Michael Biebl <email@michaelbiebl.de> (2014-08-11): > > Am 11.08.2014 21:58, schrieb Cyril Brulebois: > > It really would have been nice to put debian-boot@ in the loop though. > > Changing init when we're in the middle of trying to release d-i isn't > > really something that has no consequences… Your first mail: | Sorry for the inconvenience. | | I'll keep that in mind in the future. | I did file the bug report before your d-d-a mail, so wasn't aware there | is an imminent release of d-i. I must acknowledge this is partly my fault. Releasing this beta is long overdue (I first mentioned it in a reply to a linux upload notification back in June IIRC); there were some technical reasons, but also some personal reasons which led me to delay this on the one hand, and not to announce it on the other hand. Anyway, the end result is: I'm not going to complain if there are too many mails on -boot@ as opposed to not enough, so don't hesitate. :) Your second mail: > Btw, since "init" still pulls in sysvinit-core atm, lowering the > severity of sysvinit and sysvinit-core shouldn't have an effect for > d-i. > > If it does, could you elaborate, what happened/went wrong? AFAICT right now, nothing went wrong; quite the contrary. I had been experiencing the net.agent 30s hang since beginning of August when I started getting stuff into shape. After some bits of debugging and sharing test results (#debian-kernel), Ben and I found out that "apt-get install systemd" was a fix, so we didn't investigate any further. Surprisingly, the bug vanished suddenly today when I started checking latest parted changes! Because: tada, systemd was my brand new shiny init! I haven't been able to perform any testing on non-Linux, but I think I'm OK with a release even if we finally find out that these images need a fix at some point. Bottom line: all good so far, just surprising. :) Mraw, KiBi.
Attachment:
signature.asc
Description: Digital signature