Re: CSS3 Text - Edit suggestions

Bert Bos wrote:
> On Monday 19 February 2007 09:30, fantasai wrote:
>> Paul Nelson (ATC) wrote:
> 
>>> 4.       text-align property, left value definition. I believe that
>>> the text for vertical is not correct, but should be defined as “In
>>> vertical text, ‘left’ is interpreted with respect to the left side
>>> of the line if the baseline was rotated to horizontal.”
>>>
>>> 5.       text-align property. right value definition. I believe
>>> that the text for vertical is not correct, but should be defined as
>>> “In vertical text, ‘right’ is interpreted with respect to the right
>>> side of the line if the baseline was rotated to horizontal.”
> 
> I think that is a good definition, because it both defines and explains 
> ("to remember this, think of horizontal and vertical lines as 90 degree 
> rotations of each other").
> 
> In both our vertical modes (rl and lr), the baseline is rotated the same 
> way, isn't it? So 'left' is always top and 'right' always bottom. Maybe 
> it is useful to add that as a note: "(I.e., text is aligned at the 
> bottom.)"

This is true by default, but it may not be true for specific values of
glyph-orientation: rotating glyphs rotates their baseline. glyph-orientation
can make glyphs stand upright in a column, but it can also turn the line
180deg to face the other way.

> I don't think glyph orientation affects text alignment. One affects 
> inline layout, the other the block.
> 
> E.g., should it make a difference whether you do
> 
>     <p style="block-progression: rl;
>               glyph-orientation-vertical: upright">
>       ...
>     </p>
> or
>     <p style="block-progression: rl">
>       <span style="glyph-orientation-vertical: upright">
>         ...
>       </span>
>     </p>?

That's like asking, should it make a difference whether you do
   <p style="direction: rtl; text-align: start"> ... </p>
or
   <p style="text-align: start"><span style="direction: rtl;"> ... </span> </p>

> I'm wondering if we shouldn't add keywords 'top' and 'bottom'. They 
> would technically be aliases for 'left' and 'right', but maybe more 
> intuitive for users.

Authors should be using 'start' and 'end' wherever possible, not 'top'
and 'bottom'. If there's a good use case for adding 'top' and 'bottom'
in addition to 'start' and 'end', then we can consider it, but 'start'
and 'end' are much less error-prone: they don't align opposite to the
author's intent if 'block-progression' gets overridden later in the
cascade.

> We have some bias to horizontal text that we can't 
> change ('line-height'), but in this case we *can* change it.

How is 'line-height' biased towards horizontal text?

~fantasai

Received on Tuesday, 20 February 2007 15:23:40 UTC