If you complete your death transaction without filing out a suicide and/or falling accident permit, you will be posthumously demoted!
If you complete your death transaction without filing out a suicide and/or falling accident permit, you will be posthumously demoted!
That’s why I was saying macOS is really the only option if your definition of “year of” is suspend/resume reliability. It highly depends on the hardware for Linux/Windows
The last two Windows laptops I’ve used (last 5 years), one wouldn’t suspend correctly (in suspend, it wouldn’t fully suspend and drained >5% battery/hour) and the other, on resume, couldn’t play audio without restarting
I’ve never had a Windows laptop suspend correctly, so…
I guess it’s the year of the macOS desktop?
You walked so I could run 🙂
I loved the idea of the framework when it was announced, but I wanted to see a couple iterations proving out it was really going to be upgradable and repairable
Loving it now
Just a heads up, if you’re on the 7040 mainboard, I needed to add this to the kernel command line on Debian 13 for reliable suspend/resume. Without it, the screen would just be grey sometimes and not resume
amdgpu.dcdebugmask=0x10
Edit: may also only affect the 2.8k display
Getting rid of that forced restart will at least help me personally stay more secure and get bug fixes faster
I get wanting to move away from “master,” but why in the world didn’t we use “trunk”
It was already a standard name, and it fits “branches,” etc.
ARM isn’t plug-and-play like x86 (n.b. it could be, but no one does it outside of servers)
You have to write a big JSON like file, called a DeviceTree, that describes exactly what is in the computer
Unless Apple decides to support Hackintoshes, their OS won’t have devicetrees for other devices.
You might be able to make your own and get the OS to read it, but it still has to be for a specific machine rather than generic like before
Microsoft seems to have deleted all their old blog posts around it (at least all the links to their responses 404 now), but…
In 2011, they were going to require it to be enabled for all products if they wanted to announce Windows 8 support.
After huge backlash, they changed it to allow motherboard companies to disable it for X years (since I can’t find the original, I can’t say the exact length of time)
But their original goal was for it to never be allowed to be disabled
postmarketos has builds for the 4/5, and Fairphone has already submitted devicetree files for the 6 to the mainline Linux kernel: https://lore.kernel.org/linux-pm/[email protected]/
Did that, did a lot of that. There wasn’t any doctor here who could shine my eyes. Not even for 20 menthol cools. Was anything you said true?
Yeah, their branding makes it harder to recover.
I don’t know how they’ll change their versioning in the future, so I just went with that.
If they don’t make an obvious split to when the extension system is stable, they may never get that new beloved version like KDE managed
GNOME 2 was different and easy to customize
GNOME is still in their KDE 4.x days where it needs time to mature.
KDE 3 was loved, KDE 4 made a ton of breaking changes, and was reviled. KDE 5/6 are now butter smooth and fixed all the issues from the 3 -> 4 transition
GNOME 4/5 will probably come back into the loved category if they start stabilizing the extension system some more
So, would your suspicion be that it’s causing them more failed boards in production?
I guess if it’s reducing returns, that might be something they’re accepting as a tradeoff?
If they can take my unlocked device by force, they can probably also break my fingers to coerce me to unlock it See also: https://xkcd.com/538/
Randall is right in pointing out you need to consider your attack vectors, but this doesn’t mean you shouldn’t take reasonable precautions
Most people are more likely to run into the type of attack OP references than someone who can break LUKS encryption stealing their device
I haven’t looked that closely at laptop CPUs
My guess would be partially because there are fewer possible interfaces, and they’re directly connecting the CPU to a separate Ethernet/WiFi MAC, USB hub controller, and audio DSP rather than having a separate chipset arbitrating who’s talking to the CPU and doing some of those functions?
For most intents and purposes
SoC is from the embedded system development world - as more and more coprocessors were being put into the same chip to consolidate board space and power efficiency, it wasn’t “just” a cpu - it had the CPUs, GPUs, DSPs, and other coprocessors in one
x86 has moved a lot closer to this architecture over the years, but you still generally have a separate chipset controller on the motherboard the CPU interfaces with
System on a chip. Think like a Qualcomm or Samsung processor, or the new M line from Apple
It’s from Futurama