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.
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).
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.
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.
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.
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.
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!
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.
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.
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
> 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.
> 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)
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.
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).
All these things are visible across the board, both from developers and end users side.
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.
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.
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 should try to compile a custom kernel for your machine and strip the unneeded things out. Good luck.
Linux has diabetes.
[0] https://www.digitalfoundry.net/news/2026/06/valves-hdmi-2-1-...
So now the HDMI Forum just doesn't really care anymore I guess.
https://www.phoronix.com/news/Linux-7.3-More-AMDGPU
I’m sure it will hurt my 11k internet point account.
-- Galatians 4:16
What type of info do you get out of this for those who read this? So i can glance your insight?
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.
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.
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
This new thing is pretty cool :
https://lwn.net/Articles/1078068/ https://lwn.net/Articles/1078539/
https://lwn.net/Articles/1088991/ has all the relevant links. Probably somewhwere there is also a more detailed "neutral" article.
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.
Living through that made me wonder why people would ever swap out their Broadcom or Atheros WiFi card for an Intel one.
ty for sharing
Per @sph, as Linus provide no changelog, seems like the firm is putting it together for others
I'd count my losses and move on happy that it hit front page.
Ok actually I "solved" this by upgrading to 128GB of RAM (before the rampocalypse), but still they should fix it.
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.
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.
It's just nice when technology is not hampered by nonsense.
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'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
TVs.
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.
You can't picture such devices or what's the holdup?
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.
So this is a win for couch gamers and open source in general.
> 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.