W3C home > Mailing lists > Public > www-style@w3.org > October 2010

RE: [css3-2d-transforms] Transforms and background-attachment:fixed

From: Sylvain Galineau <sylvaing@microsoft.com>
Date: Wed, 13 Oct 2010 00:06:19 +0000
To: Simon Fraser <smfr@me.com>
CC: "www-style@w3.org list" <www-style@w3.org>
Message-ID: <045A765940533D4CA4933A4A7E32597E280134E1@TK5EX14MBXC111.redmond.corp.microsoft.com>
For 2D, I understand the performance concern. Particularly for smaller
devices. So we could leave it undefined or make it optional. I'd rather
have the spec say implementations should do this in a normative section.

For 3D, the answer depends on what you define the viewport in this context.
If the viewport can't be transformed and is always in the plane of the user's
browser window - reasonable assumptions, I think -then the answer is the same
one you suggested in the 2D intro: the fixed background is always 'flat' against 
the plane of the screen and you see it through the element, whatever perspective 
it may have at the time.

From: Simon Fraser [mailto:smfr@me.com] 
Sent: Tuesday, October 12, 2010 4:42 PM
To: Sylvain Galineau
Cc: www-style@w3.org list
Subject: Re: [css3-2d-transforms] Transforms and background-attachment:fixed

On Oct 12, 2010, at 3:43 PM, Sylvain Galineau wrote:

The red section of the introduction [1] asks:

# What do fixed backgrounds do in transforms? They should probably 
# ignore the transform completely, since - even transformed - the object 
# should be acting as "porthole" through which the fixed background can be 
# viewed in its original form.

The suggested behavior would seem to be the one authors would expect. No
browser currently does this, however. The spec should define this normatively.

[1] http://www.w3.org/TR/css3-2d-transforms/#introduction

For 2D transforms, would be possible to implement the described behavior, but
at great cost to animation performance.

For 3D, it would be extremely hard. What does a fixed background look like for
something that is angled towards you?

Received on Wednesday, 13 October 2010 00:06:53 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:38:39 UTC