W3C home > Mailing lists > Public > public-css-testsuite@w3.org > September 2010

Review Report on section 9.3.2 (left box offset) ~= 42 testcases

From: Gérard Talbot <css21testsuite@gtalbot.org>
Date: Sun, 12 Sep 2010 19:43:00 -0700
Message-ID: <6f59a7b627033a274e11abec541052dc.squirrel@cp3.shieldhost.com>
To: "public-css-testsuite@w3.org" <public-css-testsuite@w3.org>
Cc: "Arron Eicholz" <arron.eicholz@microsoft.com>
Hello all,

This report is mostly for Fantasai and Arron Eicholz.


Summary
*******

- only 2 testcases with Fractional pixel problem
- no testcase rejected
- a few testcases use a border-width with 0.2in which is 19.2px: those
do not trigger a rounding up of value in browsers but they are
nevertheless creating fractions of a pixel.



*********************************
Section 9.3.2 Box offsets: 'left'
*********************************

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/float-001.htm
http://test.csswg.org/suites/css2.1/20100815/html4/float-002.htm
http://test.csswg.org/suites/css2.1/20100815/html4/float-003.htm
http://test.csswg.org/suites/css2.1/20100815/html4/float-004.htm


   <link rel="help"
href="http://www.w3.org/TR/CSS21/visuren.html#propdef-right">
   <link rel="help"
href="http://www.w3.org/TR/CSS21/visuren.html#position-props">

should be replaced with

   <link rel="help" title="9.5.1 Positioning the float"
href="http://www.w3.org/TR/CSS21/visuren.html#float-position">

Approved with such changes

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-005.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-007.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-019.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-020.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-031.htm

Approved

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-043.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-044.htm

Fractional pixel problem!

border-left: 1cm solid red; /* 37.795px but computed as 37px: rounding
down occurs */
margin-left: -3.54cm; /* -133.795 but computed as -134px: rounding up
occurs */

Proposed replacements:
http://www.gtalbot.org/BrowserBugsSection/css21testsuite/left-043.htm
http://www.gtalbot.org/BrowserBugsSection/css21testsuite/left-044.htm

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-067.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-068.htm

0.2in is 19.2px.

Although I see no unfavorable consequences from this, out of this, it
may not be the case for all possible browsers or for future browsers.

Approved nevertheless

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-076.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-077.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-078.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-079.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-080.htm

ahem font is not necessary for these testcases

Approved nevertheless

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-091.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-092.htm

For these tests, I think
            div
            {
                font: 12pt ahem;

should be instead

            div
            {
                font: 20px ahem;

although I see no unfavorable consequences with 12pt with current browsers.

Approved nevertheless

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-103.htm
and
http://test.csswg.org/suites/css2.1/20100815/html4/left-104.htm

0.2in is 19.2px. Although I see no unfavorable consequences from this,
out of this, it may not be the case for all possible browsers or for
future browsers.

I propose to change the assert from

<meta name="assert" content="The 'left' property sets a nominal length
value in percentages.">

to

<meta name="assert" content="The 'left' property sets a nominal length
value in percentages. A percentage set on left offset of a box refers to
its containing block's width.">

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-112.htm

<meta name="assert" content="The 'left' property sets a value of 'auto'.">

could be changed to

<meta name="assert" content="The 'left' property sets a value of 'auto'.
For relatively positioned box, if both 'left' and 'right' are 'auto',
then the used and computed values are '0' (i.e., the boxes stay in their
original position).">

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-113.htm

0.2in is 19.2px. Although I see no unfavorable consequences from this,
out of this, it may not be the case for all possible browsers or for
future browsers.

 <meta name="assert" content="The 'left' property sets a value of
'inherit'.">

could or should be changed to

<meta name="assert" content="The 'left' property sets a value of
'inherit'. Here, #div1 is moved to the right by the width of its
containing block which is body thanks to 'left: 100%'. Reserved keyword
'inherit' refers to the computed value of the parent: here, it is #div1
and it is 100% which refers to the width of the containing block which
happens to be the same box. #div1 has a set width of 1in; so the used
left offset value will be 1in for the inner div.">

to better reflect what is happening in the tescase.

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-001.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-002.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-003.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-004.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-005.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-006.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-007.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-013.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-014.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-015.htm

I propose to change

<link rel="help"
href="http://www.w3.org/TR/CSS21/visuren.html#position-props">

to

<link rel="help" title="9.7 Relationships between 'display', 'position',
and 'float'" href="http://www.w3.org/TR/CSS21/visuren.html#dis-pos-flo">

because this is the only spot in the spec where elements with display
set to  table or sub-table elements are described that they can be
absolutely positioned and how the position of the box can be determined.

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-008.htm
http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-009.htm

Approved

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-applies-to-010.htm

The testcase should allocate sufficient horizontal space to display a
bullet and the expected results should mention that a bullet list-marker
must be produced.

Instead of left: 0, left: 32px or 2em would be appropriate.

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-offset-001.htm

This type of testcase would be best if the containing block (#div1) had
borders and padding. Nevertheless,

Approved

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-offset-002.htm

Approved

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-offset-003.htm

1-
#container's background is transparent: so padding-right area is not
colored or noticeable. It has no impact, no meaningful consequence in
the testcase. You can remove it and it won't affect the test in any
noticeable manner.

2-
#div1's left offset applies to an area without a padding (there is no
padding-left). So, the test ends up to be the same as if the containing
block had no padding at all.

 <meta name="assert" content="The 'left' property, for absolute
positioning, specifies the offset of the element in relation to the
containing block's left padding edge.">

but the containing block has no left padding area.

3-
To talk about containing block's padding edge is IMO a way of saying
things which may end up confusing the CSS beginner, CSS intermediate
more than it can help him/her. The nearest positioned ancestor has or
has not a padding box. The containing block for an abs. pos. box is
formed, is established by such nearest positioned ancestor's padding
box. Strictly speaking, the containing block is not a CSS box.

Proposed replacement:
http://www.gtalbot.org/BrowserBugsSection/css21testsuite/left-offset-003.htm

----------------------

Author: Microsoft

http://test.csswg.org/suites/css2.1/20100815/html4/left-offset-percentage-001.htm

Proposed replacement:
http://www.gtalbot.org/BrowserBugsSection/css21testsuite/left-offset-percentage-001.htm


regards, Gérard
-- 
Contributions to the CSS 2.1 test suite:
http://www.gtalbot.org/BrowserBugsSection/css21testsuite/

CSS 2.1 test suite (beta 3; August 15th 2010):
http://test.csswg.org/suites/css2.1/20100815/html4/toc.html

CSS 2.1 test suite contributors:
http://test.csswg.org/source/contributors/
Received on Monday, 13 September 2010 02:43:37 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 20 September 2010 17:52:03 GMT