Hi Manish, Sometimes, you can get negative positional values for Nodes, that do not appear on screen themselves, or those that already ceased to exist. I've looked over the application you have mentioned (if I guessed right :) ), and there is a 'table cell' element in the UI tree which is right under a 'table column header' element. The 'table column' has positive positional values, while the other doesn't since it's not a real part of UI, but more of a helper meta-element. Perhaps you've been accessing that? Or - if I understand it correctly - you've assigned some 'table cell' node to a variable and then added more items into this table-menu? Then possibly the whole tree-table would be reloaded with new object representation even for those items already in place before. Well, simply said, you get negative positional values for those Nodes in the accessibility tree that either don't represent actual UI elements or are not valid anymore (were closed, re-rendered etc.) I suggest that you look over your app with 'accercizer', a tool similar to sniff, that also shows object properties including size or position. Please feel free to let me know if you need any help, perhaps I didn't understand your situation correctly. You can also find me at #dogtail channel on freenode, that might be the simplest. Cheers, Vita ----- Original Message ----- > Hello All, > > I am facing a peculiar problem. I have a hardware component and a > control panel(cpl) provided for the same (can't name due to obvious > reasons :) ). Now cpl is a typical GTK app with left pane as menu > (rendered as table-cells). I make a parent-child tree out of this > table cell using position X, Y. However as soon as I add 2nd > component, the cpl returns few menu items with high negative > positional values. There can be 2 things. 1) issue is with gtk > control tree building or with the way the cpl menu code is written > or 2) the way dogtail sees it. I wanted to know the scenarios where > dogtail might see positional values high negative. > > Can anyone please help/explain this behavior. > > Thanks, > Manish K Katoch > > ----------------------------------------------------------------------------------- > This email message is for the sole use of the intended recipient(s) > and may contain > confidential information. Any unauthorized review, use, disclosure > or distribution > is prohibited. If you are not the intended recipient, please contact > the sender by > reply email and destroy all copies of the original message. > ----------------------------------------------------------------------------------- > -- test mailing list test@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe: https://admin.fedoraproject.org/mailman/listinfo/test
