daviddahl/css-transform-research.md

## css-transform-research.md

      
    Raw
  

              css-transform-research.md
            
          
    CSS transform research

Find, in the relevant standard(s), why the following works in Gecko:


elem.style.WebKitTransform

A:

It is difficult to tell from reading the spec why the browser implementations behave the way the do. Outside of the standards process, human decisions around not supporting unstandardized browser functionality and behavior drove the usage of -webkit-* and -moz-*. digging through the mailing lists will no doubt turn up the exact reasons.
For now, the techical reasons are found in code.
In Gecko there is a CSS property alias for WebkitTransform here:  http://mxr.mozilla.org/mozilla-central/source/layout/style/nsCSSPropAliasList.h#234
But the following doesn't:


elem.style[-webkit-transform]

A:

A reference to -moz-transform is in a list of CSS properties (-webkit-transform is not included), here: http://mxr.mozilla.org/mozilla-central/source/layout/style/nsCSSPropList.h#4207 but alas, not corresponding -webkit-transform.
I would love to know exactly why - not the "techical why" but I fear this will take too much of my day away.
And relatedly, why both work in Chrome.

A:


Here is the changeset for WebKit supporting un-prefixed transform: https://trac.webkit.org/changeset/181832
Bug: https://bugs.webkit.org/show_bug.cgi?id=142927

Naturally, the original implementation was probably completed before the spec was adopted by vendors. The latest W3C spec appears to still be A working draft. Whuck?
Finding relevant bugs (Bugzilla, crbug, spec bugs, etc.) and links to source code would be helpful

A:


I remember when Henri posted this, it was widely discussed: https://hsivonen.fi/vendor-prefixes/
Inside of this thread[1] it would seem that unprefixing has been discussed for some time, but due to the spec being in WD form, no one wanted to for the longest time, sigh.
[1] http://lists.w3.org/Archives/Public/www-style/2011Nov/0322.html
w3c spec: https://www.w3.org/TR/css-transforms-1/#transform-property
whatwg references w3c spec: https://compat.spec.whatwg.org/#-webkit-transform

It looks like Mozilla did some research into prefixes to help determine which -webkit prefixixes might warrant support. transform is talked about a lot here: https://bugzilla.mozilla.org/show_bug.cgi?id=708406#c6 It would seem that getting it unprefixed was the goal, however, I am unsure what the final result was - at that time. A few questions for dbaron or jet might help me understand better.