Re: ascender, descender, cap-height and x-height

Erik:

The reason for this inconsistency is historical. Font design
traditionally allows the font designer to specify these things based
on their own eye.

It's hard to imagine that foundries and individual designers will go
back and change the tens of thousands of font designs they have to
fit some standard.

What's more, I personally have yet to see any kind of font
substitution system that really worked well enough. Panose, by Ben
Bauermeister (see
http://www.amazon.com/exec/obidos/ASIN/0442211872/o/qid=948938748/sr=8
-1/104-0822201-5846809 ) is the best system so far, and it has it's
limitations in that fonts themselves must have the Panose numbers
embedded in them, and I think only Microsoft and Bitstream have done
this.

Far better than encouraging substitutions would be to further promote
the Dynamic Fonts feature already in Netscape and secure, open font
embedding. Dynamic Fonts is an excellent and secure system that's
unfortunately rarely used, even though tools to create these fonts
are readily available.

I've written about Dynamic fonts www.webfonts.com (from O'Reilly)
both at http://webreview.com/pub/98/04/03/feature/index.html and
http://webreview.com/pub/97/11/07/feature/part1.html

If web developers were encouraged to use embedded fonts, then font
substitution would be unnecessary. Since both IE and Netscape now
support this format (IE through Active X controls), this would
probably be a more realistic way to "standardize" than trying to
change all the fonts that have already been developed.

]) /\ |\| | (- |_
----------------------------
Daniel@Will-Harris.com
http://www.will-harris.com - Home of EsperFonto
http://news.i-us.com/wire/
Editor, www.webfonts.com
=================
Editor, www.eFuse.com
The friendly place to learn how to build a better web site


*********** REPLY *********** 

On 1/26/00 at 4:50 PM erik@netscape.com wrote:

>Jelle Bosma wrote:
>> 
>> It is the type designer who decides on how large the font
>> is in the EM square. There are no rules. In TrueType
>> (and OpenType) you can get some measurements such
>> as the Typographic ascend and descend as explained by Greg.
>> These values are in the OS/2 table of the font.
>> In Monotype fonts these are set to top f and bottom g
>> for the font family. Whether fonts of other foundries
>> contain reliable values I am not sure. Many foundries didn't
>> bother much about these things until recently.
>
>Do you also happen to know why those "many" foundries are now more
>concerned about these things? I'd really like to know.
>
>The reason I'm concerned is because the Internet has caused many
>different platforms and many different document types to come into
>direct contact. In the past, we had various word processing programs
>like MS Word, WordPerfect, etc, and users cursed the software
industry
>for all of the (in)compatibility problems. These problems were
>compounded by differences between the OSes (platforms).
>
>However, now that we are using documents formats like HTML and XML
>together with style sheets like CSS's on the Web, we need to
standardize
>font sizes. It is impossible to predict what kind of platform the
user
>is using, and which fonts they have available. So, CSS provides for
font
>substitution. You still want to have a "reasonable" font size, even
when
>substitution takes place.
>
>But what exactly is "reasonable"? This is kind of subjective, but it
>seems like we have some agreement that the size of a "bicameral"
font
>(one with both upper and lower case) is mostly dependent on the
>x-height.
>
>That is why CSS defines font-size-adjust. But for font-size-adjust
to
>work well, the underlying font-size must also be specified and
>implemented well. Any help that www-font experts can provide is most
>appreciated.
>
>Erik



]) /\ |\| | (- |_
-------------------------------------------
Editor, www.eFuse.com
The friendly place to learn how to build a better web site

Help a friend join the fuseletter, and get free fonts and graphics!
It's easy, just go to:
http://xilonen.netobjects.com/efuse/tellafriend.cfm now!

Save your site in the eFuse Time Capsule at
http://www.efuse.com/New/time_capsule.html

Received on Thursday, 27 January 2000 16:41:29 UTC