Absolute and relative font sizes; was RE: Response to comments

[Bob Duncan]

Jeff Bobicki wrote:
> [snippage]
>"forcing a font size" is bad business. Use only relative fonts. For 
>example:
>
><FONT FACE="Arial" SIZE="-3" COLOR="Black">
>sets the font to a size "relative" to the normal where:
>
><FONT FACE="Arial" SIZE="3" COLOR="Black">
>would force a size 3 font to the browser which would probably make it
>unacceptable to both.

Jeff's point is well meant, but I think these statements illustrate a lot
of the confusion and misunderstanding surrounding the dreaded FONT tag.
Relative font sizes are no less guilty of "forcing a font size" than
so-called absolute font sizes: *all* calls for a font size are *relative*
to the size set in a user's preferences/options setting, and both relative
and absolute sizes "force" a size on the user. And IMHO, neither is "bad
business" unless misused (which is frequently). 

My lecture on font sizes, as I understand things...

Unless a Web page author alters the default font size using the BASEFONT
tag, any "plain" text on a page is understood by the browser as size
3---i.e., the default basefont is 3. (Out of 1-7; most browsers have a
limitation of seven sizes in which they can display text at one time).
Browser size 3 is equivalent to whatever the user has set as the display
font size in his or her browser's preferences. So, in my Netscape
prefs/options, whether I have my font size set to 12pt or 24pt, this is the
size the browser treats as 3. A call for the relative SIZE=+1 or the
"absolute" SIZE=4 produces *exactly* the same result: text is rendered one
size larger than (relative to) the default. (I'd see text so coded in a
14-ish font if my default is 12pt, or 26-ish if my default is 24pt.)
Likewise, a call for the relative SIZE=-2 or the "absolute" SIZE=1 produces
the same result: text rendered 2 sizes smaller than the default. (I'd "see"
8-ish point text if default=12pt.) A "relative" -2 will be far more
disruptive to my ability to read text than an "absolute" 2. (And in the
quoted example above, "-3" would render as size 1 since basefont 3 minus
3=0, which is not an acceptable font size; and "3" should be thoroughly
acceptable since it represents no divergence from the norm.)

Altering font sizes only becomes a problem when Web authors don't
understand the implications of what they do. As a user, I set my
prefs/options so that text unaltered by a font tag is a comfortable size
for my viewing on my specific setup. As long as an author leaves the font
size alone, I can read lots of text without a problem. However, an author's
call to alter the size in which my browser displays text is in effect
saying "I don't care what *you* think is readable, here's what you're
getting from me." This is no big deal (and perfectly acceptable) when
smaller sizes are applied to disclaimer text and larger sizes are applied
to non-heading text which should stand out, but how many times have we seen
the content of an entire page coded with "-1" or "-2", rendering it close
to illegible? (Or almost as bad, "+2" or "+3"?) Sure, I can alter the size
in my browser to accomodate a specific site, but then it has to be done
again after I leave the site. I can only guess that the authors of these
pages are designing on the extreme ends of the monitor spectrum without
proper adjustments made in local browser preferences, and to look
acceptable to them (or the keeper of the coin), the text has to be coded
with the size alterations. (Sorry for the generalization; just a guess.)

Not every Web author has access to a dozen or so hardware/software
configurations for testing pages, but an awareness of the different
environments which exist and the effects of various coding on each isn't a
secret, nor is it all that hard to figure out. The FONT tag (relative or
absolute) isn't the bad guy any more than Style Sheets are. Each has its
purpose and can be used to great effect if used properly and judiciously
and with some understanding of the potential outcomes.

Bob Duncan


  ~'~'~'~'~'~'~'~'~'~'~'~'~'~'~'~'~
  Robert E. Duncan
  Reference/Instruction Librarian
  David Bishop Skillman Library
  Lafayette College
  Easton, PA  18042
  duncanr at lafayette.edu
  http://www.library.lafayette.edu/