W3C home > Mailing lists > Public > www-style@w3.org > September 2015

Re: [css-writing-modes] computed value for text-orientation: sideways or sideways-right

From: 馬場孝夫 <baba@bpsinc.jp>
Date: Fri, 18 Sep 2015 10:49:40 +0900
Message-ID: <CAAWjb-e3TJUkaMBUYZMyV61T=p=f7CARwE2UhicFjMOCTwTBYQ@mail.gmail.com>
To: fantasai <fantasai.lists@inkedblade.net>
Cc: Gérard Talbot <www-style@gtalbot.org>, Jonathan Kew <jfkthame@gmail.com>, W3C www-style mailing list <www-style@w3.org>
Hi Jonathan, Gerard, fantasai.

# This is my first time to sending www-style. Nice to e-meet you.
# I'm Baba, a colleague of Hiroshi (skk@bpsinc.jp).

I roughly agree to Gerard.

In my opinion 'text-orientation: sideways-left' might be needed in the future,
because of the following reason:

Since an inline box which has a different 'writing-mode' property to its parent
becomes inline-block, it is impossible to align original baseline between
clockwise-rotated texts and counter-clockwise-rotated texts. Mostly this is not
required because mixing them in a same line is rare, but is possibly needed in
the future. So we propose to disuccuss it in future level.

On the other hand, as fantasai said, a shorter name is more useful. Therefore,
how about the definition like below?

    'sideways': For 'writing-mode: vertical-rl', alias for 'sideways-right'.
                For other values, UA-dependent. NOTE: the future level of CSS
                may specify the detail of the behavior.

Simply I want to say that sideways-left might be required in the near future, I
want to avoid to add conflicting values to it for current level.

Baba
----------------------------------------------------
ビヨンド・パースペクティブ・ソリューションズ株式会社
〒160-0023
東京都新宿区西新宿6-20-7 コンシェリア西新宿TOWER'S WEST 2F
Tel: 03-6279-4320 Fax: 03-6279-4450
http://www.bpsinc.jp
馬場 孝夫(Baba Takao)


On Fri, Sep 18, 2015 at 9:29 AM, fantasai <fantasai.lists@inkedblade.net> wrote:
> On 09/17/2015 05:42 PM, Gérard Talbot wrote:
>>
>> Le 2015-09-17 06:51, Jonathan Kew a écrit :
>>>
>>> The recently-updated draft[1] says:
>>>
>>> # Name:           text-orientation
>>> # Value:          mixed | upright | sideways-right | sideways
>>> # ...
>>> # Computed value: specified value
>>>
>>> Given that 'sideways' and 'sideways-right' are defined to be
>>> equivalent, could/should we change this such that 'sideways-right'
>>> computes to 'sideways' (or possibly vice versa)? Then the UA can treat
>>> this as a simple alias during parsing, and avoid the need to keep
>>> track of the specified value just so that getComputedStyle() can
>>> return it correctly.
>>>
>>> Which of 'sideways' or 'sideways-right' do we want to promote as the
>>> "canonical" computed value, and which should be an alias? (My vote
>>> goes to the simpler 'sideways', fwiw.)
>>>
>>> JK
>>>
>>> [1]
>>> https://drafts.csswg.org/css-writing-modes-3/#propdef-text-orientation
>>
>>
>> I believe that 'sideways' should just be removed. Just my opinion.
>>
>> If, one day, 'sideways-left' is reintroduced, then we should also
>> reintroduce 'sideways' like in the original intent of the spec,
>> meaning :
>>
>> "'sideways' value is equivalent to sideways-right in vertical-rl
>> writing mode and equivalent to sideways-left in vertical-lr
>> writing mode."
>
>
> Since the use cases for the automatic rotation are mainly for
> horizontal scripts, and these will be using the new 'sideways-lr'
> and 'sideways-rl' values of 'writing-mode', there might not be
> much need for such a value in the future. It seems to me it would
> be better to optimize for the CJK users who will need to use the
> 'text-orientation' property with much more frequency, and so it
> makes sense to use a shorter name.
>
> The alternative, which may still be possible depending on the
> status of implementations and Web content, would be to drop
> both values and use 'sideways-rl' as a keyword:
>
>   text-orientation: upright | mixed | sideways-rl
>
> (EPUB content should be using -epub- prefixed syntax, so would
> not be affected.)
>
>> By the way, Example 8 in
>> https://drafts.csswg.org/css-writing-modes-3/#text-orientation
>> no longer makes perfect sense now that sideways-left has been removed.
>
>
> Removed. Thanks~
> ~fantasai
>
Received on Friday, 18 September 2015 01:50:43 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:08:57 UTC