It is funny with Linux (the kernel). From the outside it looks like almost nothing ever changes. You switch the computer on, it works away in the background to bring up the desktop and have everything talk. From that perspective it feels like nothing has changed in decades.
But then you look at the change log, and almost everything you see added sounds really useful to someone working on this stuff. 35 years of work on this and there is still so much to be done even if you most people using it never directly feel it.
Strongly disagree. Hardware support improved by leaps and bounds last decade and a half. Namespace isolation and cgroups driven containerization. eBPF appeared and gained adoption. Async I/O (and later just everything async) converged on io_uring. Btrfs is _stable_ - some people believed it would never be possible, ever. All kinds of scheduling were iterated and iterated over.
All these things are visible across the board, both from developers and end users side.
I have been using Linux on the desktop since the late 90's. Other than increase hardware support, I have not tangentially noticed/felt almost anything you have mentioned here. When I fire up the desktop and load Firefox, I do not feel or see the sandboxing, the scheduler, the file system or the I/O. It just works.
Yes, we interact with it all the time but it isn't something that the end user would notice and that is a great thing. If you are not directly poking around with that stuff, you shouldn't notice it.
That they are so seamless should be seen as a badge of honor.
I also have been using Linux for over 20 years and I agree that hardware support is the most noticable.
This year I have installed Linux Mint on computers from people who wanted to jump from Windows to Linux. Every time I booted via USB, installed Linux and 10 minutes later everything was working.
This would not have happened 10 years ago. Back then there would always be an issue with a video driver, WiFi driver or audio driver.
Of course there are still hardware issues for some special deviced, but today Linux just runs fine for most people. And that's a great accomplishment.
"I installed Ubuntu on peoples laptops and was extremely lucky they happened to have compatible hardware or didn't notice the stuff that didn't work and had a non-borked firmware". FTFY.
This was absolutely not everyones experience, including mine.
Only on a select combination of blessed components and BIOS/EFI not half-ass broken to the tune it only worked somewhat stable on Windows.
Anything with split iGPU/dGPU was a mess no matter what and where, WiFi that was not Intel was a mess, suspend to ram and suspend to disk were a coin flip if the system would wake up again or if it would decide to wake up on its own and overheat itself to panic.
Seems like parent thread spoke from perspective of developer/tech savvy person.
But even from end-user perspective, maybe its hard to see change when you compare one year or so, but over somewhat longer period I definitely see improvements. Over say a decade or so, wifi/bluetooth/trackpad support has significantly improved. Over last 5 years, I think gaming has become huge and you can actually use Linux as gaming machine. In last 3 years or so, I think I've seen my 5k display using Thunderbolt work much better. Fractional scaling also works much nicer.
NVIDIA GPUs are not shitty to use on Linux, and AMD has been working great for slightly longer. UEFI/TPM works much better now as well, where I don't have to worry about booting in legacy mode when using GPU drivers.
I didn't have to think about any of these above. If I think harder, I can likely come up with many more.
You may know more about what it powers. It has enabled much faster, higher quality, more stable security software on Linux. Many modern security tools are built on top of it. While end users might not know it by name, it has made Linux easier to support in the enterprise.
Well ya, it's more useful in observability usecases than desktop ones. But recently I've been using for a consumer desktop usecase. I made a file watcher with it. I'm sure you're familiar with the woes of using inotify/fanotify.
eBPF vfs hooks based watchers are really nice.
[edit] details
- you setup a single `mount` - this is the only part that requires cap_sys_admin
- then you run a single daemon, this one only needs cap_bpf, and this runs once at startup and is its own self contained program
- this daemon writes events into a ringbuf referred to by a file in that bpf mount
- you then read that from your standard user program (this can be done by whichever application needs to consume the events)
the events are create, mkdir, delete. renames are automatically decomposed since we hook at the vfs layer. the hook itself is practically free to run and thus there is no question of a performance impact.
For desktop I think it would be fine, though if you have the advanced needs of all the ZFS features I would stick with ZFS, but I had no problem.
The real problem I had was a low limit to save history. I install a lot of software, Cachy would run the snapshot every time (good), but it only kept like 5-10. And the one I needed to restore to was further back (I can't remember the exact issue), but that was more the defaults of how cachy set it up, and my use case than btrfs/snapper itself.
Whatever tool you use to snapshot btrfs, you may wish to have a custom number of saves if you install tons of software a lot...so you can avoid having too few backups available from the boot menu. I'm sure there was a way around it but I said f it and just reinstalled instead of trying to look up on my crap tablet how to get at the old files (assuming they were still there and just not accessible from the boot menu vs being completely erased).
I do not think it was btrfs that broke, to be clear.
This was like a week before the big hack on AUR, so I wonder if I got hit by a supply attack, but it was pretty shoddy, and broke my system, thus my reinstall saved me from (possibly) a lot more pain down the road.
It was the default on SUSE / OpenSUSE for years. Haven't used them for a while, not sure whether it still is. No problems with basic functionality including snapshots. I believe some RAID configurations were experimental for long time.
Been using it for a year or so on my main box. Never had any kind of issue. Snapshots saved my ass a couple of times. All in all it’s been pretty boring (in a good way)
Maybe. I think there's still no chance of repair when it blows up. Only FS since Reiser and early XFS that I've completely lost data to on healthy hardware.
Whichever one of my test machines is using btrfs is doing fine, though, so on a subjective level that one's been stable enough.
It's basically fine as long as you read the release notes carefully and don't do anything the release notes say not to do (e.g. RAID5, let the disk fill up). It still has some sharp edges, but they're all documented.
I imagine for some end users a lot of this stuff is happily invisible, just like you say.
But for me, a few kinds of changes have been highly visible as a user:
- basically all filesystem changes, especially the mainstreaming of copy-on-write filesystems
- not a change, but dramatically better filesystems performance than Windows or macOS has always been noticeable
- the addition of cgroups and cgroupsv2, and later user namespaces, which enabled lots of software that I use
- all kinds of driver updates and additions, obviously
- newer virtual filesystem stuff for introspection, set to replace /proc
- addition of KVM! I remember when this was not a given, in the Xen days
As someone who uses macOS at work for the last several years, I really miss the Linux kernel when it comes to filesystems, and feel like macOS updates often involve very few changes that I care about or even consider operating system change at all. Most macOS updates are just trumped up application updates.
In that way I've kind of felt the opposite somehow, like OS updates on proprietary operating systems feel thin and empty.
Isn't this the case with most well established software?
There is still a lot of work done on the background, perhaps even more than ever.
But unless you are a power user eagerly waiting for one of the things currently worked on, you wouldn't notice or care unless the interface changes (which usually only happens on visual interfaces).
I've been daily driving Linux since ubuntu 12.04. Things have improved a lot. It used to be the best OS for (a certain class of) programmers. It's now the OS I installed on my 7 year old son's computer. Things like connecting Bluetooth peripherals, multiple audio devices, multiple screens (with multiple DPI) are a not just possible, but easy now, and robust.
i remember back in the day, when i built my project, using 100% cpu, the whole system became lagging and choppy and you really can't do anything anymore.
Now it behaves more like Windows/Mac, even when 100%, my youtube/spotify tab can still run smoothly.
(Of course i think there will be hit in some areas, like my code would be built slower, there is no free meal).
Maybe it's distro packaging, but I suspect it's the kernel, too: In the 20 years I have dipped my toes into Linux, there is no better time to be a desktop user than today. With CachyOS and modern kernel, I have zero audio issues, zero GPU issues, (thanks to Wine/Steam Proton) no gaming issues, and much more rare instances of wifi or sleep problems. The idea of having to install hacky wifi drivers from USB stick are over.
Actually, i was waiting for 7.2 in order not to "install hacky wifi drivers from usb stick" XD, because mt7927 was not supported, so neither wifi nor bluetooth have been working out of the box
Not really. I never had this less issues on Linux with hardware. I remember 15 years back when I wanted to install it on a Desktop/Laptop, many things simply didn't work. Now you can install Linux on a self-assembled box and 99% of things will work out of the box.
Things that are unneeded, ie. unmaintained, eventually get purged. It happens quite regularly. For nearly every kconfig option, someone needs it enough to maintain it. But I agree, `make menuconfig' is overwhelming.
Does anyone understand how HDMI 2.1 support is now no problem? I remember that 2.1 support in AMD's open source driver was blocked by the HDMI forum, but I haven't heard any news about them unblocking it. What changed?
It doesn't give any specifics about what changed on the HDMI Forum side, but this article [0] from June indicates that Valve worked closely with AMD to get it pushed through prior to the release of the Steam Machine (which supports HDMI 2.1 VRR).
As far as I can recall: some internal implementation documentation leaked which made the legal position tougher for them somehow to protect it as proprietary (it made sense when I heard it explained, but now I feel like I'm missing an important detail to that), and then Valve had some talks with the HDMI Forum to convince them to let it go, presumably so they could have such support integrated into Linux for their hardware (Deck, Machine, Frame, etc).
So now the HDMI Forum just doesn't really care anymore I guess.
IIRC, a non-AMD contributor was working on adding HDMI 2.1 support to Linux's AMD driver, and then AMD managed to get the HDMI Forum to see that their position was moot and was making things very awkward for AMD.
Oh this makes sense! Obviously, it can't be illegal for a non-forum-member to add HDMI support; it's just an HDMI forum policy after all, not law. That's certainly the link I was missing for this to make sense.
presumably someone at Valve said "either you do this or we will pay an open source contributor 100k a year to reimplement every spec you make for the rest of time"
It's not like the policy ever made much sense to begin with. Hard to try to keep track of what a rationale might be enough to convince people who are already being unreasonable!
I've seen some people say that since HDMI 2.2 recently released the HDMI forum doesn't care as much about "protecting" 2.1 anymore.
If FOSS drivers support HDMI 2.2 when devices start shipping with HDMI 2.2 ports then this'll be disproven, but the HDMI forum held off on allowing AMD to implement 2.1 support in their Linux drivers for years and only relented when they released HDMI 2.2. Valve with the Gabecube probably didn't have much to do with it considering how low volume it is.
I just assumed they could tell that the world was getting close to just switching to DP if they kept making hdmi inconvenient, so the choice is be convenient or be irrelevant.
I don't mean that specifically the linux kernel mattered so much to them, but in general for consumers (product makers) all up and down the line.
what even is the alleged crime? If there was any protected IP they would be able to use the courts to sue. But that isn't happening, which suggests they know they don't have grounds. Am I missing something? Patents seem to be very much still grounds for lawsuits (even relatively baseless patents)
As a long time desktop and server user, why do I look at changelogs...
New filesystems and major changes to existing ones (large performance improvements, new features like adding better compression standards which I now know I can use, fixes to long-standing bugs which open new use cases).
New protocols like wireguard when it was new and exciting (still is now that I think of it). When that was mainlined, I knew I had some network rearchitecting to do.
New device drivers like amdgpu when it was massive news. Less so these days since most hardware is well supported, and I don't buy the absolute latest.
New kernel APIs which I can use in my programs — like landlock, when it was merged I knew it was time to improve sandboxing in some of my stuff.
Though LWN and kernel newbies do it best, already linked by others.
It's a well-formatted changelog of the most salient features of the release. Linus doesn't provide one apart from a list of added commits and merged branches, which isn't very useful unless you are a kernel developer.
=USB4STREAM for streaming data with USB4=
In addition to Thunderbolt networking, which was already supported, this release adds support for streaming data directly over a cable as well (USB4STREAM). It is very simple and basically just transfers raw packets from one host to another. The driver exposes /dev/tbstreamX devices on each side of the link that can be used to transfer data using regular filesystem operations such as read(2) and write(2):
host1 # cat /dev/tbstream0
host2 # echo hello > /dev/tbstream0
This can be useful in cases where network tooling is not available or just for existing applications like dd and cat that do not support sockets. This feature can be used at the same time with thunderbolt_net so they don't rule each other out, and it allows multiple streams to be created.
I think it is mostly "marketing" for Igalia's services. If you want to pay people to do "stuff" in Linux and adjacent projects like Mesa, Igalia seems very competent.
Exactly. My immediate reaction was: A violation of HN submission guidelines. Linux is project of many contributors / companies. For sure Igalia is a valuable contributor and I would not doubt to recommend them for serious Linux work, but this "article" is pure marketing.
We publish something about all the work Igalia has contributed with each release. Fwiw, we didn't submit this ourselves, but it's cool someone thought it was interesting enough to share. In context, the title is fine, but as shared I can definitely understand the confusion.
I'm using a recently released laptop, and Linux hardware support hasn't caught up yet. So I've been following the changelogs of recent kernel releases to understand the current state of support.
I'd say mostly sys admins type users that want to stay on top of things?
I install Linux on a wide variety of hardware for both work and personal reasons.
I don't read the detailed changelogs religiously but after years of paying some attention to them I can understand most of it and see where the industry is investing developer efforts, what cool new hardware is gaining support and what new security or performance features people are focusing on.
I look for wake from sleep fixes, nvidia the primary offender there (they have at least two bugs affecting it across several driver versions). But my intel wifi driver (BE200) also has an apparent hard lock on wake. Looks like this particular release doesn't have any fixes for me, though claude says its not the full change list.
Intel has a specific series of integrated network cards that softlock themselves and you can only fix by disconnecting power for 30s, reconnecting the power, then rebooting the PC twice.
Living through that made me wonder why people would ever swap out their Broadcom or Atheros WiFi card for an Intel one.
It's more than that: free and open support for HDMI 2.1 in general. Previously it was only proprietary implementations that didn't leak the spec which were allowed. AMD is just the only one that has really good open source drivers which can benefit from it, and didn't work around the problem by converting HDMI to DisplayPort in a proprietary on-board GPU chip.
I don't think there is much too it, it's a consulting firm. Would you hire a consulting firm with an empty tech blog portifolio? It's ticking a box, it's stating it's a consulting firm in tech, that it's target audience are companies with Linux machines etc...
Wouldn't have HN any other way. For better or worse, it's the best sorta proving grounds for writers. Pass the trial of fire and chances are you did something right.
I'd count my losses and move on happy that it hit front page.
Also amplifiers and capture devices generally only do HDMI. I haven't looked at projectors, lately, but HDMI is The Standard and a desktop computer display is the only exception.
> Okay, so ELI 5 why I would now use HDMI instead of DP?
I use DP too when I can. For example my main desktop is using DP.
But many devices have HDMI, not DP. For example I've got seven Raspberry Pi here: they all have HDMI and none has DP. There's, what, 100 million Raspberry Pi in the world by now? Don't know the number but: devices that do not support DP aren't rare.
I have been using HDMI to game on TV for most or all of my adult life. Mostly console gaming. A few years back I built a dedicated gaming rig, but I still use it with a controller from my couch connected to a TV.
So this is a win for couch gamers and open source in general.
I ran fiber optic HDMI from my office to my living room for 4K120 HDR+VRR couch gaming with no added latency or loss in quality. Works great. My LG C1 even supports FreeSync so VRR even works, but I have to route the HDMI directly to the TV. My receiver does not support FreeSync and only supports HDMI VRR. Presumably with this patch I will be able to run HDMI 2.1 VRR through my receiver and not have to treat my desktop as a special input.
It's just nice when technology is not hampered by nonsense.
I can tell you mine does: a Corsair XENEON 27QHD240. Or the somewhat popular and badly named Acer Nitro XV275K P5biipruzx
I'd say it's usual for displays released in the 2-5 years old range. HDMI 2.1 got somewhat quick monitor support for consoles, but search for "first DP2.1 monitor" gets results from 2024
Have you considered zram/swap? Since running it, i have never seen mem usage > 95%, even under absurd workloads and have never been oom-killed ever since.
I did try that back when I only had 32GB. They definitely helped (especially swap) but didn't completely eliminate the problem. I don't know how they do it but I've never had a single issue with OOMs on Windows or Mac. I know Windows doesn't overcommit and it has a proper ctrl-alt-del to let you actually control killing processes if you need to... I'm not sure what Mac does.
Every time I OOM on Windows my Laptop just becomes completely unresponsive excluding the hardware cursor (it is the only thing still movable, everything else is frozen in time), ctrl-alt-del does fuck all until the system becomes responsive again. The worst part is, this isn't limited to OOM, but generally when my disk is hammered with high I/O (e.g. caused by thrashing).
... look into earlyoom , its a pretty neat solution
just prescribe to it, all the stuff to never oom no matter what
I've never had to hard reboot anymore due to OOM
After setting up earlyoom to gracefully start shutting stuff down at a threshold before hitting oom, robustly works all the time even under intense ram + swap pressure + 100% cpu on all cores sorta workload
Linux is so very predictably stable from a desktop user perspective . It is an absolute pleasure to use , particularly when combined with TUI applications like yazi, lsd, fzf , neovim etc
Been using it since the early nineties .
At work I use Windows Server 2025 as my desktop ( the alternative would be the dreadful AI infested Windows 11).
While Windows Server 2025 is not dreadful , the Linux experience is so much nicer .
> DRM exposes an API that user-space programs can use to send commands and data to the GPU and perform operations such as configuring the mode setting of the display.
> User-space programs can use the DRM API to command the GPU to do hardware-accelerated 3D rendering and video decoding, as well as GPGPU computing.
The Direct Rendering Manager is useful for any Linux user who wants to display things on a monitor (unless you're happy with like 640x480 VESA, in which case maybe you don't need it)
It is funny with Linux (the kernel). From the outside it looks like almost nothing ever changes. You switch the computer on, it works away in the background to bring up the desktop and have everything talk. From that perspective it feels like nothing has changed in decades.
But then you look at the change log, and almost everything you see added sounds really useful to someone working on this stuff. 35 years of work on this and there is still so much to be done even if you most people using it never directly feel it.
Strongly disagree. Hardware support improved by leaps and bounds last decade and a half. Namespace isolation and cgroups driven containerization. eBPF appeared and gained adoption. Async I/O (and later just everything async) converged on io_uring. Btrfs is _stable_ - some people believed it would never be possible, ever. All kinds of scheduling were iterated and iterated over.
All these things are visible across the board, both from developers and end users side.
I have been using Linux on the desktop since the late 90's. Other than increase hardware support, I have not tangentially noticed/felt almost anything you have mentioned here. When I fire up the desktop and load Firefox, I do not feel or see the sandboxing, the scheduler, the file system or the I/O. It just works.
Yes, we interact with it all the time but it isn't something that the end user would notice and that is a great thing. If you are not directly poking around with that stuff, you shouldn't notice it.
That they are so seamless should be seen as a badge of honor.
I also have been using Linux for over 20 years and I agree that hardware support is the most noticable.
This year I have installed Linux Mint on computers from people who wanted to jump from Windows to Linux. Every time I booted via USB, installed Linux and 10 minutes later everything was working. This would not have happened 10 years ago. Back then there would always be an issue with a video driver, WiFi driver or audio driver.
Of course there are still hardware issues for some special deviced, but today Linux just runs fine for most people. And that's a great accomplishment.
The most fun I ever had with a computer was as a 17-year old, installing gentoo on a P4 machine, over dialup, stage 2 tarball, custom kernel.
I’ve seen people say that for 20 years
> This would not have happened 10 years ago.
Funny you say that, as I was installing Ubuntu on peoples laptops and it just worked, like 15 years ago... :-)
"I installed Ubuntu on peoples laptops and was extremely lucky they happened to have compatible hardware or didn't notice the stuff that didn't work and had a non-borked firmware". FTFY.
This was absolutely not everyones experience, including mine.
Only on a select combination of blessed components and BIOS/EFI not half-ass broken to the tune it only worked somewhat stable on Windows.
Anything with split iGPU/dGPU was a mess no matter what and where, WiFi that was not Intel was a mess, suspend to ram and suspend to disk were a coin flip if the system would wake up again or if it would decide to wake up on its own and overheat itself to panic.
I run Linux as my primary operating system and I have never even heard of eBPF! I'd be surprised if most end users have.
Seems like parent thread spoke from perspective of developer/tech savvy person.
But even from end-user perspective, maybe its hard to see change when you compare one year or so, but over somewhat longer period I definitely see improvements. Over say a decade or so, wifi/bluetooth/trackpad support has significantly improved. Over last 5 years, I think gaming has become huge and you can actually use Linux as gaming machine. In last 3 years or so, I think I've seen my 5k display using Thunderbolt work much better. Fractional scaling also works much nicer.
NVIDIA GPUs are not shitty to use on Linux, and AMD has been working great for slightly longer. UEFI/TPM works much better now as well, where I don't have to worry about booting in legacy mode when using GPU drivers.
I didn't have to think about any of these above. If I think harder, I can likely come up with many more.
[dead]
You may know more about what it powers. It has enabled much faster, higher quality, more stable security software on Linux. Many modern security tools are built on top of it. While end users might not know it by name, it has made Linux easier to support in the enterprise.
Well ya, it's more useful in observability usecases than desktop ones. But recently I've been using for a consumer desktop usecase. I made a file watcher with it. I'm sure you're familiar with the woes of using inotify/fanotify.
eBPF vfs hooks based watchers are really nice.
[edit] details
- you setup a single `mount` - this is the only part that requires cap_sys_admin - then you run a single daemon, this one only needs cap_bpf, and this runs once at startup and is its own self contained program - this daemon writes events into a ringbuf referred to by a file in that bpf mount - you then read that from your standard user program (this can be done by whichever application needs to consume the events)
the events are create, mkdir, delete. renames are automatically decomposed since we hook at the vfs layer. the hook itself is practically free to run and thus there is no question of a performance impact.
> Btrfs is _stable_ - some people believed it would never be possible, ever
Didn't know that. Is it worth looking into over ZFS?
For desktop I think it would be fine, though if you have the advanced needs of all the ZFS features I would stick with ZFS, but I had no problem.
The real problem I had was a low limit to save history. I install a lot of software, Cachy would run the snapshot every time (good), but it only kept like 5-10. And the one I needed to restore to was further back (I can't remember the exact issue), but that was more the defaults of how cachy set it up, and my use case than btrfs/snapper itself.
Whatever tool you use to snapshot btrfs, you may wish to have a custom number of saves if you install tons of software a lot...so you can avoid having too few backups available from the boot menu. I'm sure there was a way around it but I said f it and just reinstalled instead of trying to look up on my crap tablet how to get at the old files (assuming they were still there and just not accessible from the boot menu vs being completely erased).
I do not think it was btrfs that broke, to be clear.
This was like a week before the big hack on AUR, so I wonder if I got hit by a supply attack, but it was pretty shoddy, and broke my system, thus my reinstall saved me from (possibly) a lot more pain down the road.
btrfs is stable?? I used it a couple of years ago and really didn’t have a positive experience. I’ve found ZFS much more stable
The disk full scenario is greatly improved. Still buggy, but it is recoverable.
Raid 5 is still consider unsafe.
Don't run postgresql on it.
So..... it's "very stable, unless you do this or that" state.
It was the default on SUSE / OpenSUSE for years. Haven't used them for a while, not sure whether it still is. No problems with basic functionality including snapshots. I believe some RAID configurations were experimental for long time.
Been using it for a year or so on my main box. Never had any kind of issue. Snapshots saved my ass a couple of times. All in all it’s been pretty boring (in a good way)
Yes. I have used it for 15 years without any problems. Just keep track of snapshots and consumed snapshot space. Besides that, works great!
Maybe. I think there's still no chance of repair when it blows up. Only FS since Reiser and early XFS that I've completely lost data to on healthy hardware.
Whichever one of my test machines is using btrfs is doing fine, though, so on a subjective level that one's been stable enough.
It's basically fine as long as you read the release notes carefully and don't do anything the release notes say not to do (e.g. RAID5, let the disk fill up). It still has some sharp edges, but they're all documented.
Yes, even Bcachefs was on and off through versions
I imagine for some end users a lot of this stuff is happily invisible, just like you say.
But for me, a few kinds of changes have been highly visible as a user:
As someone who uses macOS at work for the last several years, I really miss the Linux kernel when it comes to filesystems, and feel like macOS updates often involve very few changes that I care about or even consider operating system change at all. Most macOS updates are just trumped up application updates.In that way I've kind of felt the opposite somehow, like OS updates on proprietary operating systems feel thin and empty.
Isn't this the case with most well established software?
There is still a lot of work done on the background, perhaps even more than ever.
But unless you are a power user eagerly waiting for one of the things currently worked on, you wouldn't notice or care unless the interface changes (which usually only happens on visual interfaces).
Software with user interfaces will typically refresh the UI from time to time to remind you that its developers are still hip and with it.
But there isn't a whole lot a kernel can do in the same way. Hardware support is about as user-visible as things get for Linux.
I've been daily driving Linux since ubuntu 12.04. Things have improved a lot. It used to be the best OS for (a certain class of) programmers. It's now the OS I installed on my 7 year old son's computer. Things like connecting Bluetooth peripherals, multiple audio devices, multiple screens (with multiple DPI) are a not just possible, but easy now, and robust.
i remember back in the day, when i built my project, using 100% cpu, the whole system became lagging and choppy and you really can't do anything anymore.
Now it behaves more like Windows/Mac, even when 100%, my youtube/spotify tab can still run smoothly.
(Of course i think there will be hit in some areas, like my code would be built slower, there is no free meal).
Maybe it's distro packaging, but I suspect it's the kernel, too: In the 20 years I have dipped my toes into Linux, there is no better time to be a desktop user than today. With CachyOS and modern kernel, I have zero audio issues, zero GPU issues, (thanks to Wine/Steam Proton) no gaming issues, and much more rare instances of wifi or sleep problems. The idea of having to install hacky wifi drivers from USB stick are over.
Actually, i was waiting for 7.2 in order not to "install hacky wifi drivers from usb stick" XD, because mt7927 was not supported, so neither wifi nor bluetooth have been working out of the box
Not really. I never had this less issues on Linux with hardware. I remember 15 years back when I wanted to install it on a Desktop/Laptop, many things simply didn't work. Now you can install Linux on a self-assembled box and 99% of things will work out of the box.
> From the outside it looks like almost nothing ever changes
You should try to compile a custom kernel for your machine and strip the unneeded things out. Good luck.
Linux has diabetes.
Things that are unneeded, ie. unmaintained, eventually get purged. It happens quite regularly. For nearly every kconfig option, someone needs it enough to maintain it. But I agree, `make menuconfig' is overwhelming.
Does anyone understand how HDMI 2.1 support is now no problem? I remember that 2.1 support in AMD's open source driver was blocked by the HDMI forum, but I haven't heard any news about them unblocking it. What changed?
It doesn't give any specifics about what changed on the HDMI Forum side, but this article [0] from June indicates that Valve worked closely with AMD to get it pushed through prior to the release of the Steam Machine (which supports HDMI 2.1 VRR).
[0] https://www.digitalfoundry.net/news/2026/06/valves-hdmi-2-1-...
As far as I can recall: some internal implementation documentation leaked which made the legal position tougher for them somehow to protect it as proprietary (it made sense when I heard it explained, but now I feel like I'm missing an important detail to that), and then Valve had some talks with the HDMI Forum to convince them to let it go, presumably so they could have such support integrated into Linux for their hardware (Deck, Machine, Frame, etc).
So now the HDMI Forum just doesn't really care anymore I guess.
IIRC, a non-AMD contributor was working on adding HDMI 2.1 support to Linux's AMD driver, and then AMD managed to get the HDMI Forum to see that their position was moot and was making things very awkward for AMD.
Oh this makes sense! Obviously, it can't be illegal for a non-forum-member to add HDMI support; it's just an HDMI forum policy after all, not law. That's certainly the link I was missing for this to make sense.
presumably someone at Valve said "either you do this or we will pay an open source contributor 100k a year to reimplement every spec you make for the rest of time"
I actually like that reality better
It's not like the policy ever made much sense to begin with. Hard to try to keep track of what a rationale might be enough to convince people who are already being unreasonable!
I've seen some people say that since HDMI 2.2 recently released the HDMI forum doesn't care as much about "protecting" 2.1 anymore.
If FOSS drivers support HDMI 2.2 when devices start shipping with HDMI 2.2 ports then this'll be disproven, but the HDMI forum held off on allowing AMD to implement 2.1 support in their Linux drivers for years and only relented when they released HDMI 2.2. Valve with the Gabecube probably didn't have much to do with it considering how low volume it is.
8K still not working properly
https://www.phoronix.com/news/Linux-7.3-More-AMDGPU
I hope they fix it so both of the 8k monitor users can get their displays working soon.
They can already, with DisplayPort.
I just assumed they could tell that the world was getting close to just switching to DP if they kept making hdmi inconvenient, so the choice is be convenient or be irrelevant.
I don't mean that specifically the linux kernel mattered so much to them, but in general for consumers (product makers) all up and down the line.
[flagged]
what even is the alleged crime? If there was any protected IP they would be able to use the courts to sue. But that isn't happening, which suggests they know they don't have grounds. Am I missing something? Patents seem to be very much still grounds for lawsuits (even relatively baseless patents)
Only if an AI does it, so make sure the vibe code the driver
Happily upvoting you for agreeing on principal with me. I shall likely join the flogging as a result.
I’m sure it will hurt my 11k internet point account.
“Have I now become your enemy by telling you the truth?”
-- Galatians 4:16
Who is the main audience for this type of content? Genuinely curious because I am not a serious Linux user.
What type of info do you get out of this for those who read this? So i can glance your insight?
As a long time desktop and server user, why do I look at changelogs...
New filesystems and major changes to existing ones (large performance improvements, new features like adding better compression standards which I now know I can use, fixes to long-standing bugs which open new use cases).
New protocols like wireguard when it was new and exciting (still is now that I think of it). When that was mainlined, I knew I had some network rearchitecting to do.
New device drivers like amdgpu when it was massive news. Less so these days since most hardware is well supported, and I don't buy the absolute latest.
New kernel APIs which I can use in my programs — like landlock, when it was merged I knew it was time to improve sandboxing in some of my stuff.
Though LWN and kernel newbies do it best, already linked by others.
It's a well-formatted changelog of the most salient features of the release. Linus doesn't provide one apart from a list of added commits and merged branches, which isn't very useful unless you are a kernel developer.
This isn't quite right.
This is a well-formatted changelog of the changes that Igalia has contributed to the release, AFAICT.
KernelNewbies publishes a much more general summary of changes (it is much longer than this one).
https://kernelnewbies.org/LinuxChanges
https://kernelnewbies.org/LinuxVersions
What great resource thank you !
This new thing is pretty cool :
It only covers changes made by Igalia devs. For a more general changelog I’d recommend the LWN articles:
https://lwn.net/Articles/1078068/ https://lwn.net/Articles/1078539/
ty for the context. I was not aware of the Linux dev ecosystem.
I think it is mostly "marketing" for Igalia's services. If you want to pay people to do "stuff" in Linux and adjacent projects like Mesa, Igalia seems very competent.
Exactly. My immediate reaction was: A violation of HN submission guidelines. Linux is project of many contributors / companies. For sure Igalia is a valuable contributor and I would not doubt to recommend them for serious Linux work, but this "article" is pure marketing.
https://lwn.net/Articles/1088991/ has all the relevant links. Probably somewhwere there is also a more detailed "neutral" article.
We publish something about all the work Igalia has contributed with each release. Fwiw, we didn't submit this ourselves, but it's cool someone thought it was interesting enough to share. In context, the title is fine, but as shared I can definitely understand the confusion.
> Who is the main audience for this type of content?
I make technology decisions at a large embedded shop that makes cars.
Content like this tells me Igalia has serious programmers and is relevant upstream. The memory sticks.
It's also informative, which makes me feel good.
They're a consultancy. It's good for business.
I'm using a recently released laptop, and Linux hardware support hasn't caught up yet. So I've been following the changelogs of recent kernel releases to understand the current state of support.
I'd say mostly sys admins type users that want to stay on top of things?
I install Linux on a wide variety of hardware for both work and personal reasons.
I don't read the detailed changelogs religiously but after years of paying some attention to them I can understand most of it and see where the industry is investing developer efforts, what cool new hardware is gaining support and what new security or performance features people are focusing on.
hi audience, we know about this stuff you use on a daily basis, you may be interested in our executive summary of the latest release of the stuff
I look for wake from sleep fixes, nvidia the primary offender there (they have at least two bugs affecting it across several driver versions). But my intel wifi driver (BE200) also has an apparent hard lock on wake. Looks like this particular release doesn't have any fixes for me, though claude says its not the full change list.
Intel has a specific series of integrated network cards that softlock themselves and you can only fix by disconnecting power for 30s, reconnecting the power, then rebooting the PC twice.
Living through that made me wonder why people would ever swap out their Broadcom or Atheros WiFi card for an Intel one.
It happens on MediaTek wifi adapter cards as well (and the AMD ones that are just rebadged MediaTek).
It's the sort of content I expect to see on HackerNews, rather than popular science/technology-type journalism.
I for one am excited about HDMI 2.1 support on AMD.
It's more than that: free and open support for HDMI 2.1 in general. Previously it was only proprietary implementations that didn't leak the spec which were allowed. AMD is just the only one that has really good open source drivers which can benefit from it, and didn't work around the problem by converting HDMI to DisplayPort in a proprietary on-board GPU chip.
Intel only used the converters for 1st gen Arc dGPUs. Battlemage has native HDMI 2.1 and also ran into the same issues as AMD.
Wasn’t it the same with m1 macs, it also had a converter chip and did therefore only support hdmi 2.0.
OK, now this is what I wanted to know as well. I was not aware that changelogs show all this inf.
ty for sharing
It’s intrinsically interesting
You only come here for the AI hype blogs?
I don't think there is much too it, it's a consulting firm. Would you hire a consulting firm with an empty tech blog portifolio? It's ticking a box, it's stating it's a consulting firm in tech, that it's target audience are companies with Linux machines etc...
ty. so very narrow audicence it seems.
Per @sph, as Linus provide no changelog, seems like the firm is putting it together for others
Wouldn't have HN any other way. For better or worse, it's the best sorta proving grounds for writers. Pass the trial of fire and chances are you did something right.
I'd count my losses and move on happy that it hit front page.
It's not a Linux 7.2 changelog, it's a list of things Igalia worked on that's included in this release.
Okay, so ELI 5 why I would now use HDMI instead of DP?
All my monitors support DP, my GPU has more ports for DP ...
I am genuinely interested as I've never even considered using HDMI for my Desktop.
If you’re using a TV. They rarely have DP ports.
Also amplifiers and capture devices generally only do HDMI. I haven't looked at projectors, lately, but HDMI is The Standard and a desktop computer display is the only exception.
> Okay, so ELI 5 why I would now use HDMI instead of DP?
I use DP too when I can. For example my main desktop is using DP.
But many devices have HDMI, not DP. For example I've got seven Raspberry Pi here: they all have HDMI and none has DP. There's, what, 100 million Raspberry Pi in the world by now? Don't know the number but: devices that do not support DP aren't rare.
And Linux runs on much more than desktop PCs.
First off: I am aware that Linux runs on much more than desktop PCs.
I am using it exclusively and since many decades.
I was specifically asking for desktop usage though, because that's what I am mostly working with.
If that point didn't come across, I am sorry. English isn't my first language.
HDMI is a king on consumer TVs. Linux (in form of SteamOS) is becoming a major console gaming platform thanks to Valve.
I have been using HDMI to game on TV for most or all of my adult life. Mostly console gaming. A few years back I built a dedicated gaming rig, but I still use it with a controller from my couch connected to a TV.
So this is a win for couch gamers and open source in general.
> Okay, so ELI 5 why I would now use HDMI instead of DP?
For A/V stuff: CEC and (e)ARC.
* CEC: control one device, and downstream items receive instructions so you're (ideally) not fiddling with multiple remotes.
* ARC: the TV/display can decode audio/video and pass the audio to a sound bar or other audio system.
HDMI 2.1 carries more bandwidth than DisplayPort 1.4. Plenty of GPU/monitors have that port arrangement, as recent as Nvidia's RTX 40 Series.
Yeah but this is for amdgpu and they had DP2.1 support for a while now.
I ran fiber optic HDMI from my office to my living room for 4K120 HDR+VRR couch gaming with no added latency or loss in quality. Works great. My LG C1 even supports FreeSync so VRR even works, but I have to route the HDMI directly to the TV. My receiver does not support FreeSync and only supports HDMI VRR. Presumably with this patch I will be able to run HDMI 2.1 VRR through my receiver and not have to treat my desktop as a special input.
It's just nice when technology is not hampered by nonsense.
AMD RX 6700 XT from 2021 only has DP1.4 (but HDMI 2.1).
Dell U5226KW from 2026 only has DP1.4 (but HDMI 2.1).
I don't think I have seen anything above DP1.4 in the wild for now.
Also interesting is the Steam Machine with its AMD RX 7000 series GPU only has DP1.4 even though those GPUs should be able to do DP2.1
That doesn't help when the monitors don't have DP2.1 support.
I havent seen a monitor (besides tvs) that has HDMI2.1 but not DP2.1
Plenty of high refresh gaming monitors out there that use HDMI 2.1 or DP1.4 with DSC.
DSC is just the "card up the sleeve" for DP1.4, you can push it all the way to 4K 240hz HDR. So there is little incentive to switch to DP2.1 for OEMs.
I can tell you mine does: a Corsair XENEON 27QHD240. Or the somewhat popular and badly named Acer Nitro XV275K P5biipruzx
I'd say it's usual for displays released in the 2-5 years old range. HDMI 2.1 got somewhat quick monitor support for consoles, but search for "first DP2.1 monitor" gets results from 2024
And that parenthetical is exactly why people care.
TVs.
lots of linux based boxes are used with TVs.
Because your motherboard only supports HDMI. The last three I used had HDMI, and sometimes VGA.
What’s DP? My desktop and laptop both have HDMI.
Display port. Most monitors support it.
Unfortunately not if you're running a mac
Macs definitely can use DP Alt Mode over USB-C ports, so if you get a USB-C -> DP cable it's straight up DP to the monitor.
Or you can just plug things into HDMI port without dongles.
Or you can just plug monitors into usb-c ports using usb-c cables without any special hdmi cables.
Which part of "because the device doesn't have a DP port" is not enough for you?
You can't picture such devices or what's the holdup?
I asked specifically for a desktop.
> I am genuinely interested as I've never even considered using HDMI for my Desktop.
Sorry if that didn't come across. English isn't my first language
I clicked this one with no hype. I am now exited to update the kernel of my raspberry pi 4.
What about this is more interesting than the LWN coverage?
I just want them to make memory management sane. Is that too much to ask? OOM shouldn't cause a hard reboot.
Ok actually I "solved" this by upgrading to 128GB of RAM (before the rampocalypse), but still they should fix it.
Have you considered zram/swap? Since running it, i have never seen mem usage > 95%, even under absurd workloads and have never been oom-killed ever since.
Don't you get slower performance in some cases?
Yes, but for many (not all) it's better than OOM killing your processes.
I did try that back when I only had 32GB. They definitely helped (especially swap) but didn't completely eliminate the problem. I don't know how they do it but I've never had a single issue with OOMs on Windows or Mac. I know Windows doesn't overcommit and it has a proper ctrl-alt-del to let you actually control killing processes if you need to... I'm not sure what Mac does.
Every time I OOM on Windows my Laptop just becomes completely unresponsive excluding the hardware cursor (it is the only thing still movable, everything else is frozen in time), ctrl-alt-del does fuck all until the system becomes responsive again. The worst part is, this isn't limited to OOM, but generally when my disk is hammered with high I/O (e.g. caused by thrashing).
... look into earlyoom , its a pretty neat solution just prescribe to it, all the stuff to never oom no matter what I've never had to hard reboot anymore due to OOM
After setting up earlyoom to gracefully start shutting stuff down at a threshold before hitting oom, robustly works all the time even under intense ram + swap pressure + 100% cpu on all cores sorta workload
https://man.archlinux.org/man/earlyoom.1.en
[dead]
Linux is so very predictably stable from a desktop user perspective . It is an absolute pleasure to use , particularly when combined with TUI applications like yazi, lsd, fzf , neovim etc
Been using it since the early nineties .
At work I use Windows Server 2025 as my desktop ( the alternative would be the dreadful AI infested Windows 11).
While Windows Server 2025 is not dreadful , the Linux experience is so much nicer .
[dead]
[flagged]
[flagged]
Why is DRM being implemented in the Kernel; What is the profile for an end-user who wants this in their FOSS?
People who want to play video games, or watch videos using hardware decoding, or train machine learning models? DRM here means https://en.wikipedia.org/wiki/Direct_Rendering_Manager
> DRM exposes an API that user-space programs can use to send commands and data to the GPU and perform operations such as configuring the mode setting of the display.
> User-space programs can use the DRM API to command the GPU to do hardware-accelerated 3D rendering and video decoding, as well as GPGPU computing.
Here it doesn’t mean “digital rights management”, but “Direct Rendering Manager”, a component of the kernel’s graphics stack: https://docs.kernel.org/gpu/?utm_source=chatgpt.com
This is DRM as in Direct Rendering Manager, which is the kernel interface to GPU hardware; not Digital Rights Management.
Named long before Digital Rights Management was a thing (or at least the 3 letter acronym existed for it).
The Direct Rendering Manager is useful for any Linux user who wants to display things on a monitor (unless you're happy with like 640x480 VESA, in which case maybe you don't need it)
DRM refers to Direct Rendering Manager in this context.
Different DRM, the one in the kernel is Direct Rendering Manager, which is basically managing access to the GPU.
DRM = Direct Rendering Manager. GPU scheduling.
its the Direct Rendering Manager, not the other DRM