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

Re: dependencies in tests

From: Gérard Talbot <css21testsuite@gtalbot.org>
Date: Wed, 17 Jun 2015 16:46:16 -0400
To: Florian Rivoal <florian@rivoal.net>
Cc: Public CSS Test suite mailing list <public-css-testsuite@w3.org>
Message-ID: <ed2ef819c12984cd7f074889edb641e0@gtalbot.org>
Le 2015-06-17 02:47, Florian Rivoal a écrit :
>> On 17 Jun 2015, at 05:37, Gérard Talbot <css21testsuite@gtalbot.org> 
>> wrote:
>> If you can describe the vertical-text cursor with simple words, then 
>> it would be sufficient.
> I do describe it with words.
> However, I am not testing the 'vertical-text' cursor. I am testing the
> 'text' cursor, which the UA is allowed to rotate when placed over
> vertical text.
>> What I often do is indicate that the glyphs of the words (say, "Text 
>> sample") tested should be rotated 90° clockwise (or translated 
>> vertically if 'text-orientation: upright' is being tested) and that 
>> the cursor should look like a short horizontal bar. By writing all 
>> that, you are also verifying that the user agent supports vertical 
>> writing-mode.
> But I am not interested in verifying that the UA supports vertical
> mode. Cursor and writing modes can interact, and how they behave when
> they do is defined, but cursor does not have a dependency on writing
> modes.
>  - Florian

may 	Behavior tested is preferred but OPTIONAL. [RFC2119]

should 	Behavior tested is RECOMMENDED, but not REQUIRED. [RFC2119]

An UA is not required to pass tests that are testing a specification 
sentence with a "may" or "should" in order to claim conformance with the 
spec. My understanding.

Test Format Guidelines

Test Style Guidelines

Test Templates

CSS Naming Guidelines

Test Review Checklist

CSS Metadata
Received on Wednesday, 17 June 2015 20:46:47 UTC

This archive was generated by hypermail 2.4.0 : Friday, 20 January 2023 19:58:21 UTC