It's a known and very annoying bug.
Issue was already reported in:
http://hub.qgis.org/issues/7321
http://hub.qgis.org/issues/7472
2013/4/23 Mathieu Pellerin nirvn.a...@gmail.com
Greetings,
I've filed an issue over the weekend which IMO is a really serious
regression which shouldn't be left unresolved prior to the imminent
2.0 release.
Long story short, when a layer symbology is set to categorized,
graduated, or rule-based, all features will disappear when editing
feature(s) on that layer (by moving feature or one of its vertex). If
the user then leave the edit mode (saving changes or not), all
features will reappear.
When features disappear, an error pops up in the log window saying
Already active iterator on this provider was closed.
Data itself is not impacted, but it's a super scary behavior for the
average user. It got me to cancel edits many, many times before I
understood what was happening.
Issue filed can be reached here: http://hub.qgis.org/issues/7660
Mathieu
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer
___
Qgis-developer mailing list
Qgis-developer@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/qgis-developer
