Re: percentage letter-spacing

Paul Nelson (ATC) wrote:

 > What happens and what should happen might be two different things. This
 > is a problem with basing specifications on observation. I hope that we
 > can define the right behavior and then get people to fix their
 > applications to do the right thing.

Of the two methods you've mentioned so far,
   - measuring the affected text and distributing a percentage of that width
   - using a percentage of the em square

I would much prefer the second. It gives consistent results throughout
the page, and it is much easier to implement.

What troubles me is that the size won't adjust with the font's width.
In CSS, you have no idea what the resulting font is going to be, so
this is a particularly important point. If you designed the letter
spacing for a very wide font and the used font is a very narrow one,
the result will be a very different effect. Also, we can already express
a percentage of an em with the 'em' unit: this would give CSS no extra 
functionality.

IMHO percentage letter-spacing should be relative to a metric that
changes with the narrowness of the font. I don't know enough about
fonts to know if there's a font metric we should use. If there isn't,
and typographers are used to measuring in fractions of an em, then
100% should be something like [width of lowercase alphabet]/13 or
[width of space]*4, since the "standard" length of the alphabet is
apparently 13ems, and the "standard" width of a space M/4. (I would
opt for the space*4 measure; it's more i18n-friendly.)

~fantasai

Received on Monday, 13 March 2006 07:02:53 UTC