The default should be the device's native blocksize, and some devices misreport. You also lose performance if you use a larger blocksize than necessary.
If we can, I'd like to get a quirks list in place, but there have been higher priorities.
Do each of the other filesystems have their own quirks list? That seems suboptimal. Oh, I guess it's because it's in the user space mkfs tool of each, not the kernel.
Still changing the on disk format as required, but we're at the point now where the end user impact should be negligible - and we aren't doing big changes.
Just after reconcile, I landed a patch series to automatically run recovery passes in the background if they (and all dependents) can be run online; this allows the 1.33 upgrade to run in the background.
And with DKMS, users aren't having to run old versions (forcing a downgrade) if they have to boot into an old kernel. That was a big support issue in the past, users would have to run old unsupported versions because of other kernel bugs (amdgpu being the most common offender).
bcachefs was always a module. You don’t want it in your kennel if you are not using it. The difference is that it used to ship in the mainline source code and be built as a module that was already built and on your drive.
If you build bcachefs as a module yourself (via DMKS or directly), it works the same as if you got it with your distro.
If you use bcachefs as root, the danger is booting with a kernel that lacks the module.
I hate that bcachefs is not in the kernel, and my primary distro does not use DKMS. But, if you can get a module built, there is no loss of functionality or performance.
yes, it will. But we do want to communicate properly with systemd and let the user know what's going on if mount has to take awhile because of some sort of recovery (instead of timing out), and various other things.
related, plymouth integration to let users know when their machine is booting up if a drive or the filesystem is unhealthy
What drives me personally nuts about the CentOS saga is all the “community” hand-waving about creating a bit for bit clone of a distro.
There can be no “community just shipping builds of RHEL code as, by definition, you cannot change anything. That means you cannot contribute. In my view, an Open Source “community” cannot just be people that use things for free. It is supposed to be about collaborating to build things.
At least now we have Alma Linux which strives to be ABI compatible with RHEL but builds it themselves from CentOS Stream. They actually build something. They can actually contribute (and they do). They can innovate. For example, they have continued the x86-64v2 builds even though RHEL has abandoned them. On Alma, you can at least claim to be building a community.
I do not use any of these distros by the way, in case you think I am shilling something.
Toothbrush UX is the same today as it was when we were hunter gatherers: use an abrasive tool to ablate plaque from the teeth and gums without removing enamel
As somebody who's tried using a miswak [0] teeth-cleaning twig out of curiosity, I can say with confidence it's not the same experience as using a modern toothbrush. It's capable of cleaning your teeth effectively, but it's slower and more difficult than a modern toothbrush. The angle of the bristles makes a huge difference. When the bristles face forward like with a teeth-cleaning twig your lips get in the the way a lot more. Sideways bristles are easier to use.
That’s just not what user experience means, two products having the same start and end state doesn’t mean the user experience is the same. Imagine two tools, one a CLI and one a GUI, which both let you do the same thing. Would you say that they by definition have the same user experience?
If you drew both brushing processes as a UML diagram the variance would be trivial
Now compare that variance to the variance options given with machine and computing UX options
you’ll see clearly that one (toothbrushing) is less than one stdev different in steps and components for the median use case and one (computing) is nearly infinite variance (no stable stdev) between median use case steps and components.
The fact that the latter state space manifold is available but the action space is constrained inside a local minima is an indictment on the capacity for action space traversal by humans.
This is reflected again with what is a point action space (physically ablate plaque with abrasive) in the possible state space of teeth cleaning for example: chemical only/non ablative, replace teeth entirely every month, remove teeth and eat paste, etc…
So yes I collapsed that complexity into calling it “UX” which classically can be described via UML
I would almost define "experience" as that which can't be described by UML.
Ask any person to go and find a stick and use it to brush their teeth, and then ask if that "experience" was the same as using their toothbrush. Invoking UML is absurd.
You know some of us old timers still remember a time before people just totally abandoned the concept of having functional definitions and iso standards and things like that.
Funny how we haven’t done anything on the scale of Hoover Dam, Three Gorges, ISS etc…since those got thrown away
User Experience also means something specific in information theory and UX and UML is designed to model that explicitly:
Notably, the terms "UX" and "experience" are not present in that document. UI and UX are different things. UX is a newer concept that is more based on observing users and their emotional reactions to using the product.
UML and functional definitions and iso standards are still important, it's just not UX.
Good luck never observing users using your product. Not everything is a space shuttle, recall that we are talking about toothbrushes here.
The GUI is usually the problem. I have booted Xubuntu and it's still slow. Slow systems with older GPUs simply can't keep up with newer desktops. Most of the times I need a terminal but a lightweight desktop can help if I quickly want to open a browser and search something so that I can copy past more complicated commands.
Why is 512 the default and, if 4096 is better, why is this not the default instead?
reply