Friday, March 20, 2026

Linux Bleg

Instead of getting overly technical, in this post, I'll just point you to Sk's explanation at OSTechNix of what all the systemd merge of Pull Request #40954 -- adding a "birthdate" field to userdb JSON records --  entails and why some people find it concerning.

I'm one of those people. Specifically, I consider the "slippery slope" argument compelling. Yes, the field is "voluntary" for now ... BUT!

  1. It's being done specifically to facilitate "compliance" with surveillance state laws; and
  2. Having that "compliance" hook installed at all will just encourage more such laws and eventually some kind of process that switches the field from "voluntary" to "mandatory" and comes with some kind of intrusive method for ensuring the supposed accuracy of the information in it; and
  3. The birthdate field is NOT encrypted, so any attack that grabs data from JSON records will reveal that personal info to the attacker.
IMO, that's bad from two standpoints. First, it's a surrender to / empowerment of the surveillance state apparatus; second, the "mandatory" and "verification" elements, when they arrive, will mean more bloat and less efficiency in systemd itself.

The two main OSes I use -- Linux Mint and Raspberry Pi OS -- have systemd as their system/service manager (ChromeOS, on my laptops, uses Upstart).

So -- does anyone reading this post happen to use a Linux OS that doesn't incorporate systemd? And if so, what you think of it? I prefer a Debian fork to more ... sui generis ... distros, but I'm willing to consider the latter.

No comments: