An Idea about init system

With the rise of age-verification and identity verification, systemd adding userdb records for compliance, It will be harder to use systemd in future. I had an idea that if whonix, tails and qubes developers unite to make a similar init system which have all the features of systemd, like (systemctl and service, socket and timer with sandboxing features like protectingsystem etc from services) and journalctl for easy centralized logging would be awesome for the linux ecosystem. Or fork systemd to strip all unwanted and not-needed components. I believe these efforts from these highly-skilled developers would be unstoppable and unmatchable. All of you are using systemd down the base so it will be really helpful to all others out there as well.

This is not the case. systemd did not add support for age verification, they simply added a mechanism for storing user date of birth in the same database that already allowed storing various other bits of sensitive personal information like email addresses. The feature in systemd can be optionally used by dedicated age verification software in theory, but on its own it is entirely benign. systemd could add support for storing genetic information or the full contents of the user’s brain in the same database and it would be similarly benign (though obviously using such features would be rather difficult with current technology :P)

Both of these options are:

systemd is a massive and very fast-moving project that enjoys widespread adoption. Reimplementing it would require resources that none of these projects have even put together and would result in a sub-standard product that would only be used by those specific distros and would likely die quickly. Beyond that, replacing systemd in Debian with any other init system is a monumental task that would very likely break compatibility with large numbers of existing Debian packages.

Forking systemd sounds easy in theory but in practice is not because of the speed with which systemd moves. There are already multiple systemd forks that remove the age storage patch; for instance:

Synced with upstream two weeks ago, yet already 206 commits behind. Another fork of the above, but even more out of date:

Given that the code being removed is a) benign, b) not an invasion to privacy, c) removing it would be a monumental task as demonstrated by multiple existing sorely outdated forks, and d) almost no one would actually use it (maybe no one at all), this is not something we can do or have a good reason to do.

1 Like

Closing as per:

1 Like