got a 404, anyone else can confirm?

I have a similar experience where jpeg encoding and webp encoding result in far less computing resources use that jpeg XL or AV1, and was curious at what other people used (as I might be using the wrong library).

MozJPEG encoding times are not that great but decoding is still fast.

I believe with Jpegli you can have faster encoding than with MozJPEG.

One issue is that format support detection was iffy, compared to Jpeg XL where people knows to use the accept header to declare support format.

And that's probably the gist of it, manufacturer find it cheaper to get rid of a legacy piece of infrastructure than "fixing the shielding" (and also support the vehicle that are already out there).

The funny thing is, some EVs like the Prius had AM radios all along. It seems like some manufacturers might be taking short cuts. I guess maybe interference regulations aren't strict enough to force the shielding issue.

A prius has an electric drive unit that is about 1/4 the power of a full EV. It also is only really used for getting the car off from stops.

Great analysis. This must be the crux of it, meaning the engineering tradeoff between good EMI mitigation and lower cost. Shielding is one way to go. I do not know the details here for cars, but in digital systems people tend to use spread spectrum clocking (SSC) to lessen the interference. Others on the thread seem to suggest the culprit is not the electronics, though, so that technique probably is not applicable.

I feel that in "CDN" the C is way too close to D, some uppercase character spacing seem to close, defeating the "readability" purpose

For the main purpose that they are used for, displays that don't need to be updated frequently, like e-reader or "smart" price tag display.

However updating the display is more costly in terms of power than lcd.

We can also factor in that most e-reader also include backlight.

As a caveat, I appreciate e-ink technology, and the innovation that it brought us. I would also love to get my hand on one of those color e-ink android tablet. But I am not hoping to have a huge difference in battery usage from it, e-reader are very specialized device and can afford optimization that more versatile device cannot.

The point was also probably that there is already a C++ compiler in the base, vs adding new headaches for adding and maintaining a rust toolchain in the base.

But that's an interesting point. Why is there a C++ compiler in the base? Do I find a long discussion on the FreeBSD list about whether it should have a C++ compiler? If not, why not?

Because the System has to be self-hosting, remember it's a whole OS and not just the Kernel, and as you can see ~25% of src is C++, 62% C etc.

Rewrite everything C++ is not a small undertaking with (maybe?) not allot plus-sides, and even then you still need C++ because of LLVM. LLVM and GCC are both C++, and you have to compile the Compiler (self-hosting), and even if your a 100% Rust or C OS, you still need C++ because of LLVM/GCC....well if you want a modern compiler ;)

https://github.com/freebsd/freebsd-src

But when you look what all that C++ is doing in Base, it seems like it's basically LLVM and then a lot of code that's just there to help you write more C++ code. Not a whole lot of the FreeBSD Base userspace, the uilities and so on - is actually written in C++ despite that. It's getting a free ride because the compiler uses it.

>it seems like it's basically LLVM

Not important, you have to compile your Compiler fini, btw for Linux (just the Kernel) you need Perl ;)

Because most C compilers also include a C++ compiler, or that the compiler toolchain of choice for freeBSD happens to be LLVM (it was an old version of GCC before because newer version use an incompatible license, and before that I think gcc replaced pcc in the original bsd unix, the discussion to pcc with gcc is probably lost in time) that includes C and C++ compiler.

I am guessing that performance of current standalone C compiler is not as good as the one produced by compiler suit.

Following the conversation thread, rust also use llvm under the hood, but a customized one. Which would force the base system to have two versions of llvm.

It’s easy to build Rust with a stock llvm. You may just lose out on some back ported bugs that were patched in the fork the rust team maintains.

As you might know windows is kind of cheating with the "instant boot" by creating an hibernate snapshot of the OS before login.

Otherwise, I while I would not describe the boot process of most of the linux I own as "instant" they are certainly booting quite faster that windows.

Didn't know that one, thanks.

That's for the case for MBR partition type, I really hope we moved to more modern alternatives

GPT puts no size limitations on the FAT32 EFI System Partition. Your bootloader can be as big as you want it to be, which is why just booting off of a Linux kernel image with an initrd in the same file has been a valid option for years. Not sure why Lennart feels compelled to reinvent this particular wheel again.

I think someone else was fined 355~450 millions for fraud recently

If you scam the powers that be out of some real money (or just slightly expose their financial incompetence) that's when they suddenly know how how to hold individuals responsible. Pharma bro, Elizabeth Holmes, Delvey, SBF, etc.

The system views individual lives as less important than the confidence / stability / illusion of power of the holy market and those it is meant to benefit.