Re: [RC6] c548-leadin-000 is imprecise

Le Dim 15 janvier 2012 20:22, "Gérard Talbot" a écrit :
> Hello,
>
> I have been trying to create a reftest for:
>
> [RC6]
> http://test.csswg.org/suites/css2.1/20110323/html4/c548-leadin-000.htm
>
> [nightly-unstable]
> http://test.csswg.org/suites/css2.1/nightly-unstable/html4/c548-leadin-000.htm
>
>
> line  10    div { font: 25px Ahem; line-height: 50px; width: 3em; color:
> lime; background: green; border: solid black; }
>
> line  11    span { background-color: red; }
>
>
> With such values, it is not predictable (although maybe possible to
> establish how browsers handle this) how tall is going to be the
> top-half-leading and the bottom-half-leading: it could be 12px or 13px.
>
> a) top-half-leading and bottom-half-leading could be respectively: 12px
> and 12px
> or
> b) top-half-leading and bottom-half-leading could be respectively: 12px
> and 13px
> or
> c) top-half-leading and bottom-half-leading could be respectively: 13px
> and 12px
> or
> d) top-half-leading and bottom-half-leading could be respectively: 13px
> and 13px
>
> After time spent with Firefox's DOM inspector, Opera DragonFly, Chrome
> Web Inspector (and by doing maths with offsetTop, offsetHeight, metrics
> of rendered layout) it seems that option b) is chosen by such 3
> browsers:
> top-half-leading and bottom-half-leading is respectively: 12px and 13px.
>
> The div has an offsetTop of 51px while the span has an offsetTop of
> 66px:
>
>   66px (offsetTop of span)
> - 51px (offsetTop of div)
> -  3px (border-top-width of black border)
> =======
>   12px (height of green gap between topmost black
>         border and 1st bright green stripe)
>
> The bottommost green gap (between 2nd lime stripe and the black
> border-bottom) can also be measured by doing maths in the same manner
> with some offset values provided by webpage debugger tools.
>
>
> But such manner of spliting 25px in 2 is not an ensured and predictable
> outcome in browsers.
>
> Proposed solution:
>     10    div { font: 26px Ahem; line-height: 50px; width: 3em; color:
> lime; background: green; }


The above analysis was correct although font-size needed to be dividable
by 5px.

I have updated the test to be using

line 10    div { font: 25px Ahem; line-height: 51px; width: 3em; color:
 lime; background: green; }


> --------
>
> border: solid black;
>
> The issue with such black border has been mentioned in
> http://test.csswg.org/shepherd/testcase/c545-txttrans-000/#event-4
> and also in
> http://lists.w3.org/Archives/Public/public-css-testsuite/2010Oct/0137.html
>
> Since such black border does not participate, is not involved in the
> test itself, then I think we should remove it.


I've decided to leave the black border as it is... albeit the black
border could be removed too as it does not participate in the test
itself. Since the most important pass/fail conditions of the test are
presence of colored stripes and absence of red, then the pass/fail
conditions sentence only need to indicate such.

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

CSS 2.1 Test suite RC6, March 23rd 2011:
http://test.csswg.org/suites/css2.1/20110323/html4/toc.html

CSS 2.1 test suite harness:
http://test.csswg.org/harness/

Contributing to to CSS 2.1 test suite:
http://www.gtalbot.org/BrowserBugsSection/css21testsuite/web-authors-contributions-css21-testsuite.html

Received on Friday, 6 April 2012 20:31:02 UTC