Another Look At The Bcachefs Performance on Linux 6.7

[reba]

Don't forget Devuan! The Appimage author's Wayland hate thread in github is also quite active: https://gist.github.com/probonopd/9f...5e3a9f2d1f2277

Just another rambling madman. And then I read "Edit: When I wrote the above, I didn't really realize what Wayland even was" and I closed the window...

[vermaden]

Why no ZFS also included in the tests?

Especially knowing that the tests were made on Ubuntu where ZFS is available ...

[darkbasic]

You should definitely add zfs 2.2 to the mix.

[Berniyh]

Just another rambling madman. And then I read "Edit: When I wrote the above, I didn't really realize what Wayland even was" and I closed the window...

Oh dear, for that kind of person and blog post you really need not a double facepalm, but a quadruple facepalm. Unbelievable …

Just when you look at his list "Wayland is broken by design": Wayland is broken by design

• A crash in the window manager takes down all running applications
• You cannot run applications as root
• You cannot do a lot of things that you can do in Xorg by design
• There is not one /usr/bin/wayland display server application that is desktop environment agnostic and is used by everyone (unlike with Xorg)
• It offloads a lot of work to each and every window manager. As a result, the same basic features get implemented differently in different window managers, with different behaviors and bugs - so what works on desktop environment A does not necessarily work in desktop environment B (e.g., often you hear that something "works in Wayland", even though it only really works on Gnome and KDE, not in all Wayland implementations). This summarizes it very well: https://gitlab.freedesktop.org/wayla...d/-/issues/233

The second point he already acknowledged is just wrong and the first is not a "broken by design", especially not more than X.org. It's just something that was not yet implemented, but is now being implemented by e.g. KDE, where starting with Plasma 6, everything should survive a compositor crash.

The third point really is not even an argument. Saying A is broken by design because B can do those things by design is just nonsense. List actual things that you think should be included instead of making over generic claims. Then there might be a discussion about whether or not your right, but like this it really is just nonsense.

For the fourth point I really don't see why this would be a "broken by design" thing. And it actually goes against his desire to to keep SysVInit alive, because everybody else (ok, except Gentoo folk) use systemd. You know that one init system that is used by everyone.
In the end, what he likely means here is the same thing as the fifth point, but he just listed it in a nonsense way as an additional point.
For that fifth point, you can indeed make an argument, but it's not a "broken by design" choice, since even with the Wayland protocol, you could in principle implement a desktop-independent thing that connects everything together like with Xorg. It's a decision that was made by the people that implement the Wayland protocol that they want to have multiple implementations of it.


In the end, from that blog post it's very clear that this Devuan person really lacks understanding of what he complains about so severely that you can just advise everybody to stay away from Devuan.

[LtdJorge]

For the fourth point I really don't see why this would be a "broken by design" thing. And it actually goes against his desire to to keep SysVInit alive, because everybody else (ok, except Gentoo folk) use systemd.

Hey, I am on Gentoo and use Systemd

[varikonniemi]

The results seem quite interesting, from clearly winning to having 10% of the performance of the winner. One could say it's looking good, but with much optimization still left to do.

[cj.wijtmans]

Hey, I am on Gentoo and use Systemd

unfortunately I also use systemd on gentoo. I just dont like sysvinit bash scripts although its more flexible, readable and configurable i just dont have enough experience with bash script to make a proper init script for the life of me. I wish systemd was less of a monolithic beast and more of a basic init system. Another terrible thing is that journald assumes something to be byte data when a log line contains a lot of numbers 🤷🏼♂️.

[carewolf]

Ext4 still looks competent though. Thanks for the benchmarks.

Yes. Never underestimate the performance of a well optimized inferior design

[LtdJorge]

unfortunately I also use systemd on gentoo. I just dont like sysvinit bash scripts although its more flexible, readable and configurable i just dont have enough experience with bash script to make a proper init script for the life of me. I wish systemd was less of a monolithic beast and more of a basic init system. Another terrible thing is that journald assumes something to be byte data when a log line contains a lot of numbers 🤷🏼♂️.

Tbh, I have no problem with it having so many uses, I prefer it to having 100 different ways of doing things, and if you don't like one utility, you can choose to not use it.

However, that doesn't mean having multiple choices is bad, in fact I welcome it. It's just that the Linux environment is too fractured. That's why I chose to use Systemd and as many of its tools as I could, for a more integrated and coherent system (then I use Sway, which doesn't integrate anything and makes you pick every piece by yourself unlike a DE, lol).

[Berniyh]

Yes. Never underestimate the performance of a well optimized inferior design

Or, more explicitely: if not for performance … why would you use Ext4?

Ok, it still has one or two other advantages, like support for case folding, but there isn't much apart from performance.