WebAIM - Web Accessibility In Mind

E-mail List Archives

RE: Font size


From: Jukka Korpela
Date: Jun 5, 2002 11:30PM

Leo Smith wrote:

> Please load this page into IE/PC and adjust your browser's text
> size settings - smallest, smaller, medium
> http://www.usm.maine.edu/~lsmith/test.htm

Oh, I see... I first thought that you meant the CSS keyword values for
font-size. Sorry for my confusion.

And now I see that IE really gets em sizes wrong. I never gave that much
thought, since having learned that you can use either em or % for font-size,
I decided to use % because it looked intuitively more natural. I didn't
realize there were more serious things involved.

Does anyone know why the WAI guidelines recommend em sizes, in a context
where they recommend % sizes for things like margins? (In the clause already
http://www.w3.org/TR/WCAG10-CSS-TECHS/#units ) Wouldn't it be _simpler_ to
say just "use percentages"?

Maybe they were just afraid of misunderstandings. After all, in e.g.
the percentage has a meaning different from percentages in font-size. But
such things probably just need clarifying explanations, not avoidance of
percentages. (Besides, using e.g. margin-left:15% isn't such a good idea
IMHO. I think it would be better to use em units there! After all, if a user
has, out of necessity or just by his choice, a fairly narrow window, it
wouldn't be that nice to have
one fourth of it eaten by BODY { margin-left: 15%; margin-right: 10%} for
example. Something like margin-left:2em; margin-right:1em should create
reasonable margins. For printing, the user needs to check the print settings
anyway, or the author needs to supply special CSS rules for print media.)

> If you have visual
> users who are leaving your site as they cannot read anything, that
> is an accessibility concern.

Certainly, but the problem needs to be solved where the problem lies. Your
point about using % sizes and not em sizes for font-size was important.
There we have a real choice between methods that are equally correct and
equivalent by CSS specifications but not in faulty implementations that are
common, so we can mentally adjust ourselves to, er, reading creatively the
statement in WAI guidelines. But at the general level, we cannot prepare to
deal with all the possible problems that users may have due to browser
errors and misconfigurations. Just because some browser might be by default
set to use a 9px font isn't a good enough reason to write an author style
sheet that tries to override the user agent's basic font size setting for
the document as a whole.

> The fact that they could have solved
> that barrier themselves by changing their browser's text size
> setting back to medium is an academic argument.

No, I think it's a very practical argument to say that if the basic font
size in a user's browser isn't suitable to the user, he should fix it, or
someone should fix it for him.

But we might consider the question how much we make preparations for
settings that aren't very bad but not optimal either. Specifically, if we
think that people very often have browsers that use _a little_ too small
font by default, then we, as authors, should be very careful with suggesting
font size reduction. We might think that a suitable font size setting
tolerates e.g. 25% reduction without making the "small print" too hard to
read. But if we, in fear of making text unreadably small, use just small
reduction in size, then the reduction might become (almost) unnoticeable,
and the user would miss our visual hint of some parts of the document being
less important than normal text. Or there would be _other_ accessibility
problems: Often some font-size reduction helps with tables with several
columns, so that users don't need to scroll horizontally.

So font size reduction (which we might use in special occasions) is a
difficult "optimization" problem. We can't really find an _optimum_, just a
good guess, hopefully. Personally, I avoid going below font-size:85%, except
for subscripts and superscripts (font-size:80%, and a sans-serif font -
intended to help with the problem that on many browsers line spacing may
become uneven unless you do something about the size of subscripts and

Jukka Korpela, senior adviser
TIEKE Finnish Information Society Development Centre
Phone: +358 9 4763 0397 Fax: +358 9 4763 0399

To subscribe, unsubscribe, or view list archives,
visit http://www.webaim.org/discussion/