In the migration to systemd, the cups startup has been overly simplified. It does not call anymore "portrelease 631", so the cups daemon can not listen on port 631 which is still held by portreserve.
CC: (none) => mageia
Ahh good catch... There should be an ExecPre= added to the main unit I guess.... Of course more generally, we should probably start phasing out portreserve... If systemd does all the port listening early on then there is no problem which portreserve solves... but that's probably a slightly bigger job :)
Sorry, I meant "StartExecPre=portrelease 631" But, having had a look at the package, I think we should just drop this port reservation thing in cups, or at least move it to a subpackage that people can install with sysvinit if they still want to. It would very much be a manual process. As previously mentioned, the fact that systemd should open the port very early on should mean that all will be well without the need for the portreserve stuff. WDYT? I did patch the systemd unit for cups but then thought better off it. I just get the feeling that things would be odd if we carried on using port reservation for cups (and IIRC I was the person who originally requested some magic here as some strange NFS thing was preventing cups from starting on some of my machines - long time ago now tho' :D)
it seems this has been fixed, I can open http[s]://localhost:631 again with cups-1.5.2-1.mga2.src.rpm systemd-43-1.mga2.src.rpm initscripts-9.34-5.mga2.src.rpm
CC: (none) => smiling.diego
Oh I forgot to update this bug. I updated cups a while ago and also removed the portreserve stuff from it. So yeah I'll close it now. Thanks for pinging :)
Status: NEW => RESOLVEDResolution: (none) => FIXED
just trying to be helpful :)