gf000

35 ms·

931.
by gf000 1y ago
AFAIK GrapheneOS is in the talks with a hardware company and we may end up with a GrapheneOS-first device!
932.
by gf000 1y ago
Solved dependency management would involve reproducing your artifacts in the future - isolation is a completely orthogonal thing and container tech just sidesteps this issue.
933.
by gf000 1y ago
You currently list 720 apps, Google Play Store has 3.3 million. Even if we assume that a good chunk of that may be "duplicate" in terms of functionality (e.g. todo apps), that is still just a completely different dimension of apps
934.
by gf000 1y ago
I mean, they were pretty trivial electronic components with a tiny screen wrapped in a kg of plastic. Like, what even could go wrong with them? But modern smartphones are anything like that, and people do like the more premium materials on
935.
by gf000 1y ago
Why? It's not like any of them is completely open-source hardware, so why bother? Pixels have decent hardware and have an open-bootloader. I'm fairly sure your desktop is not better/worse than that either.
936.
by gf000 1y ago
And let's be honest, it's not like there hasn't been spent a decade or more, with thousands of developers working tirelessly on making Android as good as it is today. Like linux laptops not burning up immediately is itself a
937.
by gf000 1y ago
What about problems that are not "one thing" only? Like, should a web browser be 640 piped subprocess in a trenchcoat, would that really be maintainable/easier to see how it works?
938.
by gf000 1y ago
> With Wayland, even if I start with an existing compositor, the barrier is far larger, and everything changes. I mean, no one puts a gun against your head to use Wayland, X will be on life support for decades and will likely work withou
939.
by gf000 1y ago
Containers don't solve dependency Management, they just push it a layer up so it's only someone else's problem.
940.
by gf000 1y ago
> Linux's security model is based on trusting the software you're installing from the FLOSS repositories, That's not a security model, and we don't live in fairyland. Just take a look how well this works with npm pack
941.
by gf000 1y ago
By making a soft and then a hard suspend the reality they have to abide by, or otherwise they are killed and users will think they are broken apps. Mobile apps just had to "grow up" in this environment, plus they have proper APIs
942.
by gf000 1y ago
But I mean, why not take the 100s of thousands of man-hours that went into making Android this very Linux-Kernel based mobile OS and build on top? Will you be happy with xeyes and a terminal? Like, even a technically superior solution is co
943.
by gf000 1y ago
I would argue that the legacy OS and app architecture we have on desktop OSs doesn't make sense there, either. It's a model that worked fine in multi-user setups where you ran a single executable, so that the security per user was
944.
by gf000 1y ago
Well, populism is no good reason to do something dumb. Maybe laymen should not directly have a say over experts in deeply technical discussions.
945.
by gf000 1y ago
We are not in disagreement - my point is that is is a fundamentally civil/legal problem, not a technical one. There is no technical distinction between a functional and a tracking cookie.
946.
by gf000 1y ago
I don't know, I don't think it helps all that much when you are up against Facebook's, and Google's wits on how to circumvent it. If they can open a port and side-step the security system of Android wholesale, they can p
947.
by gf000 1y ago
> They're honor system, which is dumb when you could have browsers not send that data back without opt-in. Given that there is no objective way to differentiate between functional and tracking cookies, your "technical" sol
948.
by gf000 1y ago
The only thing where Apple is consistently ahead of the rest of the industry is the processor.
949.
by gf000 1y ago
On Android you can just patch and recompile the native youtube app on your phone to disable ads with ReVanced.
950.
by gf000 1y ago
> and remember the battery with a daily charge will degrade significantly. If it's borderline on release, it'll be inadequate after a year. That has not been my experience though - having used both an Apple Watch and a Pixel Wa
951.
by gf000 1y ago
But it's not the screen that causes it to lose energy as fast, but the general purpose OS with a decent CPU.
952.
by gf000 1y ago
> The other issue is that I don't want to have to bring Yet Another Dongle™ I think reverse charging from your smartphone is a quite decent solution to the problem, which is supported by certain Android devices.
953.
by gf000 1y ago
This is a misunderstanding of what essential complexity is. If it could be subdivided into its own small, manageable piece, then we wouldn't really have a problem as human teams either. But the thing is, composing functions can lead to
954.
by gf000 1y ago
That just rotates the problem 90 degrees - now adding new functions is easy, but adding new data is hard (requires rewriting every use site). I really recommmend giving this a read: https://craftinginterpreters.com/represent
955.
by gf000 1y ago
I tried to translate the article to Rust, but seems that the article's gotcha in Haskell is not necessarily an issue in Rust if we return `dyn OurTrait`, so now I'm even more confused. If anyone could take a look what I'm mis
956.
by gf000 1y ago
AI solves "essential complexity" the same way they solve the Halting problem... These are fundamental CS concepts, you don't solve them. Also, I would first wait for LLMs to have reliable reasoning capabilities on trivial log
957.
by gf000 1y ago
> I feel as if a lot of multipliers have happened that he didn't anticipate Such as? I think his essay still stands the time that no single multiplier is even close to an order of magnitude productivity boost, with the exception o
958.
by gf000 1y ago
> this means that accidental complexity has not disappeared, it has only moved I would argue that it is essential complexity that we reuse (what the library does), at the added cost of some accidental complexity from dependency manag
959.
by gf000 1y ago
Well, you can reuse all that man-hours that went into the JVM - this is what Flix does!
960.
by gf000 1y ago
One important thing that languages with algebraic types can usually do and most others can't is being generic in this "colorness" - precisely because this is the color problem they are solving. So they can e.g. have a `map` f

More ›