not to mention Apple Education sells it for $500

i don’t see those recovered addicts you mention

what was it

Debian as more mainstream than Arch?

is it really still the same system if it’s a radically divergent userland? (see: android as a linux distribution)

Ubuntu has also gotten a lot better and promising in recent years, too.

Except Snap shenanigans. Snap always shenanigans. The Snap pushing is eternal… Though Entrop is right that it’s mostly just a power-user worry.

I didn’t say that it should be part of the OS. You’re repeating a different (compelling) argument that doesn’t change whether this field is opt-out when this entire subthread is about whether these fields are opt-out since kent_eh said the other fields were opt-out.

exactly, users should disable the optional birthDate field

a headless interface is still an interface

being accessible headless is still accessible

Yeah I see your point. I feel like it’s entirely reasonable, though. Like those who went to the PR, they saw something in the news and decided to do something about it with their abilities, throwing aside whether that’s good for a moment. I certainly would not call the optional JSON schema for user records a critical component especially as no existing fields were modified, just new ones.

(And FWIW it’s systemd, very different from the kenrel, though I do feel like you know what you meant to talk about. systemd’s code quality is relatively notorious anyways.)

Your point is true, but I’m saying its impact is also optional.

You specifically named the bills from Colorado and NY. They simply do not include those. https://leg.colorado.gov/bills/SB26-051 https://www.nysenate.gov/legislation/bills/2025/S8102/amendment/A I wholeheartedly believe my characterization of these bills are faithful.

Specifically due to how barebones it is, it is trivial to modify yourself the birthdate sent to applications, as long as you have the system password.

Coming out of nowhere just for this contribution is hecka sus though.

I’m fairly sure this is the first systemd pull request that many here have viewed. I wouldn’t say we’re coming out of nowhere.

That is the purpose, but the field is implemented as optional and modifiable with admin privileges.

(same for the birthDate field)

I agree with you but repeating your arguments in mass replies does not make it stronger.

That’s the Minnesota bill. The PR does not comply with that. You can read on how to the California law and NY and Colorado bills basically say to give the user a drop-down to select their birth date.

I don’t see what’s wrong with implementing it as an add-on to the https://en.wikipedia.org/wiki/Gecos_field as the PR in question does. It’s the most logical place as the location to store user information and is even easier to opt out of—you just edit a file—than choosing whether to compile Linux with/add to DKMS a kernel module.

Edit: One can see https://github.com/systemd/systemd/blob/8878df45c1a58afdfb500fdc53ec50e057a240ce/docs/USER_RECORD_BLOB_DIRS.md?plain=1#L103 for an example of a user record file and its path. Further documentation you can read at https://github.com/systemd/systemd/blob/8878df45c1a58afdfb500fdc53ec50e057a240ce/man/systemd-userdbd.service.xml#L36 and https://github.com/systemd/systemd/blob/8878df45c1a58afdfb500fdc53ec50e057a240ce/docs/USER_RECORD.md .

nah as an anarchist i am against silence. i’m just saying that in our capitalist society open source maintainers do not in fact have responsibility to the community, only to their market share, and this works slightly less dysfunctionally than proprietary because come what may the opposition may fork it. but that and the transparency and the ability to volunteer your labor for them are the only things that open source does guarantee.