[Webkit-unassigned] [Bug 31517] Bring WebGL support to Gtk port

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=31517





--- Comment #1 from Zan Dobersek <zandobersek at gmail.com>  2009-11-15 08:58:54 PST ---
Created an attachment (id=43247)
 --> (https://bugs.webkit.org/attachment.cgi?id=43247)
Work-In-Progress of WebGL implementation

Warning - this patch is pretty outdated (2 or more weeks), so it probably won't
apply easily or build properly.
It also contains some debugging leftovers and unnecessary stuff (methods, lines
etc.)

This implementation requires use of cairo's OpenGL backend. A recent version or
even trunk should be used, along with another patch applied. That patch will be
attached to this bug.

There are some notable problems with this implementation, but they can be
fixed.
The surface is not properly painted on the canvas. The surface is correctly
shown on the canvas only if the canvashas size of 300x150px, which is the
default canvas size. If the canvas is resized to a bigger size, the surface
resizes also, but only the rectangle (x= 0, y = 0, width = 300, height = 150)
of the surface is shown. When resized to a smaller size, the surface is shown
completely, but unevenly, relatively to the canvas's size.

The other problem is that the OpenGL content is rendered to the cairo's surface
upside down - OpenGL is using lower-left coordinate system while images use
upper-left coordinate system. This is described as twelveth pitfall at [1] and
should be simple to fix.

Other than that, the performance is, I feel, good. I've tested the webgl manual
tests and they all worked pretty well apart from the mentioned problems. I
experienced crashes on olliej's WebGL example[2], though.

[1] http://www.opengl.org/resources/features/KilgardTechniques/oglpitfall/
[2] http://nerget.com/webgl

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