Version: unspecified (using KDE 4.4.4) OS: Linux If the network connection is not up when one logs into KDE, plasmoids like the weather forecast or the newsticker do not show any content even after the network connection got established (knetworkmanager). They wait until they do their normal refresh. I wonder whether this is due to plasma not having a central signal that tells all plasmoids that a new network connection is now available or whether this is a bug in every single plasmoid that suffers from this. Reproducible: Always
Same here been strugling with this flaw. Found a work around, but in that case plasma won't start until network is up.
"I wonder whether this is due to plasma not having a central signal that tells all plasmoids that a new network connection is now available or whether this is a bug in every single plasmoid that suffers from this." neither; it's a bug in the DataEngines they are using. and it's each engine. there is no way to centralize a fix for it, but there is a central "network is up now" signal, and it is in fact working in the weather plasmoid. what it isn't doing is calling forceImmediateUpdateOfAllVisualizations() when it is updated in this fashion. this is a per-DataEngine or per-Plasmoid issue. i've started fixing the weather engine, but it's fairly non-trivial given how poorly it was implemented. the rss DataEngine's problem is likely that it only calls forceImmediateUpdateOfAllVisualizations() on the first source, and/or RssEngine::cachesUpToDate() is failing in that case. i don't think b.k.o is really the best place to track this kind of issue, in any case, since it's per-plugin. Sven: why didn't you post this to plasma-devel? you are on the list according to lists.kde.org .. hm.
SVN commit 1155751 by aseigo: actually push updates when the network updates; tested envcan thoroughly. the rest follow the same pattern, but need further testing CCBUG:245639 M +16 -3 ions/bbcukmet/ion_bbcukmet.cpp M +3 -2 ions/bbcukmet/ion_bbcukmet.h M +34 -32 ions/envcan/ion_envcan.cpp M +4 -5 ions/envcan/ion_envcan.h M +1 -3 ions/ion.cpp M +1 -5 ions/ion.h M +20 -9 ions/noaa/ion_noaa.cpp M +1 -1 ions/noaa/ion_noaa.h M +17 -3 ions/wetter.com/ion_wettercom.cpp M +1 -0 ions/wetter.com/ion_wettercom.h M +68 -82 weatherengine.cpp M +19 -10 weatherengine.h WebSVN link: http://websvn.kde.org/?view=rev&revision=1155751
(In reply to comment #2) > i don't think b.k.o is really the best place to track this kind of issue, in > any case, since it's per-plugin. Sven: why didn't you post this to > plasma-devel? you are on the list according to lists.kde.org .. hm. I thought it was a bug and that bugs belong into bko and not on the mailinglist which is rather for discussion. Is it that it affects several plasmoids which makes the difference? I'll try to remember it for the next time I find this kind of issue.
*** Bug 247127 has been marked as a duplicate of this bug. ***
I added it already to the duplicate (247127), sorry for that, but I think that's the right place for the comment: Konqueror and Choqok show the same behavior. Both don't have network access if they where started before the network was up. Only restarting Konqueror and Choqok after the network is up helps to get a connection.
SVN commit 1169287 by asouza: Make weather dataengine work if started without network Always call setData() using the original source so it's initialized and when we reset the ions, they have sources connected to them and are able to communicate back. Also, fix the ions to update their data into the weather engine and later emit the signal that will make the engine force the clients (plasmoids) to update using the new data. This mostly fix the problem of turning the computer on without network, getting network later and the weather plasmoids not updating. CCBUG: 226022 CCBUG: 245639 M +5 -1 envcan/ion_envcan.cpp M +9 -3 ion.cpp M +5 -1 noaa/ion_noaa.cpp M +9 -1 wetter.com/ion_wettercom.cpp WebSVN link: http://websvn.kde.org/?view=rev&revision=1169287
The rss dataengine is properly updating on trunk - tested with rssnow plasmoid.
SVN commit 1170005 by asouza: backporting 1155751 actually push updates when the network updates; tested envcan thoroughly. the rest follow the same pattern, but need further testing CCBUG:245639 M +68 -82 weatherengine.cpp WebSVN link: http://websvn.kde.org/?view=rev&revision=1170005
SVN commit 1170021 by asouza: backporting 1155751 actually push updates when the network updates; tested envcan thoroughly. the rest follow the same pattern, but need further testing CCBUG:245639 M +16 -3 ions/bbcukmet/ion_bbcukmet.cpp M +3 -2 ions/bbcukmet/ion_bbcukmet.h M +34 -32 ions/envcan/ion_envcan.cpp M +4 -5 ions/envcan/ion_envcan.h M +1 -3 ions/ion.cpp M +1 -5 ions/ion.h M +20 -9 ions/noaa/ion_noaa.cpp M +1 -1 ions/noaa/ion_noaa.h M +17 -3 ions/wetter.com/ion_wettercom.cpp M +1 -0 ions/wetter.com/ion_wettercom.h M +19 -10 weatherengine.h WebSVN link: http://websvn.kde.org/?view=rev&revision=1170021
SVN commit 1170055 by asouza: backporting r1169287 Make weather dataengine work if started without network Always call setData() using the original source so it's initialized and when we reset the ions, they have sources connected to them and are able to communicate back. Also, fix the ions to update their data into the weather engine and later emit the signal that will make the engine force the clients (plasmoids) to update using the new data. This mostly fix the problem of turning the computer on without network, getting network later and the weather plasmoids not updating. CCBUG: 226022 CCBUG: 245639 M +5 -1 envcan/ion_envcan.cpp M +9 -3 ion.cpp M +5 -1 noaa/ion_noaa.cpp M +9 -1 wetter.com/ion_wettercom.cpp WebSVN link: http://websvn.kde.org/?view=rev&revision=1170055