Re: [css3-writing-modes] before/after terminology alternative?

fantasai <> wrote on 2012/06/21 10:40:44
> On 06/01/2012 06:55 AM, MURAKAMI Shinyu wrote:
> > fantasai<>  wrote on 2012/05/30 1:10:43
> >> On 05/29/2012 08:56 AM, Sylvain Galineau wrote:
> >>>
> >>> [L. David Baron:]
> >>>>
> >>>> Are we sure 'head' / 'foot' are actually writing-mode-independent terms,
> >>>> as opposed to effectively being terms for 'top' / 'bottom'?
> >>>>
> >>> Right; I think the assumption here is that web authors will see head/foot and
> >>> think header/footer. Fwiw I think that is a reasonable assumption [...]
> >>
> >> Yes, that was exactly my first thought!
> >
> >
> > Unfortunately, head/foot as alternative terminology of before/after are inconsistent with Japanese layout terminology in JLREQ[1]:
> >
> > Terminology Japanese Romanized transliteration Definition
> >
> > head 天 ten a) The top part of a book or a page.
> >    b) The top margin between the top edge of a trimmed page and the hanmen (text area)
> >    (JIS Z 8125)
> >
> > foot 地 chi a) The bottom part of a book or a page.
> >    b) The bottom margin between the edge of a trimmed page and the hanmen (text area)
> >    (JIS Z 8125)
> I kindof like "sky" and "ground" for the top and bottom of the book. :)
> It sounds very poetic.
> > headnote 頭注 tōchū A kind of notes in vertical writing style, head area in kihon-hanmen is kept beforehand, and notes are set with smaller size font than main text.
> >
> > footnote 脚注 kyakuchū A note in a smaller face than that of main text, placed at the bottom of a page. (JIS Z 8125)
> Hm, these ones might be a concern.
> But then what about the header and footer of a table? We have <thead> and <tfoot>,
> and those respond to writing mode.

Yes, head/foot can be both physical and writing mode relative directions.
But in Japanese layout terminology 頭(head)/脚(foot) are usually 
physical top and bottom, or 行頭(line-head) means the start of a line.

> If a book uses 頭書 to refer to something in vertical writing, where is it
> in relation to the text? Or is a different term used for vertical writing
> vs. horizontal writing?

頭(head) means also "start" generally, 行頭(line-head, 
the start of line) is an example.
頭書(above-mentioned) can mean the start of a document in 
both writing modes, but it is not a layout terminology and 
not in JLREQ.

> > line head 行頭 gyōtō The position at which a line starts. (JIS Z 8125)
> > line head alignment 行頭そろえ gyōtō soroe To align a run of text to the line head. (JIS Z 8125)
> > line head indent 字下げ jisage To reserve a certain amount of space after the default position of a line head. (JIS Z 8125)
> This seems like it would change which side is affected depending on the writing mode, no?
> In horizontal writing it's not the same side as the "head" of the book, it's 90deg
> counter-clockwise from that.

Yes, Japanese line layout terminology is based on vertical writing mode, 
as in the term 字下げ(jisage, literally "character-down") 
and 行頭(line head) is the top side physically in vertical writing mode
(and 行頭 is also understood as "the start of line").

> > running head 柱 hashira A page element which contains information on the title of the book, chapter, section and so on, printed outside the area of the hanmen. (JIS Z 8125)
> "Running head" is short for "Running header", so I think it's okay here.
> > [1]
> >
> > In JLREQ, head/foot mean basically physical top/bottom and the 'line head' (行頭) means line-start.
> > The headnote (頭注) is positioned in physical top area in vertical writing style.
> > The footnote (脚注) is positioned in physical bottom area in both vertical/horizontal writing styles.
> > (Note: 頭=head, 脚=foot)
> >
> > Are you ok if CSS spec terminology and Japanese layout terminology are so inconsistent?
> I'm not too concerned if JLREQ is inconsistent, I'm a little more concerned
> if Japanese itself is inconsistent.

I don't think JLREQ is inconsistent. I think JLREQ terminology, 
both Japanese and English, are clear and consistent.
It never uses 頭(head)/脚(foot) for writing mode relative block directions.
Section "4.4 Processing of Tables" of JLREQ uses "header" for table headers
(for header column and header row) but in the Japanese version of JLREQ,
"ヘッダー" (transliteration of "header") is used instead of Japanese word.

> > Personally, I think before/after are better.
> But I'm pretty sure that before/after confuses many people. :/

I'm pretty sure that head/foot confuses people who learn JLREQ and CSS Writing Modes.
It makes it difficult to explain the CSS Writing Modes specification 
to Japanese people. The "before side" and "after side" can be translated 
into Japanese literally "前側" and "後側" (or "前方" and "後方") and 
they are easily understood. But "head side" and "foot side"
("頭側" and "脚側" if literally translated) are hardly understood 
and easily confused with physical top/bottom sides or logical start/end of line.

There are already Japanese tutorial documents explaining 
logical directions, start/end/before/after.
EBPAJ (The Electronic Book Publishers Association of Japan)'s
EPUB3 Authoring Guide (in Japanese) is an example.
It has the following description:
 行頭  :start (縦:top    横:left)
 行末  :end   (縦:bottom 横:right)
 行の前方:before(縦:right  横:top)
 行の後方:after (縦:left   横:bottom)

Here, "行の前方" means "before the line" and "行の後方" means "after the line".
If "before/after" is changed to "head/foot", Japanese explanation becomes difficult.

As the above description with Japanese, the logical directions 
start/end/before/after can be easily understood with the four 
sides of a line box, the start of the line, the end of the line, 
before the line, and after the line. Why many people confuse?

I don't think the logocal direction before/after conflicts 
with existing CSS specification, the pseudo elements ::before and ::after.
The pseudo elements ::before and ::after are for 
"before the element's content" and "after the element's content" 
in the DOM tree, and do not mean directions in layout.
People can easily distinguish them.

We, Antenna House, support both XSL-FO and CSS in our product, AH Formatter.
Many XSL-FO properties are derived form CSS2 and they use similar terminology.
If CSS uses start/end/head/foot while XSL-FO uses start/end/before/after,
it becomes very inconvenient for people who use both XSL-FO and CSS.


Shinyu Murakami
Antenna House

Received on Friday, 21 September 2012 07:56:34 UTC