W3C home > Mailing lists > Public > public-css-testsuite@w3.org > January 2017

Re: [css-writing-modes-3] 112 new sizing orthogonal tests in section 7.3.1 and section 7.3.2

From: Gérard Talbot <css21testsuite@gtalbot.org>
Date: Wed, 04 Jan 2017 19:07:46 -0500
To: Public CSS Test suite mailing list <public-css-testsuite@w3.org>
Message-ID: <ee444d3fa58f55f07844630e69df5a39@gtalbot.org>
Le 2017-01-04 00:31, Gérard Talbot a écrit :
> Hello,
> 
> 32 new tests for section 7.3.1:
> 
> sizing-orthog-prct-[ htb-in-vlr | htb-in-vrl | vlr-in-htb | vrl-in-htb
> ]-[001-008]
> 
> These 32 tests checks how an orthogonal block is sized when its inline
> size is given a percentage value. The 4 possibilities of orthogonal
> context are checked
> 
> http://hg.csswg.org/test/rev/8a0a8f2afe1a
> 
> - - - - - - -
> 
> 20 new tests for section 7.3.2:
> 
> sizing-orthog-htb-in-vlr-[001-024]
> 
> These 20 tests check the sizing of an 'horizontal-tb' orthogonal block
> with its inline size set to 'auto' inside an auto-sized 'vertical-lr'
> containing block under various code scenario conditions in relation
> with this complex formula:
> 
> min(max-content, max(min-content, min(available space, initial
> containing block's size)))
> 
> http://hg.csswg.org/test/rev/6f2742210ef9
> 
> - - - - - - -
> 
> 20 new tests for section 7.3.2:
> 
> sizing-orthog-htb-in-vrl-[001-024]
> 
> These 20 tests check the sizing of an 'horizontal-tb' orthogonal block
> with its inline size set to 'auto' inside an auto-sized 'vertical-rl'
> containing block under various code scenario conditions in relation
> with this complex formula:
> 
> min(max-content, max(min-content, min(available space, initial
> containing block's size)))
> 
> http://hg.csswg.org/test/rev/0664939ab9ea
> 
> - - - - - - -
> 
> 20 new tests for section 7.3.2:
> 
> sizing-orthog-vlr-in-htb-[001-024]
> 
> These 20 tests check the sizing of an 'vertical-lr' orthogonal block
> with its inline size set to 'auto' inside an auto-sized
> 'horizontal-tb' containing block under various code scenario
> conditions in relation with this complex formula:
> 
> min(max-content, max(min-content, min(available space, initial
> containing block's size)))
> 
> http://hg.csswg.org/test/rev/3cef54b15491
> 
> - - - - - - -
> 
> 20 new tests for section 7.3.2:
> 
> sizing-orthog-vrl-in-htb-[001-024]
> 
> These 20 tests check the sizing of an 'vertical-rl' orthogonal block
> with its inline size set to 'auto' inside an auto-sized
> 'horizontal-tb' containing block under various code scenario
> conditions in relation with this complex formula:
> 
> min(max-content, max(min-content, min(available space, initial
> containing block's size)))
> 
> http://hg.csswg.org/test/rev/ba4fb3bfec76
> 
> - - - - - - -
> 
> Some comments about those tests.
> 
> 1-
> Chrome 57.0.2970.0 and Edge 13 fail about 12% of all those 112 tests;
> for Chrome 57.0.2970.0, I will know exactly the numbers of failed
> tests later.

Chrome 57 fails 16 tests out of 112: 14.2% failure.

These 16 tests are:

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-prct-htb-in-vlr-001

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-prct-htb-in-vlr-002

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-prct-htb-in-vrl-001

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-prct-htb-in-vrl-002

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-prct-vlr-in-htb-001

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-prct-vlr-in-htb-002

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-prct-vrl-in-htb-001

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-prct-vrl-in-htb-002

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-htb-in-vlr-001

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-htb-in-vlr-004

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-htb-in-vrl-001

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-htb-in-vrl-004

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-vlr-in-htb-001

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-vlr-in-htb-004

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-vrl-in-htb-001

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-vrl-in-htb-004


I believe - but I can not be sure of this - that MS-Edge14 fails these 
same 16 tests. Someone else besides me would have to confirm this (I do 
not have and do not use Windows 10).

> Firefox 53.0a1 buildID=20170103030204 passes all 112 tests.
> 
> 2-
> I will need a review so that these tests are correct. There is also 94
> reference files which may need to be checked too.
> 
> 3-
> I know that about 5% of those 112 tests may have problems with the
> test harness. Several tests requires that the initial containing block
> be taller than 500px. This may not actually happen with the test
> harness. This depends on your screen monitor resolution and size. I
> will either adjust those tests (change a few values or change the
> character data of blocks) later; I am not sure right now what's the
> best thing to do... but I am sure this issue can be solved elegantly.

I have a list of reference files that I believe may need a bit of 
tuning: a) document box height varies between a few tests and reference 
files and b) the margin bottom of last sentence collapses with margin 
bottom of body which may or may not be rendered, according to spec. 
There is a way for me to work around this and I want to do this and, if 
I recall correctly, I have done so in a few tests.


Another odd thing I found, which I need to investigate a bit more: 
several reference files using a vertical writing-mode will not generate 
a vertical scrollbar in the test harness' <object> even though document 
box height is definitely taller than the object's viewport. Once outside 
the test harness, the window display a vertical scrollbar as expected. 
So, I do not if this is normal (does not seem like it) and if it could 
be a bug. Right now, I believe this only happens to Chrome 57 ...

Eg. View the reference file of

http://test.csswg.org/harness/test/css-writing-modes-3_dev/single/sizing-orthog-vrl-in-htb-001

in both Chrome 57 and Firefox 53. Chrome 57 does not render a vertical 
scrollbar when I think it should. Once outside the test harness' 
<object> and in a normal window

http://test.csswg.org/suites/css-writing-modes-3_dev/nightly-unstable/html/reference/sizing-orthog-vrl-in-htb-001-ref.htm

then Chrome renders a vertical scrollbar. At first sight, this looks 
like a bug to me.

Gérard
-- 
Test Format Guidelines
http://testthewebforward.org/docs/test-format-guidelines.html

Test Style Guidelines
http://testthewebforward.org/docs/test-style-guidelines.html

Test Templates
http://testthewebforward.org/docs/test-templates.html

CSS Naming Guidelines
http://testthewebforward.org/docs/css-naming.html

Test Review Checklist
http://testthewebforward.org/docs/review-checklist.html

CSS Metadata
http://testthewebforward.org/docs/css-metadata.html
Received on Thursday, 5 January 2017 00:08:23 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 5 January 2017 00:08:27 UTC