W3C home > Mailing lists > Public > www-style@w3.org > November 1999

Re: Various comments on selectors' Test Suite

From: Eric A. Meyer <emeyer@sr71.lit.cwru.edu>
Date: Mon, 22 Nov 1999 11:47:54 -0500
Message-Id: <v0421010ab45f22358639@[]>
To: www-style@w3.org
At 17:32 +0100 11/22/1999, Daniel Glazman wrote a whole lot of stuff, part
of which went something like this:

>5.4 Type selectors
>	On my browser, all the document is sans-serif so it is
>	difficult to check if the first rule on H1 is applied
>	I run NS4.7 on Sparc+X

   Oops!  I forgot to add a 'font-family' to the basic stylesheet for all
pages.  This has been corrected.

>5.5 Descendant selectors
>	Typo in the title : DescEndant
>	I suggest "Descendant combinators" instead of "Descendant selectors"
>	and more generally use "combinators" when appropriate (5.5 -> 5.7)

   For page titles, I went with reproducing the section titles in the
specification, but I'm willing to be talked out of that approach if enough
people agree with you.

>5.7 Adjacent selectors
>	I have one general comment on testing selectors that some browsers
>	don't implement : using - if it is correctly implemented - the
>	grouping mechanism and the error handling allows to build confirmed
>	tests :
>		E, F + G { color :red }
>	E and G are red if and only if the rule is not thrown away because
>	the selector's part is invalid...

   So, in other words, if some relevant elements are red, then the browser
is parsing the selector side correctly, thus testing a browser's
forward-compatible parsing with respect to selectors.  Right?  I just want
to make sure I understand where you're going with this.

>5.8.1 Attribute Selectors
>	A[HREF="http://www.w3.org/"] {color: maroon;}
>	I suggest testing some other element/attribute pair...

   ...as in something like this?

     A[HREF="http://www.erehwon.zzz/"] {color: maroon;}

>5.8.3 Class selectors
>	The first rule needs the implementation of the universal selector.
>	For instance NS4.7 on my platform correctly implements class
>	selectors but not "*". So your test is false but it should be true...
>	Suggestion : test only with a real type element selector

   A good point, but at the same time, I'm trying to reproduce the examples
in the specification as much as possible.  What do other people think:  is
this sort of cross-contamination of CSS2 features undesirable?
   Otherwise, good suggestions with which I agree.  Look for changes in the
next few hours.

Eric A. Meyer  -  eam3@po.cwru.edu  -  http://www.cwru.edu/home/eam3.html
 Editor, Style Sheets Reference Guide  http://style.webreview.com/
 Coordinator, W3C CSS Test Suite       http://www.w3.org/Style/CSS/Test/
 Member, WSP CSS Technical Committee   http://www.webstandards.org/
Received on Monday, 22 November 1999 11:51:12 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:26:52 UTC