W3C home > Mailing lists > Public > www-style@w3.org > March 2008

Re: [CSS3] Box Model Terminology

From: Andrei Polushin <polushin@gmail.com>
Date: Sun, 02 Mar 2008 01:45:56 +0600
Message-ID: <47C9B274.8000003@gmail.com>
To: www-style@w3.org

Ambrose Li wrote:
> Andrei Polushin wrote:
>>  European          Arabic, Hebrew   Chinese, Japanese   XSL-FO equivalent
>>  ----------------- ---------------- ------------------- ------
>>  logical-left      semitic-right    east-asian-top      before
>>  logical-right     semitic-left     east-asian-bottom   after
>>  logical-top       semitic-top      east-asian-right    start
>>  logical-bottom    semitic-bottom   east-asian-left     end
>
> While I appreciate the reasoning of the proposal, I find this to be 
> even more confusing than things are right now.

I do not propose the exact terms, but the principle of convenience. It 
should be convenient for the specification writer to use simply (left, 
right, top, bottom) throughout in the specification.

Implementers would prefer to keep (left, right, top, bottom) in their code, 
but just remap them to follow the direction of the flow. In addition, 
implementers would prefer to match their terms with the specification terms. 
That should be mostly enough to successfully implement the thing like this.

Then, international web authors might be interested to use the most 
convenient names in their stylesheets: when the stylesheet is designed for 
the Hebrew web site, it would be convenient to express the stylesheet 
relying upon the RTL coordinate system (and using the semitic-* terms).

That's all. If you are a spec writer, use "logical-left" etc. terms, and do 
not beware of terminological bugs. If you are a spec reader, enjoy reading 
it. If you are an implementer, use "logical-left" instead of left, and don't 
bother at all. And only if you are a web author, use one of the terms from 
the table above, which are most convenient for you.

> If we need writing-directon-independing wording, how about using some 
> form of "advance" ("inline direction") and "leading" ("block 
> direction")?
>
> BTW, due to western influence, I have been seeing more and more 
> left-to-right vertical text for some years already (first I saw 
> Korean, then I saw more and more Chinese). It's only a matter of time 
> you can't assume vertical CJK is right-to-left.

Once again, I propose just the principle of convenience. Would it be 
convenient, we can mention the writing mode explicitly:

     European      Arabic, Hebrew   Chinese, Japanese   XSL-FO equivalent
     ------------- ---------------- ------------------- ------
     ltr-left      rtl-right        tb-rtl-top          before
     ltr-right     rtl-left         tb-rtl-bottom       after
     ltr-top       rtl-top          tb-rtl-right        start
     ltr-bottom    rtl-bottom       tb-rtl-left         end

(Now ltr-left == logical-left from the previous table).

--
Andrei Polushin
Received on Saturday, 1 March 2008 19:46:06 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:55:02 GMT