W3C home > Mailing lists > Public > public-css-testsuite@w3.org > September 2015

Re: [CSS21][CSS22] Seeking review of position:fixed edge case test

From: Gérard Talbot <css21testsuite@gtalbot.org>
Date: Thu, 17 Sep 2015 01:11:56 -0400
To: Chris Rebert <csswg@rebertia.com>
Cc: Public CSS Test suite mailing list <public-css-testsuite@w3.org>, chris@rebertia.com
Message-ID: <f6bee98c3b9396e300b484eda6f91f2e@gtalbot.org>
Le 2015-09-13 00:30, Chris Rebert a écrit :
> On Mon, Aug 17, 2015 at 1:49 PM, Gérard Talbot
> <css21testsuite@gtalbot.org> wrote:
>> Le 2015-08-17 15:08, Chris Rebert a écrit :
>>> On Tue, Aug 4, 2015 at 9:06 PM, Gérard Talbot
>>> <css21testsuite@gtalbot.org> wrote:
>>>> Le 2015-07-27 03:21, Chris Rebert a écrit :
>>>>> 
>>>>> Hi folks,
>>>>> 
>>>>> I'm trying to get a reviewer for a CSS2 regression test:
>>>>> https://github.com/w3c/csswg-test/pull/813
>>> <snip>
>>>> Your test is a good test. Indeed iPhone 6, Safari 7+ fail your test. 
>>>> Here's
>>>> some review comments.
> <snip>
>>>> 7-
>>>> p {
>>>>   position: absolute;
>>>>   top: 100px;
>>>> }
>>>> 
>>>> It is recommended to start all tests with the pass-fail-conditions 
>>>> sentence
>>>> and, since it should not be part of the test itself, then we do not 
>>>> want to
>>>> style it in any way. Sometimes, this is not possible but, I'd say 
>>>> 99% of the
>>>> time, it is possible.
>>> 
>>> The position:fixed makes doing this somewhat tricky.
>>> If you have a specific suggestion for how to achieve this, I'm fine
>>> with changing the test, but I myself don't have any ideas.
>> 
>> The modified test I created made that possible.
> 
> Revised.
> 
>>>> Here's what could be your test with all these changes:
>>>> 
>>>> http://www.gtalbot.org/BrowserBugsSection/css21testsuite/left-offset-position-fixed-001.xht
> <snip>
>>>> One last thing. One other equivalent test could be created with 
>>>> 'direction:
>>>> rtl', 'position fixed', 'right: auto' for many reasons. Also, same 2 
>>>> tests
>>>> but with 'position: absolute'. So, 3 other additional tests with the 
>>>> same
>>>> basis code are possible here.
>>> 
>>> I'm not familiar enough with `direction`/RTL to write those variants,
>> 
>> http://www.gtalbot.org/BrowserBugsSection/css21testsuite/right-offset-position-fixed-001.xht
>> 
>> iPhone 6, Safari 7+ fail this additional test.
>> The text assert and a few other things in that 
>> right-offset-position-fixed-001 would need to be adjusted, tweaked.
> 
> Added the RTL and position:absolute tests you suggested.
> 
> Everything look good now?:
> https://github.com/w3c/csswg-test/pull/813/files
> 
> Thanks,
> Chris


Chris,

Sorry for the long delay.

I also had all kinds of difficulties with your tests, GitHub, filenames, 
browsers, etc.

As strange as it may seem, it's possible to save a file under a filename 
that has a white blank space (U+0020) appended at the end of the file 
extension: this is news to me! In such case, Firefox 40.0.3 will load 
such file but Chrome 45.0.2454.93 will not and will only say that the 
file is not found "ERR_FILE_NOT_FOUND" but that's because the address 
bar, it seems, trims out blank spaces. I think the text editor I used is 
the software with a bug: it should trim out the blank space at the end.

1-
With GitHub, I have to find the files, then I have to copy and paste the 
code into a text editor, then remove the "+" [addendum: I just noticed 
the View button for viewing the whole file without "+"], then save the 
code under suitable filename(s) and then load the tests in browsers for 
examination. There seems no other way to review tests with GitHub.

2-
The main difference between absolute positioning and fixed positioning 
is what's a box's containing block. Section 10.1 states:

"
If the element has 'position: fixed', the containing block is 
established by the viewport (...)

If the element has 'position: absolute', the containing block is 
established by the nearest ancestor with a 'position' of 'absolute', 
'relative' or 'fixed', (...).
"
10.1 Definition of "containing block"
http://www.w3.org/TR/CSS21/visudet.html#containing-block-details

About abs. positioning, we often use expressions like nearest (or 
closest) positioned container block within containment hierarchy or 
nearest (or closest) positioned ancestor.

So, your text asserts need to reflect such absolute-versus-fixed 
difference and they are not.

- - - - - - - -

The 'left: auto' and 'right: auto' is given by (or, if you prefer, is 
established by) the so-called "static-position containing block". Often, 
this is the immediate (nearest, closest) parent of such box. In your 
tests, the "static-position containing block" of your div#red is your 
div#shifted-column .

- - - - - - - -

Your href="http://chrisrebert.com" leads to 404 not found ... but that's 
a minor thing.

- - - - - - - -

Chris, for many reasons, I propose you just submit your position-fixed 
tests (along with the reference files) and leave out, drop the abs-pos 
tests. I approve your
left-offset-position-fixed-001.xht
and
right-offset-position-fixed-001.xht
along with their respective reference files. Please add the date like 
this:

     <link rel="reviewer" title="Gérard Talbot" 
href="http://www.gtalbot.org/BrowserBugsSection/css21testsuite/" /> <!-- 
2015-09-17 -->


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, 17 September 2015 05:12:27 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 17 September 2015 05:12:30 UTC