W3C home > Mailing lists > Public > public-css-testsuite@w3.org > October 2011

[CSS21] Invalid test cases for BIDI

From: Arron Eicholz <Arron.Eicholz@microsoft.com>
Date: Wed, 26 Oct 2011 00:23:02 +0000
To: "public-css-testsuite@w3.org" <public-css-testsuite@w3.org>
Message-ID: <07349ECFC3608F48BC3B10459913E70B14935E97@TK5EX14MBXC293.redmond.corp.microsoft.com>
According to CSS 2.1 the following cases I believe are invalid.

The test expects that all the characters are evenly spaced however there is nothing I the CSS 2.1 spec that says that is the way these bidi cases should work.

The first issue I see with these cases is we explicitly say for letter-spacing that the character spacing algorithms are user agent-dependent. This means that whatever the user agent is doing for letter spacing is correct in all instances.

Second assuming that the first rule issue I raised does not apply the spacing is an inter-character spacing in addition to the default space which means that it is the space between the characters of the writing direction since you have to determine where the space character would be. This would mean that between 'a' and 'b' there should be more space and there should be no space between 'b' and 'c' because the space between the 'b' character would be on the left side because we changed direction using the RLO character. This overrides the direction and makes the 'b' go right-to-left which means the space is on the left now.

>From what I can tell from these two issues is the cases are invalid (in the case of the first issue) or incorrect (in the case of the second issue). Either way the cases need to be changed in order for them to be valid. If we say the cases are just invalid then the rendering in Firefox and IE is the correct rendering. I do not see any situation where the expected result in the case is the correct behavior per the spec definition. If I am missing some piece of this from some obscure piece of text in the spec please point me to it because right now these cases just are wrong from what have read.

http://test.csswg.org/suites/css2.1/20110323/html4/bidi-005b.htm
http://test.csswg.org/suites/css2.1/20110323/html4/bidi-006b.htm
http://test.csswg.org/suites/css2.1/20110323/html4/bidi-007b.htm
http://test.csswg.org/suites/css2.1/20110323/html4/bidi-008b.htm
http://test.csswg.org/suites/css2.1/20110323/html4/bidi-009b.htm
http://test.csswg.org/suites/css2.1/20110323/html4/bidi-010b.htm
http://test.csswg.org/suites/css2.1/20110323/html4/bidi-011.htm

--
Thanks,
Arron Eicholz
Received on Wednesday, 26 October 2011 00:23:36 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 26 October 2011 00:23:45 GMT