Different size for different font

[Jonas Smithson]

In his book "CSS: The Definitive Guide" 2nd edition (pgs. 116-117), Eric
Meyer has an interesting discussion about "font-size-adjust" that was
evidently dropped in CSS 2.1 due to browser non-support. I'm wondering
if there's some way I could still get a primitive version of this
functionality, because different fonts display at very different optical
sizes for the same nominal sizes.

For example, suppose I want type to display in 13px/1.5em Georgia. But
if the user doesn't have Georgia installed, and I've simply listed Times
New Roman as the alternate font in the usual way (or if "serif" gets
interpreted as Times on their system), then the type will display in
13px Times, which is optically much smaller than 13px Georgia (and has
optically different xheight/linespacing characteristics too). Even if I
used em instead of px, I'm guessing there'd be a substantial size
difference.

So what I'd really like, in the absence of "font-size-adjust", is
something like this:

p.someClass {font: 13px/1.5em Georgia, 15px/1.4em "Times New Roman",
14px/1.5em serif;}

....but I have a feeling that wouldn't actually do what I want. Er...
would it?

Any solution?

[Bergamot]

Jonas Smithson wrote:

>
different fonts display at very different optical
sizes for the same nominal sizes.

For example, suppose I want type to display in 13px/1.5em Georgia.

Simple solution - don't specify an absolute font-size. If you leave it
at the default size (100%), then the user will see the font (be it
Georgia or not) in whatever size *they* think is optimal.

--
Berg

[Jukka K. Korpela]

Scripsit Jonas Smithson:

In his book "CSS: The Definitive Guide" 2nd edition (pgs. 116-117),
Eric Meyer has an interesting discussion about "font-size-adjust"
that was evidently dropped in CSS 2.1 due to browser non-support.

It has now been implemented on Firefox 2 (on Windows).

p.someClass {font: 13px/1.5em Georgia, 15px/1.4em "Times New Roman",
14px/1.5em serif;}

...but I have a feeling that wouldn't actually do what I want. Er...
would it?

I doubt that. It violates CSS syntax.

You might be able to do something via client-side scripting, but that would
be off-topic here.

--
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/

[Jukka K. Korpela]

Scripsit Bergamot:

Jonas Smithson wrote:

>>
different fonts display at very different optical
sizes for the same nominal sizes.

For example, suppose I want type to display in 13px/1.5em Georgia.

Simple solution - don't specify an absolute font-size. If you leave it
at the default size (100%), then the user will see the font (be it
Georgia or not) in whatever size *they* think is optimal.

You missed the point. Of course absolute font sizes should be avoided, but
that doesn't solve the problem.

For example, suppose you want to use Arial. Setting just font-family: Arial
works for a great majority of users. And there the problem is. If they have
set their browser to use a font size that is suitable for their browser's
default font, which is probably Times New Roman, they will see Arial in that
size, which is probably 12pt (16px).

While 12pt Times New Roman looks reasonable on screen, 12pt Arial looks too
big to most people. So what can you do?

If you ask me, you should just accept that. Too big font size is seldom
complained by _users_ (and quite often by designers), and users can reduce
the font size if it matters.

But many authors set e.g. font-size: 80% or even smaller. This is usually
not catastrophic, but it's not nice either, especially to users who have set
their browser to use e.g. Arial in 10pt size. The reduction takes the size
to 8pt, which is probably too small to be read comfortably.

--
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/

[Kevin Scholl]

Jukka K. Korpela wrote:

>>different fonts display at very different optical
sizes for the same nominal sizes.

For example, suppose I want type to display in 13px/1.5em Georgia.

Simple solution - don't specify an absolute font-size. If you leave it
at the default size (100%), then the user will see the font (be it
Georgia or not) in whatever size *they* think is optimal.

You missed the point. Of course absolute font sizes should be avoided,
but that doesn't solve the problem.

For example, suppose you want to use Arial. Setting just font-family:
Arial works for a great majority of users. And there the problem is. If
they have set their browser to use a font size that is suitable for
their browser's default font, which is probably Times New Roman, they
will see Arial in that size, which is probably 12pt (16px).

While 12pt Times New Roman looks reasonable on screen, 12pt Arial looks
too big to most people. So what can you do?

If you ask me, you should just accept that. Too big font size is seldom
complained by _users_ (and quite often by designers), and users can
reduce the font size if it matters.

But many authors set e.g. font-size: 80% or even smaller. This is
usually not catastrophic, but it's not nice either, especially to users
who have set their browser to use e.g. Arial in 10pt size. The reduction
takes the size to 8pt, which is probably too small to be read comfortably.

This argument is all well and good in theory, but one important part of
it really isn't based -- in my experience -- in the real world. That
part is the "users who have set their browser to use...". I've come
across maybe two or three people in my years of Web work that have
utilized user style sheets beyond a little experimentation , and none of
them was your typical mainstream user. Most people don't even know such
a thing exists.

A few more may change the default font size in the browser View menu,
but very few, certainly (IMO) not enough to compromise ideal visual
appearance.

The way I see it, font size is largely an element of design. That is,
the designer can SUGGEST what they believe works best visually, so long
as they allow the user to size as needed (by using ems or percentages).
This way, one meets the accessibility needs of the vast majority of
users, while initially displaying the intended look.

I know YMMV, and some folks here will vehemently disagree with this
approach, but I've not gotten a single complaint from a client or client
user.

--

*** Remove the DELETE from my address to reply ***

=============== =============== =============== =========
Kevin Scholl http://www.ksscholl.com/
ks*****@comcast .DELETE.net
------------------------------------------------------
Information Architecture, Web Design and Development
------------------------------------------------------
We are the music makers, and we are the dreamers of
the dreams...
=============== =============== =============== =========

[Jonathan N. Little]

Kevin Scholl wrote:

I know YMMV, and some folks here will vehemently disagree with this
approach, but I've not gotten a single complaint from a client or client
user.

http://www.ksscholl.com/

Well 75% of 16px is 12px and on my monitor is a bit small for the copy
especially when grey on white...

Maybe folks don't complain they just don't come back.

--
Take care,

Jonathan
-------------------
LITTLE WORKS STUDIO
http://www.LittleWorksStudio.com

[dorayme]

In article <JP************ *************** ***@comcast.com >,
Kevin Scholl <ks*****@comcas t.DELETE.netwro te:

Jukka K. Korpela wrote:

But many authors set e.g. font-size: 80% or even smaller. This is
usually not catastrophic, but it's not nice either, especially to users
who have set their browser to use e.g. Arial in 10pt size. The reduction
takes the size to 8pt, which is probably too small to be read comfortably.

This argument is all well and good in theory, but one important part of
it really isn't based -- in my experience -- in the real world. That
part is the "users who have set their browser to use...". I've come
across maybe two or three people in my years of Web work that have
utilized user style sheets beyond a little experimentation , and none of
them was your typical mainstream user. Most people don't even know such
a thing exists.

I assume you mean to include those who look into properties or
browser font preferences (Mac). If, in general, a user was not
comfortable with the font sizes he was seeing, then he would
likely get to do something about it. There is a sort of arms race
going on. Mostly, and ridiculously, authors use less than 100%
even for main body text (to look neat, to get more in etc). So
browsers either come with or are set by technicians or savvy
friends or users to show such comfortably. These defaults or
adjustments are made to compensate for the unnatural choices of
authors in general. When an author uses 100%, it can cause
surprise and comment about 'the bigness'.

The way I see it, font size is largely an element of design. That is,
the designer can SUGGEST what they believe works best visually,

Perhaps it behoves authors not to think this way as it is too
risky and too paper-print orientated in concept. Happier, and
much weaker of course, is the idea that font-size _range_ is an
element of design. In plain language, if a design looks good at a
few clicks either way of text size, so much better is the design
for websites. Your stated approach is different and can result in
one optimum and everything goes bad after this whereas what many
people have been thinking around these parts is that there is a
likely a cluster of nice enough and useful enough range of
font-size designs,

--
dorayme

[Kevin Scholl]

dorayme wrote:

In article <JP************ *************** ***@comcast.com >,
Kevin Scholl <ks*****@comcas t.DELETE.netwro te:

>Jukka K. Korpela wrote:

>>But many authors set e.g. font-size: 80% or even smaller. This is
usually not catastrophic, but it's not nice either, especially to users
who have set their browser to use e.g. Arial in 10pt size. The reduction
takes the size to 8pt, which is probably too small to be read comfortably.

This argument is all well and good in theory, but one important part of
it really isn't based -- in my experience -- in the real world. That
part is the "users who have set their browser to use...". I've come
across maybe two or three people in my years of Web work that have
utilized user style sheets beyond a little experimentation , and none of
them was your typical mainstream user. Most people don't even know such
a thing exists.

I assume you mean to include those who look into properties or
browser font preferences (Mac). If, in general, a user was not
comfortable with the font sizes he was seeing, then he would
likely get to do something about it. There is a sort of arms race

That's rather my point. Most people, indeed a very small percentage,
actually do look into these things. Realistically, most don't need to,
and a well-designed site allows a user to resize text through the
browser's View menu without having to set a global minimum in the
browser setting (i.e., a user style sheet). Unfortunately, there are a
great many sites -- too many -- that are not designed in such a manner.

going on. Mostly, and ridiculously, authors use less than 100%
even for main body text (to look neat, to get more in etc). So
browsers either come with or are set by technicians or savvy
friends or users to show such comfortably. These defaults or

Honestly, I don't see this happening. A browser's font settings being
set to anything but the default is an exceedingly rare occurrence in my
experience. Again, I'm not saying my experience is the be-all, end-all,
though.

adjustments are made to compensate for the unnatural choices of
authors in general. When an author uses 100%, it can cause
surprise and comment about 'the bigness'.

>The way I see it, font size is largely an element of design. That is,
the designer can SUGGEST what they believe works best visually,

Perhaps it behoves authors not to think this way as it is too
risky and too paper-print orientated in concept. Happier, and
much weaker of course, is the idea that font-size _range_ is an
element of design. In plain language, if a design looks good at a
few clicks either way of text size, so much better is the design
for websites. Your stated approach is different and can result in
one optimum and everything goes bad after this whereas what many
people have been thinking around these parts is that there is a
likely a cluster of nice enough and useful enough range of
font-size designs,

Perhaps I should have specified to this point. Certainly I expect a
well-designed site to "behave" as font sizes are changed, at least
within some reason (I doubt all but the simplest sites would not break
layout if text is resized to its absolute maximium).

--

*** Remove the DELETE from my address to reply ***

=============== =============== =============== =========
Kevin Scholl http://www.ksscholl.com/
ks*****@comcast .DELETE.net
------------------------------------------------------
Information Architecture, Web Design and Development
------------------------------------------------------
We are the music makers, and we are the dreamers of
the dreams...
=============== =============== =============== =========

[Bergamot]

Kevin Scholl wrote:

[re: font-size:75% and other tiny settings]

I've not gotten a single complaint from a client or client user.

FYI, I've had font size discussions with almost everyone I've worked
with. You might be surprised at the number of users that are secretly
dissatisfied with the tiny type size prevalent on web sites. Once I
teach them how to adjust their browser settings (usually to make the
default larger), their only remaining complaint is that it doesn't work
on every web site.

BTW, I've never gotten a single complaint from a client or client user
about font-size:100% being too big, so your lack of complaints doesn't
mean there aren't any, just they don't tell you about it.

--
Berg

[Jukka K. Korpela]

Scripsit Kevin Scholl:

This argument is all well and good in theory,

Before that statement, you quoted my entire message, practically speaking.
Comprehensive quoting generally indicates lack of comprehensive reading on
Usenet. In particular, you missed the point that my argument was practical -
actually, more practical than yours.

-- in my experience -- in the real world.

Your personal opinions don't become any more objective if you call them the
real world.

A few more may change the default font size in the browser View menu,
but very few, certainly (IMO) not enough to compromise ideal visual
appearance.

"IMO" is the key here; there's your real world: you _don't care_ about
people who set the font size.

Of course, this is just the top of the iceberg. There are millions of people
around who have never touched any relevant browser settings and to whom
their browser's basic font size is the _smallest readable_ size. So you want
to go below that, in your quest for "ideal visual appearance" (i.e., a look
that pleases _your_ eyes, or the eyes of your team, or whatever).

The default setting of 12pt is good exactly _because_ it is somewhat too
large to most people when sans-serif fonts are used. Give some thought to
the variation of people and browsing conditions, and you'll see this.

The way I see it, font size is largely an element of design. That is,
the designer can SUGGEST what they believe works best visually, so
long as they allow the user to size as needed

You don't present any concrete method so that we could evaluate it. If
you're suggesting body { font-size: 80% }, then that's usuallu not
catastrophic, as I wrote, though it prevents users from _easily_ selecting a
fairly large font size (since "largest" in the font size menu of IE is
relative to the size set on the page). But your tone of voice suggests
something more drastic. Even if it didn't, many newbies and clueless
"advanced" authors will apply your principles so that they set the font size
to the smallest value they can read in optimal circumstances. This is often
accompanied with other settings that reduce legibility further.

And, for example, the page mentioned in your sig (
http://www.ksscholl.com/ ) uses tiny text, in gray on a little lighter gray
background. Even when I set the font size to "largest" on IE, it's
inconveniently small due to insufficient contrast, and this means that for
millions of people, it's too small to read. Of course, _those_ people who
have _severe_ eyesight problems have already somehow configured, or have had
someone to configure, their browsing environment to deal with such issues,
e.g. by disabling all author stylesheets; otherwise they wouldn't be
browsing at all. But a much larger group just finds the page too difficult
to use due to its _visual presentation_.

Moreover, if you think that most people don't even use the font size menu,
you should be worried about hundreds of millions of people who just can't
read tiny gray text on gray background. They simply go away. Now you might
say that they are not your potential customers. But your potential
customers, tough young people with good eyesight, might understand the
importance of thinking about _their_ customers in a manner that you don't
seem to care about.

--
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/