Warren Steel (
Tue, 21 May 1996 08:46:43 -0500

Message-Id: <>
Date: Tue, 21 May 1996 08:46:43 -0500
From: Warren Steel <>
Subject: Font-style...

Benjamin Franz <> wrote:
>Nothing requires *YOU* to use FONT....

   This is an odd point of view for a web author.  Graphic designers
use font styles, colors, and extreme sizes in print advertisements.
If the reader finds them ugly, or illegible (nearsighted, colorblind),
then the advertiser has lost a customer.  The World Wide Web, and
hypertext markup, provide a unique solution to the problem of user
needs and preferences.  Information providers (including advertisers)
can *rejoice* in the knowledge that their work will appear different
to each viewer, but accessible to all.  If you specify font colors,
absolute font sizes, or extreme relative font sizes, you have defeated
your own purposes, since viewers cannot reverse their effects.  
This can hardly be advantageous to the author, or to a client.

>Would stylesheets be better? Yup. Can stylesheets be mis-used the same way
>as FONT? <p class="bigtext">Since H1 excludes all but physical markup but
>&lt;p&gt; allows <strong>other</strong> stuff I might like in my topic
>headers...</p> you can bet your bippy stylesheets can be mis-used.
>Most authors *don't care* about content models. Or more precisely: most
>have *never heard of* content models.

    I don't doubt that stylesheets can be abused, once their potential
becomes widely known.   I think all of us agree that Font information
belongs primarily in style sheets, where it can be disabled by the 
viewer.  And I also acknowledge that web browsers could be built in 
which the effects of the <FONT> tag could be completely disabled 
by the user.  The HTML 3.2 materials attempt to document "what 
currently works" on the Web, among "market leaders."  My point
is that <FONT> is currently *broken* on the Web, and that 
information is being lost every day on its account.  Developers
have not added the ability to cancel its effects.  By the time
they do so, stylesheets will be available to all. (Ron Newman) wrote:
>In fact, a far
>as I can see, there is no real difference between <BIG> and
><FONT SIZE="+1">, or between <SMALL> and <FONT SIZE="-1"> .

   That's true.  <FONT SIZE="+1"> and <FONT SIZE="-1"> might
be retained as an "alias" for <BIG> and <SMALL>.  

>>are many situations, unforeseen to authors, in which text can become
>>irrecoverably illegible or even invisible to a broad range of users.

>I'd like to hear more about this.

   My objections to fonts is that innocent authors trying to 
get their message across find themselves producing markup that
ensures that some viewers actually lose information.  Let me 
recap briefly:

   1. Sight-impaired users often choose large base fonts to
ensure legibility.  Users of laptops and other small displays 
often use as small basefonts as possible, to get a workable
amount of text onscreen at a time.  In either case, there 
may be a rather narrow range of font sizes that ensures legibility.  
Allowing absolute font sizes ("1") or extreme relative sizes ("+4" 
or "-3") violates that range, without allowing the user to disable 
them (as style sheets would), and *guarantees* illegibility or 
inconvenience to many users.  <BIG> and <SMALL> are sufficient.

   2. As you know, users of popular browsers have ultimate 
control over background and text colors.  While the body-color 
tags can be controlled and set by the user, font-colors cannot, 
in current implementations.  If I have a yellow background set, 
I will never even be aware of <FONT COLOR=yellow> text in a 
document.  Considering the wide variation of color schemes
in documents currently on the Web, it is no surprise that
information is being lost every day in this manner, and I 
have received several messages confirming this fact.  I guess 
I must view the source of every document, to see if I've 
missed something!  

   3. <FONT> is an unreliable method of emphasizing specific 
text.  <STRONG>, <EM>, <BIG>, <I>, <B> are all much more portable,
and cause none of the above problems.  But the truly harmful
effects of <FONT> are reserved for the browsers that *do*
recognize it.  The concept was flawed from the start, its
implementation is broken in current browsers, and it will
be obsolete in a few months.  Instead of "enshrining" it
in a specification, it should be rejected or deprecated in
the strongest terms.  It is time for the members of the W3C
to cut their losses and minimize their embarrassment over
this unrealistic and unsuccessful addition to HTML, which
results in such a loss in communication over the Web!

Warren Steel              
Department of Music              University of Mississippi