Seriously? I get that running Linux on random hardware is the modern equivilant of getting doom to run on a toaster. But, this machine has been out, what a year an a half (two?) now? And we are talking about the battery indicator? Isn't that just part of the ACPI battery spec, of which this laptop runs windows doesn't it?
All this really says is that no one should touch this hardware with a ten foot pole, given a random AMD laptop tends to work out of the box, and costs 1/2 of that thing, and likely outperforms it by 2x. And these days, most of these AMD integrated GPU laptops are in the 3-4W range for the entire machine, running linux+browser/etc. Which means your getting ~15-20hours battery life with a 65Wh battery.
So, its someones fun hobby project to get it working in linux, which will be useful when I see them on sale for $150, but until then I will stick with laptops that are supported by their manufactures and can have their firmware updated with fwupdate.
> I wouldn't touch non-ARM64 laptops for the time being.
Why? I would contribute most of that success to Apple's engineers/management not to the CPU architecture. I mean Qualcomm still seems to be behind not only Apple but also AMD and even Intel if you need both high performance and high battery life.
Is the focus of your ire the user who got Linux working, the hardware vendor, kernel developers, the person who submitted this to HN, or all of the above?
FWIW I wanted this laptop for a while and a big blocker for me was a lack of Linux support at the time. AFAIK a bunch of people have been working for a while to get this specific machine working with Linux. To me it's pretty cool.
I'll openly admit this is unreasonable for the "average consumer." So I'd never think to suggest mucking about with their systems. However, for tech-literate folk, if you really need extreme battery life, limit short/long TDP max values.
If you are willing to use a noticeably slower machine because you value battery life, then instruct your machine to behave accordingly. It takes seconds and is way easier than writing any program. Zen 3+ cores are still quite fast with a 10/15 watt long/short power limit. On the Intel side, CPUs can further be instructed to disable all but 1 P-Core.
Modern processors are pretty dynamic. I'm not pretending this will get you to arm levels of idle power draw, but things aren't an order of magnitude off, either. As a bonus, when you are plugged in, you get way more power on tap compared to any snapdragon.
My old x270 got that kind of battery life, on Linux, and that was 4-5 years ago. Granted, it was an Intel i5 and not AMD, but AMD had a negligible presence on the laptop CPU market at the time.
Every laptop I've run with Linux in the last 6 years just works out of the box with pretty much everything working: Lenovo, HP, Acer, Dell. I don't know if it's a distro thing or if I'm just lucky or what, but I never really need to do much special.
couldn't agree more. I'm all for hacking together hardware but these read like it's a solved problem, finally, linux on the laptop! Each and every time, the battery life is garbage, suspend is broken, and wifi barreeellly works, like it's 1995. And, in 1995 that was a cool feat. Today, these read like a broken record. Can we set the bar a bit higher beyond just booting into a shell?
And, these are just the issues out the gate. Unfortunately, these are dealbreakers especially when there are competitive options available that are not gimped.
Forking over 2-3k to see how far I get is something I have no appetite for.
Well without work like this and posts like these Linux would be a lot less usable. The first time I ventured outside of Windows I had to do stuff like configure ports and hunt down wifi drivers. Today for most cases it seems to "just work" and work like this ensures it stays that way.
Of course audio is not (yet) working (: