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

Re: Wording of text-transform-bicameral-*

From: L. David Baron <dbaron@dbaron.org>
Date: Tue, 1 Feb 2011 09:35:02 -0800
To: Øyvind Stenhaug <oyvinds@opera.com>
Cc: CSS-testsuite <public-css-testsuite@w3.org>
Message-ID: <20110201173502.GA17398@pickering.dbaron.org>
On Tuesday 2011-02-01 12:41 +0100, Øyvind Stenhaug wrote:
> the text-transform-bicameral-* tests all use the wording "Test
> passes if the left-hand character in each pair matches the
> right-hand one."
> 
> Some of our testers have interpreted this in a way that would
> classify characters of the same shape as "matching", resulting in
> mis-classifications of tests like e.g. this one:
> 
> http://test.csswg.org/suites/css2.1/20110111/html4/text-transform-bicameral-022.htm
> 
> I suggest that the text be clarified further. E.g.: "Test passes if
> the left-hand character in each pair is identical to the right-hand
> one (having both the same shape and the same size)".
> 
> Another potential issue is that the characters will be exactly
> identical if no suitable font is found and a replacement character
> is rendered instead. Should probably also clarify that each pair
> should be different from all the other pairs.

I believe there are some cases in which the "each pair should be
different from all the other pairs" constraint should not be true.
For example,
http://test.csswg.org/suites/css2.1/20110111/html4/text-transform-bicameral-007.htm
contains two pairs that should look like "E E", since both "Ε ε" and
"Ε ϵ" uppercase to "Ε Ε" (and likewise for "Κ κ" and "Κ ϰ",
"Σ ς" and "Σ σ" etc.).

-David

-- 
L. David Baron                                 http://dbaron.org/
Mozilla Corporation                       http://www.mozilla.com/
Received on Tuesday, 1 February 2011 17:35:44 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 1 February 2011 17:35:51 GMT