User Agent: Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:11.0a1) Gecko/20111113 Firefox/11.0a1
Build ID: 20111113031758
Steps to reproduce:
Render a page with -moz-transform-style : preserve3d; style
Example can be found here:
CSS applied to that page is:
The body also has a -moz-perspective : 1000px;
In case it might help, I tried making the direct parent have the perspective, and also tried making it not float, but this didn't seem to affect the behavior.
The element with that style is not rendered.
The element should be rendered, preserving its 3D properties. The equivalent -webkit-* styles are applied as well, so a webkit browser can be used for comparison (works with Chrome and Safari/iOS).
Reproduced on Mac OSX using Nightly
Created attachment 574432 [details] [diff] [review]
Include the current frame bounds in preserve-3d overflow calculations
This fixes the image being invisible, but doesn't cause the perspective to be applied.
As far as I can tell the perspective frame (the <body> element) isn't a parent of the transformed frame (the <img> element), so perspective shouldn't be applied.
I'll construct a reduced test case for this and ask for clarification on the spec.
The <body> is not a parent because the direct parent of the <img> is float:right? Or because the perspective has to be set on a direct parent?
The spec says
The ‘perspective’ property applies the same transform as the perspective(<length>) transform function, except that it applies only to the positioned or transformed children of the element, not to the transform on the element itself.
That property is not inherited. The "img" is not a child of the <body>. So the body's 'perspective' does not affect it.
If Webkit behaves differently, maybe we should change the spec to match Webkit. That would need to be discussed on www-style.
(Maybe perspective inherits in Webkit?)
Created attachment 574475 [details]
Reduced test case
Definitely looks like webkit is treating this as an inherit property, or they are interpreting 'children' as 'descendents'.
I'll take a look at the WebKit source code and post to www-style.
https://bugzilla.mozilla.org/show_bug.cgi?id=716524 looks similar to this one, although, still reproduces the issue.
*** Bug 724363 has been marked as a duplicate of this bug. ***