Re: [css-line-grid][css-writing-modes] line-grid:match-parent and orthogonal flows

> On 08 Dec 2015, at 20:13, Alan Stearns <stearns@adobe.com> wrote:
> 
> On 10/31/15, 11:01 AM, "Florian Rivoal" <florian@rivoal.net> wrote:
> 
>> https://drafts.csswg.org/css-line-grid/#valdef-line-grid-match-parent
>> 
>>> line-grid: match-parent
>>> Box assumes the line grid of its parent.
>> 
>> The spec needs to say what happens if the parent has a different writing mode.
>> Presumably, in that situation, "line-grid: match-parent" should do the same
>> as "line-grid: create", but that should be made explicit.
> 
> I’ve changed the definition to:
> 
> Box assumes the line grid of its parent 
>    if its writing mode is the same as its parent. 
>    If the box has a writing mode that is different than its parent, 
>    then the box creates a new line grid as create below. 

Thanks.

You wrote "writing mode" in prose, depending on how you read that sentence and the writing modes spec, I think this could either be interpreted as:

a - vertical vs horizontal, which means that vertical-rl and vertical-lr (both regardless of the text-orientation property), sideways-rl, and sideway-lr can share a grid.

While that's possible for the first 3, I don't think it actually works for sideways-lr, since the line is mirrored compared to the others.

c - comparing the computed value of the writing-mode property

That would work, but I'm not sure why we'd force a separate grid on vertical-rl+upright and sideways-rl if we don't on vertical-rl+upright and vertical-rl+sideways

b - comparing the triplet of computed values from these 3 properties: writing-mode, text-orientation, direction

Direction is part of the properties listed by the writing modes spec as influencing the "writing mode", but I'm not sure why it would matter here.

d - something else?

So:

1) I am misunderstanding this?

2) Is there an other term we could use that would group like this:
[
  [horizontal],
  [vertical-rl, vertical-lr,  sideways-rl],
  [sideways-lr]
]

3) Should we use "computed value of the writing-mode property", which would make all five independent

4) Should we depend on the computed value of writing-mode and text-orientation?

5) should we depend on the computed value of writing-mode text-orientation and direction?

I'd probably prefer 2, although if we ever want to be able to reintroduce text-orientation:sideways-left (aka text-orientation:sideways-flipped), we may find 4 or 5 to be a more conformable avenue, unless the term we end up with in 2 also differenciate vertical-? from vertical-?+sideways-left and groups the later with writing-mode:sideways-left.

 - Florian

Received on Tuesday, 8 December 2015 13:55:24 UTC