W3C home > Mailing lists > Public > www-international@w3.org > October to December 2010

RE: [css3-ruby] [css3-text] Position values and before/after definitions in LR vertical writing mode

From: Stephen Zilles <szilles@adobe.com>
Date: Mon, 4 Oct 2010 10:44:10 -0700
To: Ishii Koji <kojiishi@gluesoft.co.jp>, Alan Gresley <alan@css-class.com>
CC: Richard Ishida <ishida@w3.org>, "'fantasai'" <fantasai.lists@inkedblade.net>, "www-style@w3.org" <www-style@w3.org>, "'WWW International'" <www-international@w3.org>, "btmnk0825@gmail.com" <btmnk0825@gmail.com>
Message-ID: <CE2F61DA5FA23945A4EA99A212B157952ABFD5F3EB@nambx03.corp.adobe.com>
Koji-san,
  The point that I was making in an earlier message is that (at least traditionally) vertical Classical Mongolian script is a lefthand rotation of a semitic RTL script. That would mean that the "underline" position is on the right (as is the "after" edge.) I do not know, however, how Mongolian readers currently view this. I did distribute an example which shows an emphasis line on the right. What I do not know is whether a Mongolian reader would see that as an "underline", an "overline" or would find neither term to be very helpful. I believe you were going to check that with a Mongolian expert. It is certainly an "underline" in the original semitic horizontal RTL script.l

So, in short, no I do not yet agree with you table for the reasons above.

Steve Zilles


> -----Original Message-----
> From: Ishii Koji [mailto:kojiishi@gluesoft.co.jp]
> Sent: Sunday, October 03, 2010 2:39 AM
> To: Alan Gresley; Stephen Zilles
> Cc: Richard Ishida; 'fantasai'; www-style@w3.org; 'WWW International';
> btmnk0825@gmail.com
> Subject: RE: [css3-ruby] [css3-text] Position values and before/after
> definitions in LR vertical writing mode
> 
> I made a table to ensure we're on the same page.
> 
>           | after | under
> ----------+-------+------
> Japanese  | left  | left
> Chinese   | left  | left
> Mongolian | right | left
> 
> "after" is based on block progression. It indicates different direction
> because Mongolian has different block progression as you know.
> 
> "under" of underline-position is based on how the UA renders horizontal
> scripts like English in vertical text flow. You could also say "under" is
> the "descent" side of the baseline in font terminologies. It is the same
> for all the three languages because all of them rotates the baseline
> clockwise.
> 
> So the question really is which terminologies do we want for ruby/emphasis
> marks positions.
> 
> Is this understandable?
> 
> 
> -----Original Message-----
> From: www-international-request@w3.org [mailto:www-international-
> request@w3.org] On Behalf Of Ishii Koji
> Sent: Saturday, October 02, 2010 11:21 PM
> To: Alan Gresley
> Cc: Stephen Zilles; Richard Ishida; 'fantasai'; www-style@w3.org; 'WWW
> International'; btmnk0825@gmail.com
> Subject: RE: [css3-ruby] [css3-text] Position values and before/after
> definitions in LR vertical writing mode
> 
> This isn't logical and I think I should have explained this better, but I
> didn't:
> 
> > You can ask any East Asian developers and I'm pretty sure that most of
> > them would tell you that "underline" is mapped "left", and "overline"
> > is mapped to "right" in vertical text flow.
> 
> So allow me to try again.
> 
> Assuming we're under the common understanding that we need single word that
> means one direction in horizontal flow and different direction in vertical
> flow, why "under=left", not "under=right"?
> 
> It comes from when we need to render horizontal scripts (like English)
> within the vertical text flow. In that case, we rotate the baseline
> clockwise by 90 degree and render the script. And in that case,
> "under=left" makes perfect sense.
> 
> As you pointed out, it may not make sense when you render regular
> Japanese/Chinese characters which is upright, but as written below, we
> don't want to call it "side" anyway, so there's no appropriate word for
> that.
> 
> Given these two conditions, the logical thinking would give us that
> "under=left" is the best compromise.
> 
> Back to the Mongolian case. As said, and also as Stephen Zilles pointed
> out, "before=over" and "after=under" in Japanese and Chinese, so we didn't
> see any issues and didn't care much until we think about Mongolian.
> 
> But in Mongolian, if we define under|over in the sense above, it becomes
> opposite, and now we found the need to distinguish between block
> progression and the side of baseline.
> 
> This is why I think we need under|over to describe directions against
> baseline. I hope this makes sense, but I'm more than happy to discuss
> further if anything is still unclear, or if I seem logically incorrect.
> 
> 
> Regards,
> Koji Ishii
> 
> -----Original Message-----
> From: Ishii Koji
> Sent: Saturday, October 02, 2010 1:58 AM
> To: 'Alan Gresley'
> Cc: Stephen Zilles; Richard Ishida; 'fantasai'; www-style@w3.org; 'WWW
> International'; btmnk0825@gmail.com
> Subject: RE: [css3-ruby] [css3-text] Position values and before/after
> definitions in LR vertical writing mode
> 
> Hi Alan, thank you for your reply.
> 
> So you think that we need new pair of words. I'm glad to hear that.
> 
> > So the underline is on the right side and means 'side-line'.
> > Instead of before/after or over/under, have you thought of side or
> > side-line (like in along-side or be-side) for emphasis?
> 
> This is a great question. Yes, this is one possibility.
> 
> There's one more thing, however, we need to consider. Text flow direction
> for East Asian is a formatting property. In word processors, it's under
> "Format" menu. We may change it multiple times during the edit process just
> like we change fonts. That said, by changing the direction, side-line is
> automatically changed to underline, or vice-versa. I'm not familiar with
> RTL languages, but I think this is the biggest difference between East
> Asian vertical and RTL. Vertical is a formatting property and is likely to
> be switched multiple times during the edits, while RTL is more fundamental
> script property (I hope I understand this correctly for RTL :)
> 
> This is the fundamental reason I'd like these naming be "logical". When
> switching the direction, we don't want to go through the CSS and replace
> all "side-line" to "underline". We need single word that represents one
> direction in horizontal flow, and another direction that is most
> appropriate to be mapped in vertical flow.
> 
> So, "left-side-line" is fine with me if you're fine to call underline as
> "left-side-line" in horizontal text flow as well, but clearly you don't
> want to do this, and I don't want to do this either.
> 
> If we knew this years ago, this is a much easier problem. We can discuss
> and come up with the best words that describe behavior in either text flow.
> The challenge comes when we found the need after we have defined a word for
> horizontal flow. In the case of underline, we have already used under|over
> for underline, and we then found the need to put on either side in vertical
> text flow.
> 
> One option is to stop using under|over, and invent a new pair of words. But
> it'll break the backward compatibility. So, as a compromise to keep
> compatibility, we tend to think that "this (left or right) is the direction
> to call 'under' when flow is switched to vertical". It may not be optimally
> make sense as you said, but somewhat understandable for us, and works
> consistent when you switch the direction.
> 
> You may not believe this, but East Asian developers are used to deal with
> this issue for more than 20 years since the invention of word processors,
> so we don't feel as weird as you might feel to use "under" meaning "left"
> in vertical text flow. You can ask any East Asian developers and I'm pretty
> sure that most of them would tell you that "underline" is mapped "left",
> and "overline" is mapped to "right" in vertical text flow.
> 
> There's slight weirdness left, you're right about that, but the benefit to
> use single word wins over it. I hope you'll understand that, when you
> change one property (text flow), and if you have to go through all your CSS
> files and have to replace all "under" to "left-side", you'll feel something
> is broken, right?
> 
> So, assuming the use of logical word is understood, I guess we don't have
> much options:
> 1. Since under|over is already done, keep the same rule in everywhere.
> 2. under|over is considered as an unfixable bug, just leave it and stop
> spreading it. Invent a new one.
> 
> Uh...I can think of only these two. Within these two, I'd vote 1, since
> under the constrain to use single word, it's quite difficult to completely
> feel natural. Whatever word we choose, there will be some compromises. If
> that's the case, I'd rather choose to remember just one mapping than two,
> and be consistent within CSS. It's also consistent with what we've been
> developing for 20 years.
> 
> Does this make sense? If my English is too poor to understand, or if I
> skipped something important for you to understand, please feel free to ask
> further. I'm more than happy to explain these things.
> 
> 
> Regards,
> Koji Ishii
> 
> -----Original Message-----
> From: Alan Gresley [mailto:alan@css-class.com]
> Sent: Friday, October 01, 2010 6:00 PM
> To: Ishii Koji
> Cc: Stephen Zilles; Richard Ishida; 'fantasai'; www-style@w3.org; 'WWW
> International'; btmnk0825@gmail.com
> Subject: Re: [css3-ruby] [css3-text] Position values and before/after
> definitions in LR vertical writing mode
> 
> Ishii Koji wrote:
> > I understand this:
> [snip]
> >>        d. Before/After make sense in both vertical and horizontal
> writing-modes. Over/under does not make (obvious) sense in a vertical
> writing-mode.
> >
> > Well, I think "make sense or not" depends on who you ask, and is a little
> weak to do logical discussions, so let's remove this from both.
> 
> 
> Before and after only makes sense in Latin if we have this.
> 
>    The quick brown
>    fox jumped over
>    the lazy fox.
> 
> The examples of Mongolian that I have seen (the attached file from SZ and
> this one [1]) shows block progression flowing from LTR so in Latin, it
> would be like this.
> 
>    the lazy fox.
>    fox jumped over
>    The quick brown
> 
> 
> Does before/after over/under make sense? Not really from my perspective.
> 
> 
> > Given pros you listed up, thank you for writing this, I think there are
> two points to discuss.
> >
> > First, there's a concept we did not recognize until now. And I guess you
> agreed with it in your point (a). So it's a balance between the complexity
> and the importance of old Mongolian. We all know that current spec is
> enough for all languages except old Mongolian. "do we want to add one more
> level of abstraction just for old Mongolian, or it's just not enough to do
> that" is probably the question we'd like to discuss.
> 
> Koji, in an earlier message in this thread [2] you wrote this:
> 
>   | The same thing actually happened for Japanese and Chinese.
>   | Underline in Japanese vertical writing is drawn on right
>   | as you might know. In Japanese, it's called "傍線", which
>   | means "side-line", so neither "under" nor "over" is the
>   | correct translations. We chose to name it "overline",
>   | because "over" is correct if you look at alphabet
>   | orientations in vertical text.
> 
> So the underline is on the right side and means 'side-line'. Instead of
> before/after or over/under, have you thought of side or side-line (like in
> along-side or be-side) for emphasis?
> 
> 
> [snip]
> > Regards,
> > Koji Ishii
> 
> 
> 1.
> <http://fantasai.inkedblade.net/style/discuss/vertical-
> text/diagrams/mongolian-lr.jpg>
> 2. <http://lists.w3.org/Archives/Public/www-style/2010Oct/0001.html>
> 
> --
> Alan http://css-class.com/
> 
> Armies Cannot Stop An Idea Whose Time Has Come. - Victor Hugo
Received on Monday, 4 October 2010 17:50:38 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 4 October 2010 17:50:39 GMT