On Monday, 2009-08-24, David Baron wrote: > On Monday 24 August 2009 13:48:41 Kevin Krammer wrote: > > > KDE will not start if I have killed them. Runs fine without them until > > > Ilog out. > > > > Hmm, interesting. > > Didn't know there was a dependency on the system bus. > > > > > Since these were there before 4.3 and other recent Sid upgrades, they > > > in themselves are not the problem. Something in 4.3's core or in > > > recent Xorg is exercising them. > > > > Right, xorg depends on D-Bus and HAL for detecting devices. > > If Xorg needs them, so does KDE (I restart Xorg for new session because of > another longstanding bug). Not sure I understand this, but Xorg needing it does not imply KDE needing it. I though that KDE didn't need the system bus unless certain system integration features are wanted. > Once Xorg has started, I would assume it has its devices set unless dbus > reports, from udev, something added. So ... this mess should not be taking > CPU cycles any longer. Things should long have quited down. Right. That's what happens, doesn't it?. If I use htop, most of the time htop itself is the top entry. Sometimes it is X, sometimes plasma (I have a CPU monitor applet running), sometimes an application like Akregator. At the moment the first daemon listed it udev, after init and konversation. Currently running session daemons are: D-Bus session bus gpg-agent kded ksmserver kwalletd klauncher knotify ksysguardd pulseaudio nepomukserver nepomukservicestub akonadiserver akonadi_control 3 akonadi resources (distlist, vcard, ical) dcopserver (for some KDE3 apps, which means I have a second kded as well) sometime KIO slaves, when KDE apps do I/O All idling until used. Cheers, Kevin
Attachment:
signature.asc
Description: This is a digitally signed message part.