Re: first-word pseudo-element

Peter S. Linss wrote:

>And that answer always bothered me. I accept that a UA can't always determine
>what a word is (or that the selected element may not even contain words), but so
>what? Why can't the pseudo element simply be defined to only apply to those
>elements where the UA can determine what a word is? (and maybe define a word in
>the cases where it can be defined, or at least make a note that UAs should be
>careful about what languages they apply this selector to, so that they don't
>just look for spaces in Japanese text, for instance)
>
Hi Peter,

Let's consider a Dutch city name beginning with s' (s'Gravenzande for 
instance) in a french text. I don't want, as a web author, to see one 
browser displaying the name correctly because its dictionary and 
grammatical reference is ok, and another one displaying correctly only 
the s because it considers it is a french sentence (like in "Il s'est 
cogné") !
And that is an easy case to solve...
I am also thinking of japanese combinations of chinese ideograms 
composing only one japanese word but perhaps two chinese words.

>Not all of CSS makes sense in all circumstances, it's unnecessarily limiting to
>try to pretend that it has to. What does the ::first-line selector select in an
>audio-only presentation?
>
Nothing. It should only be applicable in visual environments and perhaps 
tactile too. We miss that constraint. Probably an errata. Thanks for the 
hint.

But I see your point and can explain the difference : it is a media 
difference, not a language or writing system difference. CSS should make 
no difference at all between languages and writing systems. All 
text/visual properties should apply to all languages and writing system. 
I don't think that our actual definition of 'word-spacing' satisfies to 
this criterion.

</Daniel>

Received on Friday, 18 May 2001 04:10:57 UTC