The discussion was already made a thousand times about pratically all the points, from why there is no issue to udev being an external project and not supportin

The discussion was already made a thousand times about pratically all the points, from why there is no issue to udev being an external project and not supporting exclusively systemd (with an additional "fuck you gentoo developers" in the ML), to extreme appetite to fagocitate everything as fast as humanly possible, to pretty big bugs ("oh sorry, journald is not fast enough to write your gigabyte-sized core dump, let me cut it in half using this constant I randomly decided was good enough!" left alone for years) and to all the problems of the design of systemd [1]. Maybe it is my impression, but this blog post cheapens it - a lot. Sure, it was pushed hard by RH (inspired by Sun's SMF), it had some (quick) wins, it solved (some) issues for distro mantainers - though they have to remember to patch some components to not use defaults that make no sense, like google NTP servers - and the DEs jumped on it because it simplified mantaining their code (making them not portable in the process, but who cares!). It seems it "won"... Like HAL did win, a lot of time ago, in a similar way (IIRC). We'll see. It's not that heavily deployed in production yet. [1] http://blog.darknedgy.net/technology/2015/10/11/0/ http://blog.darknedgy.net/technology/2015/10/11/0/