W3C home > Mailing lists > Public > www-style@w3.org > July 1997

Re: the *precise* definition of 1em

From: Joel N. Weber II <devnull@gnu.ai.mit.edu>
Date: Sun, 20 Jul 1997 20:39:08 -0400
Message-Id: <199707210039.UAA26408@mescaline.gnu.ai.mit.edu>
To: davidp@earthlink.net
CC: www-style@w3.org
   From: "David Perrell" <davidp@earthlink.net>
   Cc: <www-style@w3.org>
   Date: Sun, 20 Jul 1997 00:35:32 -0700
   X-MSMail-Priority: Normal
   X-Priority: 3
   X-Mailer: Microsoft Internet Mail 4.70.1161
   MIME-Version: 1.0
   Content-Type: text/plain; charset=ISO-8859-1
   Content-Transfer-Encoding: 7bit

   Joel N. Weber II wrote:
   > My point is that when your approximation of the size for one is 12
   > points, and the approximation of the size for the average is 10
   points,
   > then when you specify 1em, the child element will now have an
   approximated
   > size of 10 points, and so the closest match for font one will now be
   > nine points, where the parent size of one was 12 points.

   Ah. I see. You are revising the first choice to produce a more
   consistently sized set. That seems like the right thing to do. Child
   elements are likely to look better when smaller than the parent, not
   larger.

But there's about an equal probablity that the child elements will become
larger instead of smaller if we approximate them.

And it seems inconsistant to me if we have a more consistant size only
when we specify 1em.

   > Hmm... if you're trying to make something as tall as the text, it
   would
   > need to take the size of the largest font, I think.

   That too makes sense. 

OK.  Another idea then is to use one value for 1em for selecting font-size,
and another for any other uses.  font-size is already strange, in that 1em
for font-size uses the parent, while 1em for anything else refers to
font-size of that element.

I think that in general, 1em for things other than font-size ought to be
the biggest size of any one font, since we want to make sure we have
enough space for the largest font.  For font-size, I'd prefer to use
the approximated value if it's the same for all fonts, otherwise the
user-specified value.

(Actually, I'd rather just use the user-specified value.)

   > I don't think X11 can scale bitmap fonts.  You can have scalable
   fonts,
   > but some fonts can't be scaled.

   I believe scalable bitmap fonts are supported in X11R5, and SPEEDO
   (Bitstream?) vector fonts. And what about FontTastic? Hey, why not
   write your own font server?

Scalable fonts are supported, but not all fonts are scalable.

If we can't scale the font, we still have to do something.
Received on Sunday, 20 July 1997 20:39:06 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:53:50 GMT