W3C home > Mailing lists > Public > www-style@w3.org > June 2014

Re: [css-transforms] Initial value of transform-style

From: Simon Fraser <smfr@me.com>
Date: Fri, 06 Jun 2014 11:29:28 -0700
Cc: Matt Rakow <marakow@microsoft.com>, Rik Cabanier <cabanier@gmail.com>, "www-style@w3.org" <www-style@w3.org>
Message-id: <DF8A8920-6751-4991-A5AE-CE0333463C97@me.com>
To: Dirk Schulze <dschulze@adobe.com>
On Jun 6, 2014, at 12:45 AM, Dirk Schulze <dschulze@adobe.com> wrote:

> Hi Simon,
> 
> On Feb 27, 2014, at 8:03 PM, Simon Fraser <smfr@me.com> wrote:
> 
>> On Feb 27, 2014, at 10:41 AM, Matt Rakow <marakow@microsoft.com> wrote:
>> 
>>>> Simon recently presented a solution at the CSS F2F and posted a document to public-fx [1] to clear some things up.
>>>> 1: https://docs.google.com/document/d/1mNF7Z67WnnV05RqXa37PmfvRbgAZwj7-h-7Y_uQ_UPE/edit?pli=1#
>>> 
>>> Ah, thanks for the pointer.  It looks like the presentation was on the day I wasn't able to make it unfortunately :(
>>> 
>>> Simon, how do you feel about the "auto" value vs. using inheritance as I suggested in my other mail [1]?  Right now it sounds like "auto" would act as "inherit, unless a transform is applied in which case flat" which still seems a bit confusing to me.  For example, it seems unintuitive that these two examples would produce different results:
>>> 
>>> <div style="transform-style: preserve-3d; transform: rotateX(30deg);">
>>>  <div>
>>>      <div>
>>>          <div>
>>>              <div style="transform-style: preserve-3d; transform: rotateX(30deg);">
>>> 
>>> <div style="transform-style: preserve-3d; transform: rotateX(30deg);">
>>>  <div>
>>>      <div style="transform: scale(1)">
>>>          <div>
>>>              <div style="transform-style: preserve-3d; transform: rotateX(30deg);">
>>> 
>>> Thanks,
>>> -Matt
>>> 
>>> [1] http://lists.w3.org/Archives/Public/www-style/2014Feb/0783.html
>> 
>> Making it inherited is an interesting suggestion; Iíd have to think about that. The performance characteristics are obviously a concern for documents with html { transform-style: preserve-3d; } but itís possible that UAs could avoid allocating additional resources for elements with no 3d-transformed descendants in this kind of content.
> 
> It actually sounds very reasonable and the goals of an inherited properties seem to fit with the goals we had in mind with a new value Ďautoí quite nicely. Did you have more time to think about it?

Hi Dirk

I did a little more editing on <https://docs.google.com/document/d/1mNF7Z67WnnV05RqXa37PmfvRbgAZwj7-h-7Y_uQ_UPE/edit?pli=1#> but havenít had time to finish it off.

The most recent changes addressed the fact that the previous text resulted in a behavior where a single 3d-transformed element would render in front of everything else on the page, which is not desired behavior (and doesnít match current implementations). The current text results in a behavior where a single 3d-transformed element only punches through to the front of its CSS stacking context.

Simon
Received on Friday, 6 June 2014 18:29:59 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:39:22 UTC