means that as from those dates a radio equipment, if it listed in Annex Ia and is capable to be recharged by means of wired charging at power above 240 W, must incorporate the harmonised charging solution up to 240 W.

The Commission will continue to update the technical specifications set out in Annex Ia, in order to reflect scientific and technological progress or market developments provided that they meet the objectives of the common charging solution.

Haha nice!

There is the Mint version that is based directly on Debian, instead of Ubuntu: LMDE

I haven’t used it myself, maybe someone with actual experience can comment on it.

Link to where that sentence appears: https://mullvad.net/en/blog/force-all-app-traffic-into-the-tunnel

So if you say they are a lucky couple, are you offering to be their third?

DNS blocking, like with a Pihole, famously does not remove Youtube ads. So no, the mechanism is totally different.

Not at all, if you actually read the article, in which the connection to Signal is laid out clearly.

What’s reason we are talking about Signal here.

That’s very easy to answer. Signal comes up here because that’s what actually happened in the case that brought the bug to light.

The FBI recovered message content from the Apple notification service that happened to come from Signal, and used it to secure a conviction against people who used Signal in planning their anti-ICE activities

Unlucky for Signal from a PR perspective but that’s just what actually happened, so people write about that rather than hypotheticals where other encrypted messaging apps would have suffered the same issue.

That’s not what I read unfortunately. When booted in secureboot, the kernel enters lockdown mode which disables all hibernation, regardless of the swap being encrypted or plain text.

It seems there are two kernel patches available to enable hibernation in lockdown mode, but not in mainline.

This one is more of an admin override, where you take the risk of root replacing the swap contents

https://gist.github.com/kelvie/917d456cb572325aae8e3bd94a9c1350

And this one is complicated but uses the TPM to ensure only the kernel, not root, can write the hibernation image in a way that causes it to be trusted on waking, so there is no reduction in assurance compared to clean booting a signed kernel with secureboot:

https://mjg59.dreamwidth.org/55845.html

But that’s all too much for me, I intend to turn it off again.

I recently learned that the option to hibernate goes away when Fedora is booted with SecureBoot. It was surprising to me, and might be good to know for you, that’s why I’m mentioning it.

True most motherboards, even the normal ones, now come with 2.5G included. But upgrading to 2.5 G feels like a wasted middle step if the next tier of external connectivity is at 10G, so I’ve not done that either haha

It’s only more profitable if there isn’t competition. He lays it out quite well in his blog post. It’s not like the Swiss ISPs are all publicly owned.

Just for reference Init7 offers 25 Gbit/s for 65 CHF a month. Thats about 83 USD.

They have the same monthly price for 1 Gbit/s 10 Gbit/s and 25 Gbit/s. Only the initial install for the higher speed optics costs 77 CHF or 222 CHF more respectively.

I’m still on their 1Gbit/s service because I’m too lazy and cheap to replace my router and LAN with 10 Gbit/s equipment.

work united on 1 distro and 1 distro only to make it better for everyone

Can’t wrap your head around multiple specialized distros existing, or what’s the issue?

She answered that in her blog post that the Phoronix article links to:

Which GPUs does this work with? Is it only AMD GPUs?

Whether or not your GPU can benefit from it depends on the kernel driver - more specifically, whether it sets up the dmem cgroup controller.

amdgpu and xe both have support for the dmem cgroup controller already. In theory, Intel GPUs running the xe kernel driver should benefit as well, although I’m not sure anyone tested this yet.

For nouveau, I have sent a patch for dmem cgroup support to the mailing lists. This patch is also included in my development branch, so if you use my AUR package it should work. In other cases, you will need to wait for the patch to be picked up by your distribution, or apply it yourself.

The proprietary NVIDIA kernel modules do not support dmem cgroups yet, so this won’t work there.

deleted by creator

I suggest reading the first paragraph, it addresses your point.

I believe the error was in the AV1 license NOT having a “if you enforce patent-license-fees on this codec, THEN you can’t use this codec” type of coercion…

I thought those provisions were usually enforced among the members of a patent pool, to ensure that any licensing customer can trust in the pools word to not be shaken down a second time by an individual pool member later.

So since AM1 isn’t forming a patent pool to sell licenses, and Dolby isn’t part of the Alliance for Open Media, it wouldn’t really apply either way, no?

Oh wait, actually there is something like this, see point 1.3 here: https://aomedia.org/license/patent-license/