In the environment that I am trying to use dthlmXtree, with about 700 nodes, the collapse action seems to be slower than the expand. We have a theory that this may be because the tree is saving of the state of all of the child nodes within the node that is being collapsed (since the tree does seem to retain the expand/collpase state of all child nodes). We were wondering if there is a was to turn off the processing that is retaining the state of the child nodes, in hopes that this might improve the performance of the collapse action.
The saveOpenStates functionality doesn’t force item rendering, so it must be pretty fast event for huge datasets ( and must not cause problems for 700 nodes for sure )
The default collapse|expand functionality will not affected by preserving states of child items ( basically child items ignored, so they must not affect performance at all )
If problem still occurs for you - please provide any kind of sample ( you can send it directly to firstname.lastname@example.org )