Jordan Maris 🇪🇺 🇺🇦 #NAFOMar 25
@BuuBuu @aral @GrapheneOS @VoidLinux the patch literally adds a date of birth field to user data, that is all. It's not an API for age verification. In addition the information is set by the user.
Show threadh3mmy Mar 25

@jmaris @BuuBuu @aral @GrapheneOS @VoidLinux There is no legitimate reason an OS needs to have a DOB field for user data.

It's a fundamental concept for Data Minimization

"Don't store any more information than you actually need"

If an application or other user-space extension has a need for a dob, it should be implemented there, not as part of the init daemon.

Show threadJordan Maris 🇪🇺 🇺🇦 #NAFOMar 25
@h3mmy @BuuBuu @aral @GrapheneOS @VoidLinux systemd is not just an init daemon at this point. And again, if you don't want to fill in the field, simply don't fill it in. Problem solved.
Show threadh3mmy Mar 25

@jmaris @BuuBuu @aral @GrapheneOS @VoidLinux I feel like you're not hearing me. This is a design and process issue. I shouldn't have to opt-out of including PII. The OS should not require it unless there is a justifiable need. It's a change that impacts a wide population, and there wasn't even a proper design discussion.

I also understand the context, with the new California law that is mandating age gates. This change can be seen as theater to comply without implementing tighter controls. However, it's not set in stone yet, which is why other projects did not implement equivalent changes.

There is a background issue with how large systems incrementally constrain populations. Minority groups tend to be sensitive to that because it's happened to us a lot. That is an explanation for some of the response you see generally. People see how "innocent" and "optional" fields today become "mandatory" and oppressive fields tomorrow. So, I understand the reaction, and dislike when valid concerns are dismissed.

Show threadJordan Maris 🇪🇺 🇺🇦 #NAFOMar 25
@h3mmy @BuuBuu @aral @GrapheneOS @VoidLinux again, you are under no obligation to provide it. It isn't 'opt out". At worst there will be a field at install time to input it, just like there is a field for full name, and if you don't fill it in, sobeit. Also we are talking about open source here. It is specifically resilient to this kind of bullshit.
Show threadAral BalkanMar 25
@jmaris @h3mmy @BuuBuu @GrapheneOS @VoidLinux Right, if people don’t like it, they can just fork their operating system. I’ve forked three this week and it’s not even Thursday.
Show threadJordan Maris 🇪🇺 🇺🇦 #NAFOMar 26
@aral @h3mmy @BuuBuu @GrapheneOS @VoidLinux as I said, they can just not fill out the field.
Show threadAral BalkanMar 26
@jmaris @h3mmy @BuuBuu @GrapheneOS @VoidLinux You realise this isn’t just about a field, right, but what it represents and what it enables?
Show threadJordan Maris 🇪🇺 🇺🇦 #NAFOMar 26
@aral @h3mmy @BuuBuu @GrapheneOS @VoidLinux again. How can this impact you if the field is not filled in?
Show threadKim Spence-Jones 🇬🇧😷
@jmaris @aral @h3mmy @[email protected] @GrapheneOS @VoidLinux
You do sound a bit naïve. There are both social implications and architectural ones. Socially, it’s the thin end of a wedge. Architecturally it is fundamentally wrong that the OS should be managing PII. For a start, many Linux systems don’t have an associated ‘user’, and some are multi-user. I can’t see why anyone is defending systemd’s position on this.
Mar 26 at 11:18amWeb
Show threadJordan Maris 🇪🇺 🇺🇦 #NAFOMar 26
@KimSJ @aral @h3mmy @BuuBuu @GrapheneOS @VoidLinux so do you envisage a world where filling in this field becomes compulsory?
Show threadKim Spence-Jones 🇬🇧😷Mar 26
@jmaris @aral @h3mmy @[email protected] @GrapheneOS @VoidLinux
We seem to be moving in that direction.