> P.S. About performance concerns: Some of the code that the RootLayer > thread executes scales pretty bad, for example the algorithm used for > rebuilding layer clipping. Before we worry about saving performance > by > trying to make locking unnecessary, we need to worry about the > algorithm first. Otherwise it is a ridiculous discussion. And before > we > even worry about performance at all, we need a correct > implementation. > Anything else is premature optimization. I agree. I know Adi has put a lot of work into the window management code, which RootLayer is part of. I also know that we are all working toward the same goal. Unfortunately, I am not up-to-date on the state of that code - keeping up with all the changes that all of you have been making is definitely not easy if you're already familiar with the code and nearly impossible if you're not. Here's a question to point the discussion in the right direction, if someone would be so kind as to enlighten me: what is the current one and its pros & cons? Likewise, what kinds of algorithms are there out there which could also work. Is there a better one? --DW