Version: (using KDE 4.3.0) OS: Linux Installed from: SuSE RPMs I'm using the "Globe" wallpaper with satellite view, mercator projection, and following the sun on 300s updates. It works fine when I shutdown the computer and start it again but if I suspend to ram the wallpaper starts at the same position after the resume. It updates regularly, on 300s intervals, but it as a difference from what should actually be. The difference is the time lapse between suspending and resuming the computer. I have to choose other wallpaper and then the globe again to have it in the correct position again.
Confirmed, on 4.3.0 from Fedora 11 rpms. Doesn't reset correctly after suspend to ram or suspend to disk. Same problem with the world-clock widget. cheers M.
Still there with 4.3.2 on F11.
*** Bug 211310 has been marked as a duplicate of this bug. ***
I believe that this bug resides in the Marble libs. When "Follow the sun" is enabled, the code actually calls: MarbleMap::sunLocator()->update(); Seems like the sun locator does not (yet) take these interruptions into account.
Hi, I can confirm that this bug is present also in KDE 4.4.1 in Debian unstable/experimental-snapshots. My setup is slightly different from the one of the original poster: my map wallpaper is not in "follow the sun" mode, but fixed to a given location. The wallpaper should show the "daylight" satellite map and the "night" satellite map cycling according to the time of the day. If the PC is on it works fine, but if I suspend it, the position of the maps is not updated correctly after resume, and the difference is the time lapse between suspend and resume.
this will happen with any and all time based components: there is no "we've just unsuspended" signal available to us (at the OS level). this will hopefully be fixed one day (our powermanagement devs have been trying to work with the linux projects around powermanagement to get this fixed), but until then there is precisely zero we can do about these situations.
Hi Aaron, thank you for your answer! Things are clearer now. Just guessing: I think that the first time the wallpaper is painted, it checks the clock to set the starting point for the day/night separation. Then, it just updates the wallpaper by a constant interval at every 300 seconds. Am I correct? Couldn't it be possible to perform the check of the clock time at every update, and not only the first time, in order to see if the clock has actually changed by more than 300 seconds? In this way the effects of this bug could be greatly mitigated, since there would be a mismatch between the actual time and the wallpaper position only in the first 5 minutes after resume (at most). I know that it is not a perfect solution, but for the time being it could make the wallpaper usable again for notebook users, until a proper fix is found. Just my 2c. Ciao Mario
I'll take a look at it.
*** Bug 233462 has been marked as a duplicate of this bug. ***
*** Bug 225130 has been marked as a duplicate of this bug. ***
Do you reproduce with kde 4.4.4 or 4.5 beta ?
I'm using 4.4.4 under Debian Unstable and the bug is still there. I've not yet tried 4.5 beta, though.
*** Bug 189634 has been marked as a duplicate of this bug. ***
*** Bug 198828 has been marked as a duplicate of this bug. ***
I have the same problem KDE SC 4.7.1.
Since updating to 4.8, I decided to try this again.... and I think everything works. Can anybody check whether that is also true for them?
I was among the ones experiencing this bug up to KDE 4.4.4. After having ceased using the globe wallpaper because of it, I decided to give it another try under KDE 4.9.5 (unofficial Debian repositories). I can confirm what the previous poster says: everything seems to behave correctly now, and the sunlight is updated to the correct position even after a suspend. I think that the bug can be closed, and I am removing my votes from it.
Thanks for checking and reporting back, Mario.