As far as I understand, audio cards hold a buffer of the audio that should be played at any time. If the CPU can’t keep up producing new audio, it will loop to the beginning of the buffer. My guess is that when you suspend, the CPU stops producing new data before the audio card stops consuming it. And that’s why you hear the last part for a short instant.

It also happens on my devices, and it’s always happened on all my previous devices as far as I can remember.

Disclaimer: this is based on my understanding + a lot of suppositions. It might not be accurate.

Sure, that’s what I thought. And Ecosia doesn’t do the same?

I do have an opinion about Ecosia, but it’s just based on feelings, so it doesn’t even make sense to share it. Apologies for not answering your question.

Instead I would like to focus on this point:

Ecosia isn’t very private, since it sends data to Bing

Also DuckDuckGo does this, but they aggregate and anonymize that data before forwarding it to Bing. That’s probably the best they can do without building their own first-party infrastructure. I would imagine Ecosia does the same.

I am not OP, but that would be the ideal solution for me. Unfortunately, KPXC does not support communication with the GPG agent and the team is not interested in adding this feature due to it being «[…] far more complicated than ssh-key management. There are already excellent tools for this, Kleopatra being the best».

Perhaps a bit unconventional, but CloudFlare R2 gives 10GB of free storage accessible as S3 with rclone.

Not FOSS, but OneShot: World Machine Edition is a great game based on this idea.

You might want to look into Island/Insular. https://f-droid.org/packages/com.oasisfeng.island.fdroid

Depending on your level of commitment to privacy, this might be a suitable solution for you.

[…] re‑engineered flash physics by replacing silicon channels with two‑dimensional Dirac graphene and exploiting its ballistic charge transport.

By tuning the “Gaussian length” of the channel, the team achieved two‑dimensional super‑injection, which is an effectively limitless charge surge into the storage layer that bypasses the classical injection bottleneck.

That’s some seriously technical jargon.

ChatGPT seems to be able to explain, not sure how accurate it is though.

Flash memory traditionally uses silicon channels to move charges (electrons) into a storage layer. These researchers changed that by replacing silicon with Dirac graphene. Graphene is a single layer of carbon atoms arranged in a honeycomb structure. It’s called a Dirac material because its electrons behave like massless particles, moving extremely fast and with very little resistance.

This leads to ballistic transport: electrons move without scattering, like a bullet in a vacuum. This is far more efficient than silicon, where electrons bump into atoms and lose energy.

Tuning the “Gaussian length" likely refers to modifying the shape or spread of the electric field or potential in the channel (possibly shaped like a Gaussian curve, i.e., a bell curve). By adjusting this, they control how charge flows.

Achieved two-dimensional super-injection means they were able to push a large amount of charge very efficiently from the graphene channel into the memory storage layer, and in a 2D way (across the flat graphene surface), rather than through a narrow point.

Effectively limitless charge surge: normally, in flash memory, there’s a bottleneck where only so much charge can be injected due to energy losses and scattering. But with graphene’s ballistic transport and this super-injection method, that bottleneck is gone—or drastically reduced—enabling faster and more efficient memory writing.

Fair enough, but then it’s the same thing as open-sourcing the code but not providing support nor binaries.

I mean, personally I also prefer it to FUTO’s proprietary license, that’s for sure. But I’m one of the few privileged users who can build from source.

If this license doesn’t impose any extra restrictions on the code (and as you say, anyone can fork and provide prebuilt binaries), then this would just increase the risk of spreading malware, with no real benefits for the original developers.

In my opinion, if you want to monetize your software without going proprietary, all you have to do is provide the users a convenient way to get it. There are some paid FOSS apps on Google Play, as well as some paid FOSS games on Steam. You don’t want to distribute binaries? Fine, okay, that’s alright and I respect your choice. You don’t want to provide support to non-paying users? Fine, that’s very reasonable in my opinion. But…

…do you want to impose extra restrictions on your code? Fine to me, but then you are no longer doing open source, don’t try to pretend you are. And if you are not imposing any restrictions on the code then you are imho just going to hurt small users. We shouldn’t fight small users imho, we should fight the big corporations exploiting FOSS code for their proprietary businesses. But if there are no extra restrictions on the code, then big corporations wouldn’t care.

That’s my opinion.

I’m not a lawyer, but this doesn’t seem to be compatible with (A)GPL licenses.

I would say this is going to harm small users more than big corporations. As a small user I might be unable to build from sources myself, so I would have to pay. But as a big corporation building from source would be something I can certainly do trivially, then I wouldn’t be subject to the restrictions imposed by this license.

Imho, if someone wants to force their users to pay, then they are not doing open source. Please let’s not try to pretend we are by adopting a OSI-approved license and slapping extra restrictions on top of it.

Just go AGPL for datacenter-oriented softwares, or GPL for drivers and embeddable code, or a proprietary license such as FUTO’s for end-user software.

If you don’t want proprietary drivers the choice is quite straightforward: AMD. The official drivers are open source.

As for my experience, I’ve had absolutely no problems in the last few years with AMD, but I have to admit that I have always been using an iGPU, which has always been good enough for my needs.

I used to have problems with Nvidia proprietary drivers, but that was at least a couple years ago, things might have changed. I’ve never had issues with the free unofficial drivers, besides worse performance.

I filled your survey. It would be nice if you could share the results once it’s completed.

GrapheneOS is certainly on my wishlist too, but Pixels are quite pricey. I guess Rethink is the poor man’s version. Just a per-app firewall.

You should install Rethink and see how much garbage your phone constantly transmits and receives. And this is not even a kernel-level firewall, so who knows how much data Google actually exfiltrates…

I don’t know about a constant audio stream, nor about keywords, but I noticed that Google Keyboard sends out some data every time you type anything. It’s not even that subtle.

I use DDG for the privacy as well, but personally I think it works better than Google in my field (software development). The only issue I personally have with DDG is that it lags behind Google in terms of updates, I notice when searching for something that came out or happened only recently.

I thought so. Although almost nothing for modern standards, 60MB is not exactly tiny. Sorry about that.

On a different note, a repository is always a good thing imho. If you’d rather not have to worry about the dependency-pull step you can always include the dependencies with your sources, or just limit your code to using features included in the standard library.

JavaScript through Node.js, or TypeScript through Deno if you like typed languages. They both check all your boxes (just check the size of the executables to make sure that it’s what you would consider “small footprint”).

Both languages and runtimes are quite popular, so you will find any answers on StackOverflow.

They are both single-executable with little dependencies, and Deno can also compile your scripts to self-contained executables.

As a bonus, both support the vast and extensive NPM package repository where you can find all sort of libraries for even the most complex tasks.

And they work with your favourite IDE or editor, not just syntax highlighting, but also contextual suggestions.

I’m still a bit confused by the use of this “Driver Store”. Since when does Wine support device drivers? Or are we talking about something else?

Phoronix seems to explain a bit more, but I still did not understand: https://www.phoronix.com/news/Wine-9.16-Released

Could anyone share their insights?

I’ve never had the chance to work with the RAW format, but I think Photoprism should handle it transparently. Depending on your area of knowledge, the setup might feel a bit convoluted though.

So… this is going to be a debloated lightweight distro (or just a DE?) with official support for not-so-fresh hardware? I don’t fully understand.