Re: font-size and accents, again

On Mon, 22 Nov 1999 16:27:54 -0800, "Tantek Celik"
( wrote:
> >> Those units are defined in terms of "font size", and since font-size
> >> corresponds to "em", I think those units should continue to be based on
> >> the em, rather than the maximum height of the glyphs.
> >
> > The reason I think this is a bad idea is that it is not backwards
> > compatible with most current behavior, and the current behavior makes
> > any line-height above 1.0 "safe" (i.e., it cannot cause overlap).
> Not the "current behavior" in IE for Mac.  It is certainly possible to have
> a font which is rendering long flourishes etc. far outside of its em-height
> which overlaps with adjacent lines of text formatted at line-height:1.0 or
> even a bit higher.

Hmmm.  That's not what happens in the screenshot I was sent [1].
(Behavior of other browsers is listed on [2], in the fourth column of
the results table.  Note that NN 4.x's behavior falls into the "safe"
category, along with the browsers marked "O".)

> > This
> > would mean that things that were once reasonable suggestions could now
> > be unsafe.
> They are "unsafe" as you say, right now, with shipping browsers, today.

The only browsers I see where it's unsafe are Opera 3.6 and WinIE5 [2].

> > Since scaling factors (i.e., 'normal' or a number) are the only safe
> > way of suggesting line-height because of inheritance, I think they should
> > be kept safe in all respects.
> I disagree with the general philosophy that it is CSS's responsibility to be
> "kept safe".  We should make things easy to understand and author to, but
> not go out of our way to restrict flexibility and empowerment.
> Who are we to decide what is safe and what is art?

This approach doesn't make CSS any less flexible.  Everything that was
possible before is still possible.  In fact, I would argue that it
makes suggesting small line-heights possible while it is not possible
the other way, since authors need not fear overlap if a font is
substituted because the suggested font is unavailable.  CSS should be
geared towards making it easier for the author to suggest styles that
make sense across platforms.

Or would you prefer that everybody on the web use Windows?? ;-)

> > New units can't be introduced to CSS for at least 4 years or so in any
> > useful way, since many existing browsers will treat them as pixels.
> That would be in violation of CSS1 Section 7.1 Forward Compatible parsing,
> now wouldn't it?


-David "section 7.1 is the most important section of CSS1" Baron


L. David Baron    Sophomore, Harvard (Physics)
Links, SatPix, CSS, etc.     <URL: >
WSP CSS AC                      <URL: >

Received on Monday, 22 November 1999 19:57:32 UTC