W3C home > Mailing lists > Public > www-style@w3.org > December 2014

Re: [css-align] Combining 'stretch' with 'true'/'safe'

From: fantasai <fantasai.lists@inkedblade.net>
Date: Wed, 17 Dec 2014 09:08:30 -0800
Message-ID: <5491B88E.8090301@inkedblade.net>
To: Javier Fernandez <jfernandez@igalia.com>, "www-style@w3.org >> www-style list" <www-style@w3.org>
On 12/17/2014 02:34 AM, Javier Fernandez wrote:
> Hi,
>
> Thanks for taking care of this and resolve all the pending issues we had
> about alignment and grid layout. I'll give my opinion below.
>
> On 12/17/2014 03:21 AM, fantasai wrote:
>> Tab and I ran into a point of disagreement this afternoon with some of the
>> edits he checked in a few months ago.
>>
>> Previously the syntax of align/justify-self was
>>
>>    auto | stretch | <baseline-position> | [ <item-position> &&
>> <overflow-position>? ]
>>
>> where
>>
>>    <item-position> = center | start | end | self-start | self-end |
>> flex-start | flex-end | left | right;
>>    <overflow-position> = true | safe
>
> Actually there were several changes on this regard. The 'stretch' value
> has been part of the <item-position>  set because it was a different
> concept from the one defined for <content-distribution>. however, the
> syntax had the intention of avoiding the combination of 'stretch' and
> <overflow-position>.
>
> I asked explicitly to get back the definition of 'stretch' for items
> because it was a lot of sense to have such behavior. We could perhaps
> tweak the syntax as it's suggested below, but we should definitively
> have different 'stretch' concepts for items and content.

We updated the wording defining the <content-distribution> keywords,
so the wording should actually now work for both cases.

>> So we'd like feedback on this issue:
>>
>>    Option A: Allow 'true'/'self' in combination with 'stretch'.
>>    Rationale: This is meaningless, consistent with how combining these
>> keywords with 'start' is meaningless.
>
> I think this option is the clearest one; the case of 'start' is exactly
> the same as 'stretch' regarding the <overflow-position> keyword.
>
>>
>>    Option B: Disallow 'true'/'self' in combination with 'stretch'.
>>    Rationale: This is disallowed, consistent with how combining these
>> keywords with the <content-distribution values (space-between |
>> space-around | space-evenly | stretch) are disallowed in 'align-content'.
>>
>
> The key is that for Content Distribution alignment we have perfectly
> separated the distribution keywords from the positional ones. That's why
> a syntax disallowing the combination with <overflow-position> is clear
> and makes sense.
>
> If we want to have a similar syntax for item positioning, we could
> perhaps define a new <item-distribution> keyword, which would have
> obviously 'stretch', but also some other <content-distribution> values.
> This would allow to define the distribution of several items placed in
> the same grid cell, for instance.

Since it's logical to at some point add fallback alignment for 'stretch',
I don't think it makes sense to allow the combination with 'stretch',
but instead to allow the combination with the fallback alignment, once
it is added [whether in this level or the next].

~fantasai
Received on Wednesday, 17 December 2014 17:09:11 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:08:49 UTC