@pcwalton By the way, two questions that just popped into my head regarding Pathfinder:
1. Since context switches are expensive, do you render each tile to a texture atlas? Or does each tile get its own float texture?
Yeah. This is the main downside of Pathfinder 3. I’ve thought of ways to mitigate the issue, though, mostly using extensions to do the coverage calculation and compositing in one pass. One nice thing, though, is that the mask can be cached from frame to frame in many cases.
-
-
I only thought about the problem for a little bit, but it seems to me like the core issue is that for proper compositing, you don't just need to know how much each pixel is covered, you need to know how much it is covered *by each layer*. So you need multiple coverage masks?
-
Hard to explain my idea in a tweet, but the basic idea is to composite each layer “just in time” to save memory bandwidth.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.