4A says "The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures" so I would think the provider of the service could consider their data to be "their papers/effects" but is the provider a member of "the people" if it's not a sole proprietor?
Yes, providers can absolutely deny requests that aren't lawful. A company is within its right to say the data is their property stored on company property and a warrant is required to search it.
They probably could refuse, but isn’t selling access to surveillance information about you part of their business model? As they say, “we value your privacy…”
While that is true, both Espressif and the Pico have their own SDKs, and they're really well written too.
The Arduino SDK is the simplest to use, sure, but the Pico framework (I don't have experience with the Espressif one) is extremely good, and the Pico's PIO is a godsend. I used it to implement 3 wire SPI (data bidirectional on the same wire) at almost 'real-time', which is to say, at half the speed of the hardware SPI controller (half the speed because the interface clock is put up one cycle and down the next; this also gives enough time for data shuffling).
Why does the Arduino SDK necessitate a huge markup on Arduino boards, when $0 of every computer I buy to run Linux on goes to GCC?
Just because most of the free software ecosystem relies on unpaid volunteer work does not mean it is a desirable state of affairs, especially with billion dollar companies building on top of said work while hardly contributing anything back.
Both Espressif and Raspberry pi (pico) target OEMs who will buy millions of their chips. They've both embraced the hobbyist market as well, but it's not how they've recouped their investment.
Arduino targets the hobyist market where customers will buy one (or at best a handful) of their boards. Arduino simply has no other way of recouping their investment than selling expensive hardware.
So I don't think it's fair to say that Arduino is being greedy. Also FWIW, Espressif's official dev boards are also pretty expensive. Not Arduino expensive, but several times the price of identical "clones" based on the same reference design and using the same official esp32 module.
If you think the price is unreasonable, don't buy. You have listed what you seem to think are better options. I agree that there are better options. If somebody else wants to spend their money in different ways than I do, let them. If Arduino thinks they can make money this way, let them try. If it works, good for them, I guess. If it fails, I guess the joke will be on Qualcomm. Honestly, Arduino could slash their price to be $1 less than a Milk-V Duo and I'd still by the Duo. If the Arduino was $1 less than an ESP32, I'd still by the ESP32. So I'm not sure lowering prices wouldn't just hurt them.
I have never bought an Arduino. I have bought a few Picos, a few ESP32s, and a couple Picos. And a clone of an Arduino Nano integrated in a system with a Pico for 5V logic, specifically, to implement a PS/2 controller. I don't see any advantage an Arduino has over an ESP32, aside from 5V logic support.
> Precisely why brand drugs continue to make money over fist for pharmaceutical companies even after patents expire.
Generics may have the same active ingredient but (vastly) different pharmacokinetics - i.e. different absorption rates/retention in the body. For basic stuff such as painkillers that's one thing, but for more sensitive medication such as insulin, antidepressants or anything related to the cardiovascular system (heart rate, blood pressure and clotting) one has to be very careful when switching between brands.
Yep. And that's exactly why the EU has the structure it does.
Unfortunately the only country that ever left proceeded to shoot itself in both knees, light itself on fire and jump in a pool of gasoline. For NO reason.
This is not an ABI stability problem, this is a UX problem.
iOS and Android do backwards incompatibility all the time. If you find a mobile app that hasn't been updated in 10 or 15 years, you won't be able to install it on a modern device. But, iOS and Android ship apps to billions of users, nobody complains about ABI stability and nobody uses a Win32 compatibility layer.
glibc already has decent UX in that it doesn't allow you to load an app built for a newer version on a host with an older version. Unfortunately, the message is not user friendly at all "could not load @@GLIBC_X.YZ@@foo()", instead of something more readable, but the restriction is, in itself, good.
The problem is that there is no system to trigger backwards incompatibility at any point, nor is there a simple way to target an older Glibc version without spinning up a docker container.
And glibc is a HELL of a codebase, I would NOT want to be responsible for implementing those features, so I understand why they're not there. But 'our Linux build for our game we are still selling and advertising as working on Linux is no longer compatible with new Linux distributions, so let's rebuild it' is a much easier decision for a developer, publisher, etc. to make than '"some" Linux users are reporting issues with our game, should we dedicate resources to investigating this?'
> nor is there a simple way to target an older Glibc version without spinning up a docker container
? I feel like you might be falling into the trap of assuming that the compiler always targets the current running system, but you 100% should always always always be doing a Linux-to-Linux "cross" compile with a sysroot... you don't need a docker container: you just extract an old Linux distribution into a folder and tell your current/modern compiler to target it.
Spinning up a Docker container the way I'm kind of assuming you are talking about is going to involve running the compiler in there, which is super annoying -- and extremely limited, as you often are going to get stuck with old versions of all of your toolchain components as well -- and certainly way more difficult than just extracting the distribution to disk and using gcc --sysroot=/path/to/glibc-2.31.
Because if function foo's third flag parameter supports FOO_ENABLE_FEATURE_X starting from 2.31, and a program only occasionally calls foo with this flag, there is no good way to detect this incompatibility on a 2.30 system at link time.
The restriction is only annoying because there is no way to trivially link against an older "minimum requirement" ABI.
Yes but there is simply no reason to have two devices. There are a large number of Windows tablet-laptop combo machines that work perfectly well and prove touch apps work perfectly well on a desktop OS.
Yeah, that took a long time for MS to get to not suck after Windows 8, but touch and tablet interactions on Windows 10 and Windows 11 work perfectly well.
Having owned at least 3 such devices, I have to disagree. It "works", but desktop apps expect desktop interactions, and touchscreen functionality feels cobbled together at best. There are a handful of apps developed specifically for touchscreen PCs that work well, everything else is a toss-up. On the other hand, apps developed for a tablet OS support touch as a first class interaction, and have OS support for hardware keyboard and (usually) mouse input if you so choose. Not to mention that the vast majority of combo machines I have used are too heavy to use as a tablet for any reasonable amount of time, or have an incredibly clunky transition method. I have yet to see a platform where touch and mouse can both coexist as first-class input methods. Even the cognitive load of transitioning is an irritation.
Once again, the EU coming 30 years later to do something everyone else is doing, and everyone will credit them for some odd reason.
In the 90s, there were dozens (hundreds?) of phone charging ports. A couple of years ago, there were only two. The dozens -> 2 simplifications occurred on the purely free market. And the EU mandated a simplification from 2 -> 1 and gets the credit for the entire simplification.
What is the point of this license? Either the GPL is invalid in the EU, in which case why aren't companies moving to the EU to infringe on the GPL? Or it is valid, in which case a a bunch of EU lawyers were given a bunch of MY money to do nothing of value to anyone.
Whenever something from the EU emerges on HN, whatever that is, a comment exactly like yours is made.
Today it's you, tomorrow someone else.
Always the same "points" and the poster always seems to feel obliged to bring their own misshapen view of history, forgetting that a lot of good changes in this so-called "free market" happened because of the EU.
Sure more tech became popular out of the US but let's not be blinded:
1. This tech isn't/wasn't the only innovative tech ever created
2. A lot of reasons for the popularity has to do with US position over the rest of the world
3. That this same tech has been kept at bay because of the US
I'm not saying the EU is perfect but I prefer its direction over the US, especially when it comes to making legislations which benefits the people rather than multinational corporations.
I simply do not think that the popularity of US tech is due to anything other than the quality of US tech.
I live in Romania, and the people that were rushing to pirate Windows in the 90's after communism fell because they couldn't afford licensing weren't doing so because of any US imposition, but simply because they wanted to use personal computers, and Windows was the best OS for most people at the time for that purpose.
Just like when phones came around they rushed to buy Nokia phones; when smartphones came around they rushed to buy Samsung phones; when they wanted DLSRs they bought Canon and Nikon cameras and now that they want easily transferable digital cash and cheap tech trinkets they opened up Revolut accounts and order stuff off Temu.
Not because of any "influence" or "position" of Finland, South Korea, Japan, the UK or China, but simply because they are the best offer on the market as perceived by consumers.
What tech does Europe lead in? To be fair there are still some fields, like Aerospace (Airbus), Lithography (ASML) or Pharmaceutics (BioNTech). But on the consumer tech market, the phones, laptops and streaming services people want? The EU has no presence. Even the auto industry is going to be eaten up by China, because Europe simply pivoted too late to EVs. I know someone who works at Renault and they're just terrified of the cars that are coming out of China.
I see your point but that stops being true once the bigger companies start stifling competition (both locally and internationally) to become even bigger and more pervasive.
There are European alternatives for many products/services that are currently US-based but they either don't have the same marketing budgets, or international reach or can offer lower prices. All those are often due to first-to-market tied with anti-competitive practices (further tied to governments not having much power against these organisations) which makes these companies move even more to the top. None of that was about better tech but rather everything else.
And sometimes yes, better tech, but let's not fool ourselves into thinking that "better tech" is not somehow tied to more money available to spend on improving such tech. Big capitals being moved around the US has an impact to this tech.
By the way, there are European alternatives Amazon (just not big enough to be international, but many countries have their own version), Microsoft (i.e. Linux), Google (ProtonMail/ProtonDrive , Nextcloud, etc.), AWS (the ones which comes to mind are Upcloud and Hetzner), Android/iOS (Ubuntu Touch)
That was not done by the European Commission. It was a document with no legal power. And it was an acknowledgment of an existing market trend. non-iPhones had already began converging on USB charging (including mini-USB and micro-USB) in the years before.
By the time the EU actually first proposed regulation on the matter, which was only in 2020, there were in effect three ports on the phone market: the Apple Lightning was still used in iPhones, and the rest of the market was undergoing an orderly migration from the cheaper micro-USB port, to the more expensive but better USB-C port.
So yes, the free market was entirely responsible for transitioning from dozens of ports to 3 ports, and would have very likely eventually transitioned to 2 ports. The fact that the EU made recommendation to that effect years ago after the trend had begun that was purely voluntary is an entirely irrelevant datum.
That's not really what happened. The European Commission made gave the manufacturers a choice: they could sit down and come up with standard that they voluntarily accepted to follow or the EC could write the standard and force the manufacturers to follow it [1].
I imagine they would have eventually converged on USB anyway, but when the upcoming rules (or "rules") were announced, you definitely could not count on being able to charge your phone using anyone else's charger (or one of the many that had come with your previous phones).
Counterfactuals are tricky, and we'll never know for sure what might have been, but seeing laptop manufacturers dragging their feet, I really can't see how you could feel so certain that the market would have fixed the mess that existed just as quickly.
... Eh? It's from 2007. Licenses like this weren't at all common at the time; there was the original Affero license, but AGPLv3, the first version to see widespread use, didn't come out until later.
Well, because nothing in the transformer architecture supports such learning ability. All AI researchers and most serious AI users are aware of this, so I'm not sure I understand the question.
> And if you have a compelling thesis, why hasn't this spread to the investing community?
The investing community believes that they can make money. That's feels pretty much orthogonal to whether the metaphorical intern can learn, and much more related to whether clients can be made to buy the product, one way or another.
They mean it doesn't learn from experience/mistakes after spending time on your codebase. There are workarounds like documenting in CLAUDE.md/CODEX.md/.roorules which dump it back into the active context but are hit and miss in my experience. It's definitely better than nothing but Claude still routinely ignores important directives whenever it's in the right mood.
Can anyone using GrapheneOS report if Firebase notifications come in consistently and reliably via sandboxed Play Services?
I'm in the market for a new phone, and I'm going to buy a Pixel 9a this week for GrapheneOS if I can reliably get notifications on it. (I already have an A05 for banking apps)
I'd be happy to check for you, but will need concrete steps. Signal working reliably is pretty much all I can confirm, since I don't use any others apps on that device that would give me notifications. Signal afaik uses this Google Cloud Messaging thing or whatever their TCP connection marketed as push service is called. Maybe that's now called Firebase?
But Signal can also fall back to websockets, which I use on my personal phone and that's working great without any battery loss so I don't know how to tell the difference
Yes, all notifications work fine with sandboxed Play Services installed. All my banking apps also work fine. I haven't really had any problems with app support or any other problems for the many years I've run GrapheneOS as my daily driver.
Ok, but there's nothing stopping from writing a "discoverable" GUI or TUI or CLI for this interaction.
To the user it doesn't matter.
To the developer, it makes a difference.
/proc/<pid>/ctl is not any more or less simple or discoverable for the end user than TerminateProcess or kill(2). It's all opaque esoteric stuff.
But for the developer /proc/<pid>/ctl is MORE discoverable than TerminateProcess or kill(2), since it's a file you can list, so you know it exists. Yeah, you still have to read the man page (you always have to read the docs as a developer), but you know there's a ctl file that probably controls some stuff about the process there. And you already know how to interact with it (the same IO interface as everything else).
And that the client and provider can sign a contract forbidding the provider to disclose the information except under a warrant.