[Webkit-unassigned] [Bug 98913] Change transformation precision from double to float

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=98913





--- Comment #4 from Simon Fraser (smfr) <simon.fraser at apple.com>  2012-10-10 20:33:08 PST ---
(In reply to comment #3)
> One thing we should be careful of is propagation of rounding errors.  For WebKit's purposes, it may not matter as the various floating point operations eventually convert into whole pixel values, but I have seen cases in scientific computing where the vagaries of floating point/double conversion lead to difficult to debug problems.
> 
> If we aren't doing much math where floating point values are being raised to large powers, or trigonometric functions results are being chained together there may be little downside.

There are certainly cases where rounding errors would be compounded; we map points and rects through accumulated transforms for hit testing etc, and it's not hard to create content (like almost edge-on layers) where rounding matters.

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