W3C home > Mailing lists > Public > www-style@w3.org > April 2011

Re: canvas origin?

From: Alan Gresley <alan@css-class.com>
Date: Sat, 09 Apr 2011 19:10:54 +1000
Message-ID: <4DA0229E.9010907@css-class.com>
To: Anton Prowse <prowse@moonhenge.net>
CC: www-style@w3.org, Jack Smiley <zxcv_890@hotmail.com>
On 9/04/2011 4:37 PM, Anton Prowse wrote:
> On 09/04/2011 02:12, Anton Prowse wrote:
>> On 08/04/2011 18:50, Jack Smiley wrote:

>>> (assuming ltr mode) For example, if I do html: {margin-left: -50px},
>>> would the canvas origin be 50px to the left of the top left corner of
>>> the viewport?
>>
>> No. It would (as always, by definition) be incident with the top left
>> corner of the initial containing block, which is typically the rectangle
>> seen in the viewport when the page loads.


This is incorrect in what browsers do.


>> The leftmost 50px of the root
>> element's margin area box would be to the left of the initial containing
>> block (and usually impossible to scroll left to in browsers, although
>> that's not mandated anywhere AFAIK).
>
> Sorry, I meant to say: "The leftmost 50x of the root element's _border_
> box would be to the left of the initial containing block".


This depends on what properties are given to out of flow boxes. There 
can be two ICB and this is what browsers show.


> (The top left margin area corner is incident with both the top left
> corner of the initial containing block and the canvas origin [and
> typically with the top left corner of the viewport when the page first
> loads], as I said above.)
>
> Cheers,
> Anton Prowse
> http://dev.moonhenge.net


Again, this depends on what properties are given to out of flow boxes. 
Take these test cases.


http://css-class.com/test/css/visformatting/icb4.htm

http://css-class.com/test/css/visformatting/icb5.htm


For floats, the ICB is wider than the viewport and relates to the left 
margin edge of the root element in the first test or right margin edge 
of the root element in the second test.

For absolutely position elements with offset values of auto for left, 
the ICB is hidden outside the viewport and relates to the left margin 
edge of the root element in the first test.

For absolutely position elements with offset values of 0 (or positive or 
negative integers) the ICB relates to the viewport.

Note that the overflow is hidden in the first test and visible (with a 
scroll-bar) in the second test.


On 9/04/2011 10:12 AM, Anton Prowse wrote:
> On 08/04/2011 18:50, Jack Smiley wrote:
>>

>>>> This should probably be made a little clearer in the spec.
>>
>> I agree. How can we make that happen?
>
> It's enough that this conversation is taking place on this mailing list
> to ensure that it'll be considered for the Errata to the upcoming
> publication of CSS21 as a Proposed Recommendation.


Yes for the Errata.


-- 
Alan http://css-class.com/

Armies Cannot Stop An Idea Whose Time Has Come. - Victor Hugo
Received on Saturday, 9 April 2011 09:11:22 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:39 GMT