New comment by TeusLollo on void-packages repository https://github.com/void-linux/void-packages/issues/38367#issuecomment-1207867875 Comment: 1. I was not suggesting a "takeover", at best a fork. In previous months, although I can't find a link anymore, one of the Void devs was saying that `mkinitcpio` was under-maintained, thus Arch devs were planning to switch fully to `dracut`, which however is going deeper into systemd hard-dependencies. That's obviously concerning for Void Linux, and I had never heard of `booster`. 2. Even then, sometimes `dracut`, at least on certain of my systems, fails loading the `intel_pstate` driver, although it's totally random and can't be replicated. This never happened when I used Arch with `mkinitcpio`, hence I assumed it was further proof of `dracut` ditching non-systemd support. 3. Hell, I thought `mkinitcpio` on Void was somewhat of an _experimental/it's your problem_ approach, but it seems I was wrong and it's fully supported instead. I may actually try it. 4. Or I may go with the new boy in town, `booster` I just wanted to remark that I was not suggesting any "takeoverist" approach. I was just wishing for developers not to depend upon software that appears to be hostile towards non-systemd approaches. Hell, the whole `eudev` story is similar, so I was just presuming eventually a similar joint-effort approach would appear for inittramfs generators, which may not be necessary anymore with the arrival of the new boy in town, `booster`.