W3C home > Mailing lists > Public > public-css-testsuite@w3.org > December 2004

Re: text-indent test

From: Ian Hickson <ian@hixie.ch>
Date: Thu, 2 Dec 2004 15:11:08 +0000 (UTC)
To: Michael Day <mikeday@yeslogic.com>
Cc: public-css-testsuite@w3.org
Message-ID: <Pine.LNX.4.61.0412021456380.11059@dhalsim.dreamhost.com>

On Thu, 2 Dec 2004, Michael Day wrote:
> 
> A test for text-indent (t1601-c547-indent-00-b-a) seems to depend on a 
> particular equivalence between pixel units and absolute units.
> 
> It states that there should be a "tall solid unbroken column of blue" in 
> the output, which is actually made of numerous indented blue boxes which 
> should all line up.
> 
> However, one of the boxes is indented 24px, while another is indented 
> 0.25in; doesn't this require a particular output device, given that 
> pixels in CSS are relative units intended to correspond to display 
> angle?

I will be adding documentation at some point to clarify the assumptions 
made by the test suite at the moment, but they include:

 * The device is a full-color device.
 * The device has a viewport width of at least 640px (approx).
 * The resolution of the device is 96 CSS pixels per inch.
 * The 'medium' font-size computes to 16px.
 * The initial value of 'color' is black.
 * The canvas background is white.
 * The Ahem font is available.
 * The user stylesheet is empty (except where indicated by the tests).
 * All other user settings are at their default values.

For devices that cannot be made to fit these conditions, a second test 
suite will probably have to be made. (This wouldn't be particularly hard.) 
I don't think assuming the above is a problem. It makes testing things a 
lot easier. Those assumptions actually are made by most of the existing 
CSS-based Web content anyway, so for most CSS UAs it's not a problem. Of  
course it can be a problem on mobile devices and in UAs not intended to be 
showing or printing typical Web content.

At some point we should probably go through and mark the tests as 
requiring some of the above assumptions to be true -- for example, the 
tests that need Ahem are marked with '-a' at the end, maybe we should add 
'-r' for tests that require 96dpi and 16px initial font-size.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Thursday, 2 December 2004 15:11:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 20 September 2010 17:51:54 GMT