- From: John Daggett <jdaggett@mozilla.com>
- Date: Mon, 7 Nov 2011 18:18:03 -0800 (PST)
- To: Koji Ishii <kojiishi@gluesoft.co.jp>
- Cc: www-style@w3.org
Koji Ishii wrote: > OpenType spec defines how to calculate Ideographic Em-Box[2]. > Historically speaking, East Asian font designers used to use > head.unitsPerEm to store this information, so the logic described in > the spec falls back to head.unitsPerEm if the information is missing. > Authors use "em" units for this purpose today for the same historical > reasons, which works for such fonts, but it can be different. > > What we'd like to propose is to make inline-direction length of the > Ideographic Em-Box as a unit, so that authors can specify the number > of Ideographic Em-Box for lengths such as box width, margins, column > width, and column gaps, and so that the value works as authors expect > for any fonts. Isn't what you're describing something more akin to 'ideo-ch', the advance of a full-width character in the inline direction? Note that 'em', 'ex', 'ch' and 'rem' are all defined relative to the font-size so you need to define something in a similar fashion. I think what you're trying to define is the ratio of the width/height of the ideographic em-box to the width/height of the em-box itself in the inline direction, multiplied by the font-size. What's the use case for this unit? In other words, what are the cases that using em-units won't solve? Seems like the only situation is compressed or expanded fonts. Whatever definition you come up with, I think you need to test and verify that your definition works with the actual fonts and layouts you're interested in supporting. Cheers, John Daggett
Received on Tuesday, 8 November 2011 02:18:31 UTC