Affine Connection @ AffineConnection @lemmy.world

Posts

0

Comments

131

Joined

2 yr. ago

ChatGPT came up with

Memory
Navigation
Enhancing
Methods
Of
Name
Integration
Coding.

So many of the results I see are incredibly obvious fakes.

Yes, but they were satisfied when IBM was given a license to use JSLint for evil.

Windows contains a lot of stuff proprietarily licensed to Microsoft, so they don't have an option to turn all of it into free software even if they wanted to.

I just love people that don’t beat around the bush and are straight to the point.

It's obviously possible to be stern and direct while maintaining composure instead of having a temper tantrum.

One can sternly address serious mistakes by a subordinate without being outright mean about it. Doing so calmly and seriously is usually more effective anyway.

That doesn't mean he should be a jerk about it.

how is that even allowed

lobbyists, I presume

Just because he's not a Doctor of Medicine doesn't mean he's not a doctor. A Doctor of Chiropractic is exactly that, regardless of its questionable merits.

A medical doctorate would be more relevant to a neurologist.

I was under the impression that the assertion that chiropractic neck manipulation causes vertebral artery dissection is often suggested, but that evidence of such a causation is inconclusive. I certainly believe it, but I can't assume. Chiropractors may twist the inconclusiveness into suggesting that such neck manipulation is safe, but that's a fallacy.

That's simple, but it's a completely unnecessary waste of I/O. You could create a writable snapshot of the btrfs root as a subvolume, edit the fstab and any other relevant files within that new subvolume, reconfigure the bootloader to specify that subvolume as the root filesystem (as a Linux kernel command line argument) instead of the btrfs root, and then reboot. After rebooting, the original btrfs root can be mounted, and everything unwanted from the original root (other than the new subvolume and its ancestor directories, obviously) can be deleted. Do not delete anything that you didn't want to lose the changes to on the original root subvolume that you did after creating the snapshot, as the snapshot only remembers what you did before, as well as the changes made specific only to it (like the fstab).

If one wanted to create multiple subvolumes for different purposes, the above procedure can be modified. For instance, if one wanted a separate subvolume mounted at / vs /home, then one can create two writable snapshots, empty out the contents of home in new subvolume 1 (but not the /home directory itself because you want the directory to exist for something to mount onto it), empty out everything outside of home within new subvolume 2, move the contents of home therein up one directory and remove the /home directory itself. Now, one can edit the fstab in new subvolume 1 as appropriate (not forgetting to have new subvolume 2 mount at /home), edit any other relevant files, reconfigure the bootloader to tell the Linux to use new subvolume 1 as the root subvolume, then reboot. Finally, one can remove the unnecessary files from the original root.

Edit:

It is arguably better to manually specify the new root when booting in the Linux kernel command line, and not reconfigure the bootloader until you successfully boot. After success, (if the following is relevant to your system) use update-grub, and it should look at fstab to automatically reconfigure the bootloader accordingly to use the appropriate new subvolume as specified at fstab.

This is what I did years ago to one of my own systems, although I don't know anything about Timeshift and how it requires things to be set up (I have my own backup scripts that are run by cron). I could have just snapshotted the btrfs root directly for snapshots, but I wanted the snapshots to be cleanly separated from the subvolume used as the Linux VFS root (except when I explicitly mount them).

It was less useless for that purpose when cp and cat were less I/O efficient compared to dd with the appropriate block size, which isn't as much the case now as it used to be.

dd does not stand for "disk duplicator". That's a modern backronymization that doesn't reflect the original general usage of the command which is to "convert and copy". Efficiently (with respect to I/O) copying raw data is only one of its intended purposes; it also converts text encodings.

Assuming /dev/device is not a symbolic link, you might as well

        cp image.iso /dev/device
  

You might as well call cp the same thing

        cp source "`realpath destination`"
  

But dc is a reverse polish calculator Unix program. It's even in the Bell Labs' Unix 1^st edition manual.

/usr/local/share/porn

What if I had a pet Gorgon?