RE: What's an em

(I don't dare cc www-style, since that seems to
cause multiple copies on each of those lists.)


> -----Original Message-----
> From: Clive Bruton [mailto:clive@typonaut.demon.co.uk]
...
> Sure, what does 7mm rasterise to at varying dpi screen rates 
> 72/96/120...?

Ideally (yes, ideally, I know that this is not the case in practice
(yet)), for a desktop/laptop screen at a zoom of 100% a 7 mm
ex height SHOULD result in rasterisations such that the glyph
images have an ex height close to, guess what, 7 mm!  But that
is of course too obvious to be true yet...

And if things turn out too big, or too small, zoom is good.
Only available in Opera and MS Word as yet.  I have suggested
a default zoom for desktop/laptop screens to be about 130%,
and have that specified in CSS.


...
> I think you're also indicating that line-height should also flex in 
> sympathy with font-size-adjust? This seems particularly odd, since 
> line-height is yet another factor of perceived size. Scaling 
> Verdana to 
> match your concept of perceived size is going to reduce the 
> line-height.

My suggestion (I'm not sure what anybody elses would be)
is to set the line height to something like '1line' or 
a little bit bigger, as usual, like '1.1line',
where "line" is the (nominal) Åp-height of the current font.
(And that's what I've said I don't know how many times by now.)


> As to accents... some may consider these to be at least 
> partially outside 
> the em square. Some practice determines that accents may not 
> be set on 
> caps or all caps text anyway (ie in French setting this is often the 
> case).

Not in modern French.  Accents are always to be kept (they are meaningful).
As for Finnish, Swedish, Norwegian, Danish, Vietnamese, Russian, etc. etc.:
don't remove what you think of as 'accents'.  They are NOT optional!

> It is far more likely that acsender/descender clash will happen than 
> accent/descender. Ascender heights in most fonts are set to ascender 
> rather than accent extremities.
> 
> I think x-height is a good value to have for meta data, but 
> the use you want to put it to is misconceived.

You have not convinced me one bit that the current kaos is in
any way better than having author requested ex height (or cap
height) adjustments applied for implicit (font substitution)
or explicit typeface changes.  I don't see how that would be
better for anyone.  And no, I'm still not asking for ultra-high
precision, how could I?  I'm happy with "sufficiently good".

I don't think 'font-size-adjust' is a good way of doing this,
though, since that relies on the author to give an aspect value,
and in addition 'font-size-adjust' does not apply to explicit
typeface changes.


		Kind regards
		/kent k