Linux 4.17 Will Be Another Exciting Kernel Cycle

[phoronix]

Phoronix: Linux 4.17 Will Be Another Exciting Kernel Cycle

While the Linux 4.16 kernel release is still three weeks or so away, the Linux 4.17 kernel is already shaping up to be another exciting cycle...

http://www.phoronix.com/scan.php?page=news_item&px=Linux-4.17-Will-Be-Exciting

[hax0r]

4.17 would be nice if it placed more respect toward Desktop users, they could get rid of old rotten CFS scheduler, it sucks at latency and iowait when any storage device in the system gets hit with 100% I/O load, desktop users should have a choice and not be stuck with CFS that's optimized for workloads with systems having >1024 CPUs, mainlined MuQSS would be nice. 4.17 would be nice it there was decent filesystem in Linux, sadly there's plenty of old rotten filesystems, XFS is nice (thanks SGI) but not for desktop users, BTRFS is unstable mess, it would be nice to have mainlined ZFS that's on par with FreeBSDs superb ZFS implementation, 'beadm' and boot environments would so nice. 4.17 would be nice if TCP/IP performance matched that of FreeBSD, Linux TCP/IP would be nice if it could withstand DDoS attacks or high throughput., Netflix knows better. 4.17 would be nice if kernel developers actually benchmarked their patches and remedied performance regressions at least every major release.

[starshipeleven]

4.17 would be nice if it placed more respect toward Desktop users, they could get rid of old rotten CFS scheduler, it sucks at latency and iowait when any storage device in the system gets hit with 100% I/O load, desktop users should have a choice and not be stuck with CFS that's optimized for workloads with systems having >1024 CPUs, mainlined MuQSS would be nice. 4.17 would be nice it there was decent filesystem in Linux, sadly there's plenty of old rotten filesystems, XFS is nice (thanks SGI) but not for desktop users, BTRFS is unstable mess, it would be nice to have mainlined ZFS that's on par with FreeBSDs superb ZFS implementation, 'beadm' and boot environments would so nice. 4.17 would be nice if TCP/IP performance matched that of FreeBSD, Linux TCP/IP would be nice if it could withstand DDoS attacks or high throughput., Netflix knows better. 4.17 would be nice if kernel developers actually benchmarked their patches and remedied performance regressions at least every major release.

Service notification: Your Return Key is broken.

[Djhg2000]

Actually Allwinner H3/H5 SoCs are also scheduled to get HDMI support with 4.17: https://linux-sunxi.org/Linux_mainli...anned_for_4.17

[sa666666]

4.17 would be nice if it placed more respect toward Desktop users, they could get rid of old rotten CFS scheduler, it sucks at latency and iowait when any storage device in the system gets hit with 100% I/O load, desktop users should have a choice and not be stuck with CFS that's optimized for workloads with systems having >1024 CPUs, mainlined MuQSS would be nice. 4.17 would be nice it there was decent filesystem in Linux, sadly there's plenty of old rotten filesystems, XFS is nice (thanks SGI) but not for desktop users, BTRFS is unstable mess, it would be nice to have mainlined ZFS that's on par with FreeBSDs superb ZFS implementation, 'beadm' and boot environments would so nice. 4.17 would be nice if TCP/IP performance matched that of FreeBSD, Linux TCP/IP would be nice if it could withstand DDoS attacks or high throughput., Netflix knows better. 4.17 would be nice if kernel developers actually benchmarked their patches and remedied performance regressions at least every major release.

Why don't you just use FreeBSD, and stop wasting energy complaining about things in Linux if you hate it that much?

Also, as another user pointed out, the return key is your friend; use it.

[Vistaus]

Why don't you just use FreeBSD, and stop wasting energy complaining about things in Linux if you hate it that much?

Also, as another user pointed out, the return key is your friend; use it.

Or Solus. They tweaked the kernel for desktop users recently.