16003 2012-02-16 18:18:56 +0000 CSSTransformValue is defined, but there's no way to get one 2012-02-23 05:20:42 +0000 1 1 1 Unclassified CSS Transforms unspecified All All RESOLVED LATER P2 normal --- 1 ayg smfr ayg cmarrin dbaron dino dschulze eoconnor glenn smfr public-css-bugzilla oldest_to_newest 64177 0 ayg 2012-02-16 18:18:56 +0000 http://dev.w3.org/csswg/css3-transforms/ defines a CSSTransformValue interface. However, unless I'm missing something, it provides no way to obtain any object that implements this interface, such as operations on other interfaces that return a CSSTransformValue, or attributes on other interfaces of type CSSTransformValue, or a constructor on the CSSTransformValue interface. Either some way of obtaining a CSSTransformValue needs to be defined, or the interface should be removed from the spec. Cf. bug 15443. 64178 1 smfr 2012-02-16 18:33:26 +0000 stylesheet.cssRules.item(0).style.getPropertyCSSValue('transform'); would give you a CSSValueList, and each member of that list is a CSSTransformValue.. 64179 2 smfr 2012-02-16 18:33:57 +0000 However, given the state of the CSS OM standardization, I agree that we might want to just remove all the CSS OM-related stuff from the spec. 64180 3 ayg 2012-02-16 18:40:21 +0000 getPropertyCSSValue() is not defined by any specification that I was able to find. When I checked before, it didn't seem to work at all interoperably across browsers. I think we should keep the CSSOM stuff, but only the stuff that actually has specs written for it (even if they're unstable or inaccurate). 64181 4 dschulze 2012-02-16 18:46:32 +0000 (In reply to comment #3) > across browsers. I think we should keep the CSSOM stuff, but only the stuff > that actually has specs written for it (even if they're unstable or > inaccurate). Sorry, that makes no sense at all. On the one hand you say we don't need it because we can't access it. On the other hand you say we should leave it in the Spec? If we leave it in the Spec, why not making it really useful, instead of leaving it broken and not useful at all? 64188 5 ayg 2012-02-16 20:15:48 +0000 (In reply to comment #4) > Sorry, that makes no sense at all. On the one hand you say we don't need it > because we can't access it. On the other hand you say we should leave it in the > Spec? No, I'm saying we should leave stuff in the spec only if it's actually defined. For instance, getComputedStyle() is defined by CSSOM, so we should leave in the stuff that relates to getComputedStyle() and try to make it as accurate as possible. But if there's no way that's defined to get a CSSTransformValue, we shouldn't spec it. Either we should define a way to get one, or remove it from the spec. Either way is fine by me, if non-WebKit implementations have support for it or are interested in adding it. 64189 6 dbaron 2012-02-16 20:26:08 +0000 Per http://lists.w3.org/Archives/Public/www-style/2003Oct/0347.html getPropertyCSSValue has been deprecated and is intended to be replaced; we shouldn't be specifying new features on top of it, so I agree this should be removed. 64202 7 dbaron 2012-02-17 00:34:27 +0000 To be clear, my proposal is to remove the entirety of section 17 (DOM Interfaces) from css3-transforms. 64203 8 dbaron 2012-02-17 00:35:55 +0000 And this issue has been raised multiple times before, e.g.: http://lists.w3.org/Archives/Public/www-style/2009Sep/0120.html http://lists.w3.org/Archives/Public/www-style/2011Nov/0710.html 64237 9 ayg 2012-02-17 16:21:22 +0000 (In reply to comment #7) > To be clear, my proposal is to remove the entirety of section 17 (DOM > Interfaces) from css3-transforms. I concur. 64239 10 glenn 2012-02-17 16:29:22 +0000 (In reply to comment #7) > To be clear, my proposal is to remove the entirety of section 17 (DOM > Interfaces) from css3-transforms. +1 64473 11 dschulze 2012-02-23 05:20:42 +0000 This gets delayed to CSS4 Transforms.