I’m sure the obvious answer is tariffs, but the US phone market has been rough for anyone who isn’t Apple or Samsung due to the outsized role carriers play in phone selection in that market. Google has made some inroads by being Google, but HMD is not Google.

I like cooking with fire. Temperature changes (especially reduction of heat) are much faster than resistive electric, and when cooking on an unfamiliar stove, it’s easy to tell what’s going on; I don’t have to guess what “6” means on a dial because I can look at the fire and see.

Both the awareness that gas stoves are a significant source of pollution (mostly nitrogen oxides) and availability of induction are fairly recent and not universally distributed. I’d accept the pollution for a better cooking experience than resistive electric, but induction is pretty compelling all things considered.

It’s likely their priority is continuing to collect all the fees they can for as long as they can rather than the fine itself.

I’ve used Linux as my primary OS for many years, but I keep a copy of Windows on my current laptop for gaming. I know the gaming story on Linux is pretty good now, but having to hibernate and reboot is enough of a barrier to launching a game that it helps me stay more productive.

At home, my laptop sits on a stand with great airflow, but when trying to play The Witcher 3 on a desk while traveling, it overheated and throttled to the point it wasn’t playable. On Linux, the thinkpad_acpi driver allows setting the fan level to “disengaged”, which sounds like “off” but actually means unregulated and results in a considerably higher speed and cooling performance than the usual maximum. Some research led to the conclusion that while manual fan control is possible with certain apps on Windows, there is no way to exceed the maximum automatic speed.

It only took a couple minutes to set up Lutris and Proton to run the game, and as expected the mild abuse of my laptop’s fan does make it playable. What I didn’t expect is considerably faster load times, but I got those too.

This is not one of the claims made by the ICEBlock developers; their claims are only to do with notifications.

If you want to claim that a locked Android device is substantially easier for law enforcement to break in to than a locked iPhone, please cite up-to-date (from 2025) sources.

It makes me suspect they’re not talking about the stock systems OEMs ship.

The developers of GrapheneOS, an independent, security-oriented Android distribution are probably not only talking about stock OEM Android. What they’re saying is true about stock OEM android though.

That’s a separate issue from whether users are forced to get all their software from a specific source, which is also separate from whether users will actually use other sources when given the option.

On Android, developers can offer users a way to install an app that isn’t easily traced to their identity and on iOS they can’t. Furthermore, an Android app can be both on the Play store and available from other sources; there’s no exclusivity.

It’s true that FCM will result in more reliability and a better UX than other ways to implement notifications. Doing something else is still the right choice for certain use cases, such as those where privacy or keeping the entire codebase open source are top priorities.

Maybe they want that, but the statement on their website is not wrong on a technicality because it’s oversimplified; it’s wrong because it asserts a privacy difference between the two operating systems that does not exist.

The link in the comment you’re replying to says which part is not true, but since you seem more willing to comment than to click a link and read, I’ll summarize:

The part about the Apple Push Notification service requiring less information that can identify an individual user than Google’s Firebase Cloud Messaging is not true. Both use a similar token system. Furthermore, it is possible to build android apps with notifications that do not use FCM.

I think generating and sharing sexually explicit images of a person without their consent is abuse.

That’s distinct from generating an image that looks like CSAM without the involvement of any real child. While I find that disturbing, I’m morally uncomfortable criminalizing an act that has no victim.

If it was all of Europe, I’d agree. That explanation seems improbable for just two countries.

I’m sure there are several I would consider fine for me.

I’m skeptical that they can be fine for someone who doesn’t know what federation means, isn’t especially upset that a handful of megacorporations control most human communication, and already finds the fact that I’m asking them to use anything different from what they’re used to annoying. XMPP has more things for the end user to think about than Signal does even if a client is very polished.

SMS works perfectly well for sending https://signal.org/install

That’s a risk, and a reason I’d like to see something federated succeed in this space. Unfortunately neither Matrix nor XMPP has managed to achieve quite the level of UX necessary for mainstream adoption, nor have the average person’s tech skills and comfort level improved.

Signal’s status as a well-funded nonprofit gives me hope that the current situation is reasonably stable.

No. WhatsApp came first, but later adopted Signal’s key exchange and encryption. WhatsApp was an independent company for years and had rejected several acquisition offers, but $19 billion is a big number.

One of WhatsApp’s founders is now chairman of the Signal Foundation and a major financial backer of the project.

Mastodon’s federation is not at all consistent even when it could get much closer with a little effort.

Servers don’t remote fetch old posts from recent follows for example, nor replies to off-server posts from people on a third server. There’s work being done on both, but I’m surprised it wasn’t prioritized much earlier. Some other Fediverse software handles these situations better.

I think the fediverse has a built-in legal risk in that any time someone posts, data is sent to a large number of servers when then make it available via the web or sometimes push it to additional servers (e.g. by user boosts or community subscriptions). This is currently done without any explicit license for the IP contained in that post.

I’m inclined to think that irrevocable permissions are the right thing here, in large part because it’s impossible to guarantee that any subsequent signal from the original poster propagates to everyone who has a copy of that post, or that the server software responds how someone else expects it will.

There is a legal basis: congress passed a law, the president at the time signed it, TikTok sued, and the Supreme Court unanimously ruled against TikTok. That’s a legal basis by definition.

Perhaps you mean that there is no rational basis. That’s a reasonable position you can argue for.

And that is what I would recommend against, even on a server that does not ban that age. If someone’s (young) age is relevant to a discussion they wish to participate in, I would suggest a throwaway account.

How were they revealed?