Obligatory history lesson: 'em' is a classical typographical term to talk about relative measurements within a typeface: it's the phoneticisation of the letter 'm', and denotes the total width of the 'm' glyph. 'en' is used for 'n'; other useful relative measurements include 'x-height' to denote the upper edge of 'straight' lowercase letters relative to the baseline.
In digital typography, it's a lot more vague. For one, it is no longer intended for internal measurements of the typeface, rather the relative font-size — and whether set in pixels, picas, or whatever, font-size is a little arbitrary: all measurable aspects of the glyphs 'M' or 'm' (or indeed any other glyph) set in Arial are smaller than 12px; with Calibri they are smaller still. The purpose of em-based sizing on the web is to provide a relative font-size: h1 {font-size: 2em} will make top-level headings twice as big as the default font-size. An irritating aspect is that the default font-size I mentioned is subject to other ems. If that h1 is inside a div with font-size: 2em, it will end up four times larger than the base. This is useful because certain screens have higher pixel densities (12px-sized fonts are fine on my 96ppi desktop but illegible on my 326ppi phone), and some people, particularly those with bad sight, may prefer different resolutions for type. Practically speaking: I don't see a convincing argument for having layout elements infer their size based on font-size. I tried this once, and ended up with 30em as a total width that worked well on desktop with my base font-size of 12-pixels — but for those people who need higher resolution base heights, this is a disaster: the layout requires massive panning left and right to be able to make sense of it (especially on the phone, which has pretty limited screen size). Better keep the font-size based on the em, but keep the layout measurements fixed or fluid based on fractions of the viewport (%). If you're not interested in any of these aspects, use pixels: they're precise, unambiguous, and suffer none of the headaches of the cascade. Regards, Barney Carroll barney.carr...@gmail.com +44 7429 177278 barneycarroll.com ______________________________________________________________________ css-discuss [css-d@lists.css-discuss.org] http://www.css-discuss.org/mailman/listinfo/css-d List wiki/FAQ -- http://css-discuss.incutio.com/ List policies -- http://css-discuss.org/policies.html Supported by evolt.org -- http://www.evolt.org/help_support_evolt/