Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Linux kernel 6.14 is a big leap forward in performance and Windows compatibility

zdnet.com

Post reply on HN
Print view View original

31–40 of 135 posts

Re: Linux kernel 6.14 is a big leap forward in performance and Windows compatibility

#31

Post by shmerl »

mhurron wrote: post #6
mouse_ wrote: post #2

From a YouTube comment: > please don't hype NTSYNC. yes it has better compatibility than ESYNC and FSYNC, yes it's marginally faster in selected title. The phoronix article reports benchmarks with WINESYNC vs NTSYNC, the gains are there only if you were not already running FSYNC (on by default in most titles running under proton). > By overhyping features they ultimately end up underdelivering because people expect i…

If you've been anywhere around linux gaming circles for the past 6-8ish months, you've seen nothing but people not understanding the benchmarks hype NTSYNC. Everyone trying to set appropriate expectations is pretty quickly silenced in whatever manner the platform allows (downvote, etc)

Must be some wrong circles. From what I've seen, it's mostly well known that ntsync is comparable to existing solutions like esync / fsync.

Re: Linux kernel 6.14 is a big leap forward in performance and Windows compatibility

#32

Post by lenerdenator »

TheCleric wrote: post #4

So it's early Monday morning (well - early for me, I'm not really a morning person), and I'd love to have some good excuse for why I didn't do the 6.14 release yesterday on my regular Sunday afternoon release schedule. I'd like to say that some important last-minute thing came up and delayed things. But no. It's just pure incompetence. Because absolutely nothing last-minute happened yesterday, and I was just clearing…

He really has grown as a person. Ten years ago, there would have been no mention of it unless someone called it out, and the response to that would have boiled down to "f#ck off, it's free".

... Which also would have been relatable, and valid if we're being honest.

Re: Linux kernel 6.14 is a big leap forward in performance and Windows compatibility

#33

Post by dayjah »

andy_xor_andrew wrote: post #20

I'm super excited for this! But it makes me curious. What's the process like for getting such a low-level primitive added to Linux? Especially a low-level primitive that 1) exist basically just to emulate behavior of a completely different kernel, and 2) is only needed for a subset of users to play games? I'm not complaining, just curious. I would assume a patch to add the above would be heavily scrutinized. Is it be…

I don’t know if this answers your question directly: but the article points out that this is a module, not a primitive that’s in the kernel itself.

The idea here is that wine/proton can interact with this module directly if it’s loaded, or continue to use their wrappers around existing fsync if not.

I expect that means a little less scrutiny since if it segs the user can choose to not load it and achieve a stable OS.

Re: Linux kernel 6.14 is a big leap forward in performance and Windows compatibility

#34

Post by tdb7893 »

Inityx wrote: post #9
jmclnx wrote: post #7

>This driver is designed to emulate Windows NT synchronization primitives To me, anything that emulates windows primitives is a bad thing. Again these changes being pushed by Microsoft make me glad the BSDs exist. Lets hope the BSDs can continue staying independent of Large Corporations.

Well, one of the biggest drivers of consumer Linux adoption in recent years has been the Steam Deck, and compatibility with Steam games. The article explicitly calls out how this improves performance for this use case. I think your opinion is an unpopular one.

Yeah, I know multiple people running Linux at home and Windows compatibility us a critical feature for them. For the people I know not in tech about half would switch if Linux got up to par in:

1. Ability to play any video game

2. Compatibility with hardware (e.g. one of my friends is booting windows temporarily as his graphics card has an issue on Linux)

3. Usability/looks. This is subjective but for people used to Windows for the last couple decades people often find Linux more difficult to use and honestly uglier (Linux is customizable but most people I know are not confident enough to go far from the defaults).

This is just my friends, some of whom are engineers and some are normal people when it comes to technical ability, but currently a lot of them want to like Linux but have specific requirements holding them back.

Re: Linux kernel 6.14 is a big leap forward in performance and Windows compatibility

#35

Post by caycep »

lenerdenator wrote: post #32
TheCleric wrote: post #4

So it's early Monday morning (well - early for me, I'm not really a morning person), and I'd love to have some good excuse for why I didn't do the 6.14 release yesterday on my regular Sunday afternoon release schedule. I'd like to say that some important last-minute thing came up and delayed things. But no. It's just pure incompetence. Because absolutely nothing last-minute happened yesterday, and I was just clearing…

He really has grown as a person. Ten years ago, there would have been no mention of it unless someone called it out, and the response to that would have boiled down to "f#ck off, it's free". ... Which also would have been relatable, and valid if we're being honest.

I assume he takes advice/feedback i.e. the famous ESR post and the concerns from that contributor from intel and others to heart. Or at least has forced himself to learn how to do so.

Re: Linux kernel 6.14 is a big leap forward in performance and Windows compatibility

#36

Post by kjrfghslkdjfl »

jchw wrote: post #30
cube00 wrote: post #17

Earlier quoted context omitted.

> But no. It's just pure incompetence. While I don't agree with how he conducts himself on the mailing lists; I respect he uses the same language when talking about himself.

You know, though, his conduct nowadays isn't really that bad; I don't think I'd be bothered by it in a workplace environment. He makes strong assertions, but not mean-spirited, and I think Linus has a lot of reasons to feel highly confident when he makes strong assertions when it comes to Linux. Of course, there are some moments, but almost everybody has their moments. When push comes to shove, Linus seems to handle…

[dead]

Re: Linux kernel 6.14 is a big leap forward in performance and Windows compatibility

#37

Post by ASalazarMX »

cube00 wrote: post #17
TheCleric wrote: post #4

So it's early Monday morning (well - early for me, I'm not really a morning person), and I'd love to have some good excuse for why I didn't do the 6.14 release yesterday on my regular Sunday afternoon release schedule. I'd like to say that some important last-minute thing came up and delayed things. But no. It's just pure incompetence. Because absolutely nothing last-minute happened yesterday, and I was just clearing…

> But no. It's just pure incompetence. While I don't agree with how he conducts himself on the mailing lists; I respect he uses the same language when talking about himself.

Does he willingly take that treatment from others? I don't really know, but that would be the actual test of congruence.

Re: Linux kernel 6.14 is a big leap forward in performance and Windows compatibility

#38

Post by nialv7 »

andy_xor_andrew wrote: post #20

I'm super excited for this! But it makes me curious. What's the process like for getting such a low-level primitive added to Linux? Especially a low-level primitive that 1) exist basically just to emulate behavior of a completely different kernel, and 2) is only needed for a subset of users to play games? I'm not complaining, just curious. I would assume a patch to add the above would be heavily scrutinized. Is it be…

Technically this is not a low-level primitive. NTSYNC isn't added as a, say, system call like futex2[0] was. It's instead added as a character devivce - which you can build your kernel without, and functions via ioctls.

[0]: https://docs.kernel.org/userspace-api/futex2.html

Re: Linux kernel 6.14 is a big leap forward in performance and Windows compatibility

#39

Post by ChocolateGod »

velcrovan wrote: post #25
jmclnx wrote: post #7

>This driver is designed to emulate Windows NT synchronization primitives To me, anything that emulates windows primitives is a bad thing. Again these changes being pushed by Microsoft make me glad the BSDs exist. Lets hope the BSDs can continue staying independent of Large Corporations.

You think Microsoft wants Windows games to perform better on Linux?

It's not as if they're the main stakeholder of Windows games on Windows anyway.

Re: Linux kernel 6.14 is a big leap forward in performance and Windows compatibility

#40

Post by KennyBlanken »

cube00 wrote: post #17
TheCleric wrote: post #4

So it's early Monday morning (well - early for me, I'm not really a morning person), and I'd love to have some good excuse for why I didn't do the 6.14 release yesterday on my regular Sunday afternoon release schedule. I'd like to say that some important last-minute thing came up and delayed things. But no. It's just pure incompetence. Because absolutely nothing last-minute happened yesterday, and I was just clearing…

> But no. It's just pure incompetence. While I don't agree with how he conducts himself on the mailing lists; I respect he uses the same language when talking about himself.

One does not get a pass on being rude and hurtful to others by speaking about themselves negatively.

The fundamental thing that people never seem to understand in the "Linus is being an asshole" vs "I love Linus, he tells it like he is and to manage a project you have to be willing to do that!" debate is that it is completely unnecessary to speak at all about someone's person on a mailing list like lkml.

"This code sucks!" - no. "This code does not meet the linux kernel's coding conventions, has race conditions, and this algorithm could likely be implemented in a far more performant way."

"You're just WRONG" - no. "The position that you're being forced to learn a new programming language is not correct. You have not been asked to write or even understand Rust."

I think the people who think 'you have to be that way to manage/lead' have never been under a good manager or leader and I daresay that a lot of people in technical supervisory and management positions unfortunately grew up respecting Linus and thinking his behavior was how you "get things done", not realizing that Linus's behavior almost certainly kept a lot of people from risking contributing and if anything has hurt the Linux kernel project - and kept people from feeling they could safely point out problems in someone's code because

T'so is a great example; a huge ego. As a result, people simply have gone elsewhere to work on filesystems. The ext filesystem(s) were not terribly reliable or tolerant of real-world situations, is still not very performant, and has evolved at a glacial pace consistently lagging behind its peers.

Post reply on HN