Version: (using KDE 4.2.90) OS: Linux Installed from: SuSE RPMs When the plasma wallpaper is using marble, and there is no internet connection, the .xsession-errors file is filled with lines that say: Download rejected: Will try to download again in some time.
Hm, is there something we can do about this without removing the qDebug()s ?
@Jaime: Sure that SUSE is not shipping debug builds? @Torsten: Isn't that related to marble itself rather than the wallpaper?
Yes, I have the debug builds (to triage bugs). I did not tested that with marble (the application) running. But as wallpaper, there is an option to move the earth by clock time. With this option, the downloads are retried and retried every time the earth must be retraced.
Ok, now I got it. I think retrying to download should happen after a certain time-period instead of with every redraw. When used from inside a KDE session shouldn't Marble be able to watch the network state via Solid?
Sascha: That's not quite what happens. All the bitmap data that Marble shows is provided in small chunks ("tiles"). Whenever you move the globe then Marble will of course request a different tile that shows the matching location at the requested zoom level. Of course all this is done asynchronously. So most messages are due to different download requests and not due to trying the same tile over and over. If you want to "see" these tiles just start Marble with "marble --tile-id" Of course we don't retry to download tiles for every redraw (that would be absolutely incredible dumb - We have never done something that stupid during the whole Marble development). Instead we reschedule the download for about a minute until we retry once again. If the tile is still not available we continue to reschedule fetching the tile three times in total until we give up after a few minutes. So the vast amount of debug messages can be attributed to different download requests. Regarding the Solid request: Yes, if somebody provides a patch we can use Solid here. However this doesn't solve the issue as the problem would still be there for the Qt-Only version.
@tackat: Thanks for the insights. To sum it up, being verbose in a debug-build is hardly a "bug", can we close it?
Well, I fear that quite a few distributions don't switch off debug messages even for the release build, right? So what I am wondering about is whether there is a way to avoid that these kinds of messages end up in the .xsession-errors file ...
First, sorry for my poor description of what could happen. Reviewing the code, I vote for remove the qDebug()s, as they do not provide much information (for example, the tile to retry), and the code is working :-)
shall we close this one?
Well, looking at the code it seems to me that the implementation doesn't work as it should (it should only run into this code once for each tile and not on every single repaint). So I think that we should still keep this open. I'd suggest however that we create a define for this specific debug output that would be off by default.
*** Bug 202375 has been marked as a duplicate of this bug. ***
SVN commit 1046166 by nienhueser: Remove repeated (every minute) useless debug output. CCBUG: 196598 M +0 -1 ExtDateTime.cpp M +0 -1 SunLocator.cpp WebSVN link: http://websvn.kde.org/?view=rev&revision=1046166
I believe this one can be closed, I re-checked and no debug messages seem to be generated