i wonder, if the current 4.5 binary is already speed optimised, as the ui seems to be pretty slow.
example: i use the “primarySelectionChanged” signal to sync a filter for a treeview in my plugin.
this signal has already an uncool behaviour, as if the user _switches_ selections it first emits with NULL, and than a 2nd time with the new selected node, means it always works 0 > 1 > 0 > 1… means my filtered treeview also replicates this.
in ds3 this was at least not visible, but with ds4.5 (not yet tested in ds4.0) the - of course unwanted - intermediate “no selection = no filter” state is clearly visible for a few moments. not a deal breaker, but it looks nasty; and also leads to the above question.
i may of course build some buffer around the results of this signal to prevent this, but it’s of course not really elegant coding to do so.
ps: but i’m happy to see, that the bug with the wrong numSelectedNodes count is now corrected
pps: at least in chrome the line-breaks of forum posts seem to work strange; the line height within paragraphs is a little large already, and to get some gap between paragraphs, i need to insert 2 empty lines…