[Webkit-unassigned] [Bug 33520] Mapping from local to container coord space is O(N^2)

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=33520





--- Comment #16 from James Robinson <jamesr at chromium.org>  2010-02-02 14:36:51 PST ---
(In reply to comment #15)
> (In reply to comment #14)
> > Yes, I've seen comments to that effect of doing all coordinate transforms via
> > RenderObjects throughout the code.  Is there a bug or a set of patches tracking
> > that code?
> 
> No, it is not being worked on.

OK.  I don't think I fully understand what this means.

My understanding is that currently a RenderLayer represents a set of
RenderObjects that
exist in the same coordinate space.  Is the idea that a RenderObject would
either exist
in its container's coordinate space or define a new one (in which case it would
also
have a RenderLayer hanging off of it?), and all mappings between
local<->container
coordinate spaces would traverse through RenderObject::container()s?

I don't quite understand why the mapping is sometimes done through the
RenderLayer
hierarchy and sometimes through the RenderObject container hierarchy currently.
 Is it
always the case that in the absence of transforms the two are equivalent?

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.