Thanks for shedding real data on the issue. I was skeptical myself. I checked out Poettering's Google+ profile and see no mention of this yet, and I was kind of surprised.
I am kind of disappointed that there is some baity qualities to this article, specifically referencing how Linus chewed out a systemd developer. He did that, but I recall it not being directly related to his work on systemd and it negatively impacting the kernel. Kay Sievers is a well-known problem causer as Linus is concerned, so this is not news.
Keep in mind if you find the mailing list thread referred to, Greg Hartmann (gregkh), the release maintainer of the Linux kernel, arguably part of the inner echelons, is responsible for the kdbus branch eventually getting merged into the mainline kernel, that is the kernel driver that will internalize dbus as a main (if not only) IPC of the kernel and reducing the overheard of using dbus now (reducing 12 operations per dbus call to 3 inside the kdbus driver, IIRC from Lennart's video). Again, this is the work of Lennart Poettering pulseaudio fame, and now much more heated systemd fame. So to pretend the Linux kernel is opposed systemd work is not truthful. Some core devs have taken it on and are staking themselves on it. If this gregkh tidbit does not make that obvious I do not know what does.
Can someone who knows more comment on what the substantive changes are thus far? Is the kdbus work a prime motivator of this? I love systemd hate as much as the next guy, but I was hoping we would get more facts from the HN crowd.
(EDIT: I know I will get downvoted, but I do use systemd and I am not its biggest fan; I just used Arch and got used to it; everyone has a right to choose their tools, and init systems ain't different.)
