[haiku-commits] Re: r38473 - haiku/trunk/src/servers/app

  • From: "Axel Dörfler" <axeld@xxxxxxxxxxxxxxxx>
  • To: haiku-commits@xxxxxxxxxxxxx
  • Date: Tue, 31 Aug 2010 10:56:02 +0200

Stephan Assmus <superstippi@xxxxxx> wrote:
> When reading the patch, I figured as much, but it's between the lines 
> and a clear example of too little comments or explainations in the 
> code. 

I agree that we have too little documentation, but Window::Backmost() 
is actually documented. Removing it without even checking what it does 
is not between the lines, and should be avoided.

> I think what Clemens tries to do should still be made possible. Even 
> with the constraints taken care of, it should be possible to mess 
> with 
> the window order in a deterministic way, and that means SendBehind() 
> can 
> actually raise a window too. But I don't think the computation of the 
> dirty region, or rather the region that is still clean, was anywhere 
> near complete after the change.

Sure, it is definitely possible to implement this, I just don't think 
that it matches the semantics of BWindow::SendBehind() anymore. It 
would even cause bad user experience in case the window *is* already 
behind the specified window, because your application would change the 
window order without any benefit.
It might have a use case (even though I can't think of one outside of 
manually grouping windows -- in which case the subset window approach 
could be used instead), but it should not clobber the semantics of the 
existing SendWindowBehind() call.


Other related posts: