Re: [CSS21] text-decoration/visibility

On Thu, 12 Nov 2009 18:38:18 +0100, Tab Atkins Jr. <jackalmage@gmail.com>  
wrote:

> On Thu, Nov 12, 2009 at 11:28 AM, Øyvind Stenhaug <oyvinds@opera.com>  
> wrote:

>> - It doesn't say that the decorations are propagated when specified on  
>> an
>> inline (unless I'm misunderstanding the meaning of "box generated by  
>> that
>> element" and/or "text of an element")
>
> They shouldn't be propagated (in fact, there's a note that doing so is
> a mistake), but they should still be applied to the entire inline,
> descendants and all.

Right, I meant the cases without floats etc., such as an otherwise-unstyled
<span style="text-decoration:underline">A<span>B</span></span>
where I assume (haven't found anything explicit in the spec) that the  
first span generates a single inline box that contains A but not B's box.  
For "other elements" there's wording about an anonymous inline box, but  
not for "inline element"s (whatever that means; display:inline?).

>> - It's not very obvious that visibility:hidden on an element  
>> (apparently)
>> applies to decorations specified on that element (making them invisible,
>> also when propagating) but not to propagated decorations specified on an
>> ancestor
>
> Yeah, that's not clear.  I'm not sure where that's being gotten from.
>
>> - As for decoration specified on a given element being ignored if the  
>> same
>> type (e.g. underline, line-through) is propagated from an ancestor, is  
>> that
>> what "cannot have any effect on the decoration of the ancestor" is  
>> meant to
>> say? If so, maybe it would be clearer if, say, "of" were to be replaced  
>> with
>> "propagated from".
>
> Hmm, I assumed the reverse - that it simply wouldn't reach up and
> change the ancestor's decoration.  But your reading makes more sense.

I always read it that way too when skimming this section in the past, but  
yes, it didn't make much sense.

-- 
Øyvind Stenhaug
Core Norway, Opera Software ASA

Received on Thursday, 12 November 2009 18:13:58 UTC