W3C home > Mailing lists > Public > www-style@w3.org > January 2011

[css3-writing-modes][css3-text] Feedbacks from an informal spec review in Japan

From: Koji Ishii <kojiishi@gluesoft.co.jp>
Date: Sun, 9 Jan 2011 06:27:10 -0500
To: "www-style@w3.org" <www-style@w3.org>
Message-ID: <A592E245B36A8949BDB0A302B375FB4E0AA8CACE94@MAILR001.mail.lan>
I've got valuable feedbacks from an informal spec review for the writing modes in Japan. Here's my memo from the review.

5. Line Orientation

*1: Arabic "ABC" in Figure 7 is displayed as LTR -> probably a problem in FF; IE/Safari displays correctly.

5.1. Orienting Text: the 'text-orientation' property

*2: The description says "this property specifies the orientation of characters in a non-native 'writing-mode'", and it is correct for "vertical-right" and "upright", but "rotate-right" and "rotate-left" are different (affects native scripts as well).

*3: It's nice to have sample pictures.

6.1.1. Orthogonal Flows

*4: What happens if a vertical block within a horizontal document overflows a page in paged media? -> Probably the block should go to the next page (prohibit page break within the vertical block), but the behavior is undefined in the current spec.

*5: What happens if a vertical block within a horizontal document is higher than the page height in paged media?

*6: What happens if the width (logical height) of a vertical block within a horizontal document exceeds the page width in paged media?

*7: How the UA should determine the width (logical height) of a vertical block within a horizontal document when it's less than available width? IE7 uses 100% while IE8 uses shrink-to-fit.

*8: In above case, if shrink-to-fit, can authors specify the block aligned left/center/right? Logical thinking gives that 'margin-left:auto' should do, but it doesn't work in IE8.

*9: The expected behavior for 'auto' logical widths (marked as an issue in the spec). Although we could not reach to consensus, the discussion was valuable. I'll follow this up separately.

*10: Can author specify logical width to be shrink-to-fit behavior?

*11: Can author specify logical width to be "all the rest of available logical width" behavior? bottom:0px?

7. Glyph Composition: the 'text-combine' property

*12: It's nice to have HTML/CSS code for the sample picture.

*13: A request for the 'text-auto-combine' property (it's currently marked as an issue in the spec). A use case was given; authors using text editors would want to set the default value of the 'text-auto-combine' for the whole document, review the results, and then fix where inappropriate (one mentioned as "often context-sensitive" in the issue section of the spec). That requires far less work than applying 'text-combine' to every occurrences of numeric words. Whether we should have this in Writing Modes Level 3 or not was split; some argued we should, while some argued most users would use authoring tools and we should keep Writing Modes Level 3 as primitive as possible.

Miscellaneous: the original intention was to review both Writing Modes Level 3 and Text Level 3, but we ran out of the time. Instead of reviewing CSS3 Text, we had a short free discussion session where I got following wishes.

*14: Wish: non-uniform glyph scaling -> It could be done by SVG/2D Transform if it's a graphic rather than text layout, but we don't have a good method in text layout. Some disliked the idea of non-uniform scaling of glyphs. Some wants more, like different scaling factor for Latin/Asian characters.

*15: Wish: text to fit within a specified space using non-uniform scaling -> Same as *14?

I'll work on these issues, but any additional feedbacks/opinions are appreciated.


Regards,
Koji
Received on Sunday, 9 January 2011 11:25:22 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:35 GMT