☃️karolherbst☃️

Mostly working on Rusticl and NVK

Victim of the woke mind virus

languages: en, de

pfp is a head shot of a snowman with a red pot as its hat

header is a photo of a full moon and leafless trees in front at night

Last Notes

2026-08-21 10:11:13 UTC

- reply

@nprofile…sqs7 well at least if you keep up for long enough this would open up the option for you to work freely within the EU, which does come with benefits on its own and knowing that you could just move to a different country for work if things become terribly politically is kinda freeing tbh.

2026-08-04 22:58:54 UTC

Autism season for sure can't come quick enough because this summer is truly hell....

2026-08-01 21:41:29 UTC

I have a theory about what sorts of programmers like to use AI and what sorts totally doesn't like it and I feel it's kinda related to how one "thinks" while in the act of programming.
If one "thinks" in their spoken language, AI might be a good fit, but if one does not, the translation from the "abstract thoughts" to "spoken language" is too high for AI to be beneficial in any sense for that person.
I'll have to keep thinking about this.

2026-07-15 14:53:13 UTC

Okay.. we all know what forking a project means. We all know that.
What I do not know is what sporking a project would mean...

2026-06-18 19:00:54 UTC

- reply

@nprofile…2l72 Nooot sure if discard=always is a thing with btrfs, but maybe the driver accepts it anyway, but couldn't find proof of it.
The docs seem to specify "discard, discard=sync, discard=async, nodiscard" as valid options (https://btrfs.readthedocs.io/en/latest/ch-mount-options.html)
And it sounds like async is preferred, but maybe try what you find first and see if it helps. But might want to check kernel or the mount option in case discard=always got rejected.

2026-06-18 17:32:26 UTC

- reply

@nprofile…2l72 uhh.. I haven't considered compression, so maybe that changes things as well, so writing zeroes might not do much actually.
I know in the past it was a reliable thing to try, but we could be past those days.
But yeah, asking on the cachyos forums might be a good idea for now.
It could be that cachyOS trimms regularly, but maybe for some reason that doesn't work on your SSD. You said that fstrim doesn't work, right? Might be relevant here.
Could also be something entirely different.

2026-06-18 16:47:39 UTC

- reply

@nprofile…2l72 but yeah... not quite sure what's wrong there... slow write and high read performance usually hint towards trim related issues with SSDs.
Could also try filling your entire disk with a zero file (dd if=/dev/zero or something), deleting it and see if that fixes it.

2026-06-18 16:43:32 UTC

- reply

@nprofile…2l72 is the discard flag enabled on your mount points? If so, then it should auto trim anyway.

2026-06-18 15:49:17 UTC

- reply

@nprofile…2l72 maybe it's trimming related? At least I hope for you it's trimming related, because that's fixable at least...

2026-04-02 12:23:23 UTC

- reply

@nprofile…2l72 though I'd argue that if you put emphasize on "jobs are lost" it's not a true statement. While modern tech can replace some human labor, those jobs normally never go fully away anyway. And it's also not the goal in itself, but rather it should be making jobs less mundane or exhausting.
So I think "new tech to replace jobs completely" -> tech bro
"new tech to help people do their jobs, but this could make us needing less people" -> not tech bro.

2026-04-02 12:19:41 UTC

- reply

@nprofile…2l72 yeah.... that's a tricky one

2026-04-02 12:12:09 UTC

- reply

@nprofile…2l72 first try: "I scored 0% on the Tech Bro Assessment" uwu

2026-02-03 20:55:51 UTC

My thunderbolt dock randomly disconnects or gets randomly disconnected and it's kinda annoying because it's annoying..
Not seeing much in dmesg, just:
thunderbolt 0-0:1.2: retimer disconnected
thunderbolt 0-0:1.1: retimer disconnected
thunderbolt 0-1: device disconnected
and I wonder if it's a bad cable or what's up there. Anybody any ideas how to figure out what's going on there?

2026-01-04 09:34:37 UTC

- reply

@nprofile…7hfy I think the problem is, that we often have to do the thing to learn that something wasn't great and the environment and users change over time.
Which might be bad today, might have been a good enough solution in the past.
Of course we have to acknowledge that things change and we need to be open for adaption and exploration of better solutions.
I do agree that some cool projects are created out of spite, but I think I have a problem with spreading negativity around.

2026-01-03 21:41:19 UTC

We can only have true collaboration in FOSS if we stop treating it as a competition (a.k.a. "this project is shit and my project is so much better for those irrelevant technical reasons that are in fact just opinions but they matter really much because I am the best coder on this world and decided so")

2025-12-31 03:29:45 UTC

- reply

@nprofile…0j2v mhhh yeah.. maybe it's something I should look into next year, but at least here it's not installed by default, but that might as well be due to it being an old install of fedora or because I removed it a few years ago or some other reason I'm not aware of...
anyway, same global shortcut on gnome it seems.

2025-12-31 01:30:13 UTC

- reply

@nprofile…0j2v I sometimes wonder how I'd use my computer after I was in an accident or whatever and my eyes are either permanently or temporarily not really usable (hopefully that never happens, but you never know)
and I think the only real conclusion here is that screen readers should be a hard dependency always installed, and it should be possible to enable it via a global keyboard shortcut or something so that none of those guides are necessary in the first place.

2025-12-30 11:01:37 UTC

- reply

kinda surprised that dnf doesn't use it for system upgrades, because it does skip a potential full disk encryption prompt