Re: Action items from 1/23/07 meeting

Hi all,

On 23/01/07, Loretta Guarino Reid <lorettaguarino@google.com> wrote:
> Gez - review techniques for How To Meet 1.4.5, 1.4.6 for correctness,
> completeness

I've made some comments below.

Best regards,

Gez

TECHNIQUE: USING PERCENT FOR FONT SIZE [1]

In the user agent and assistive support notes section, the first
bullet contains the following point:

"When font size is given in pixels, The Text Size menu commands in
Internet Explorer 6 and earlier do not resize the text"

This is true for all absolute units of measurement, such as points.
The 'T' in "The" for "The Text Size menu" shouldn't be in uppercase.

The second bullet point states:

"When High Contrast Mode has been set from the Accessibility Control
Panel on Windows, IE6 increases the size of the page text using
percent."

I'm not sure what this is trying to say, but guess that it's raising
the same issue as the first bullet point. I'm not sure why percent is
being singled out, as the text is also resized for other relative
units of measurement, such as em. Maybe it would be better to combine
this point with the first bullet point, or state something like:

"When High Contrast Mode has been set from the Accessibility Control
Panel on Windows, IE6 and earlier will not honor the text-size for a
particular high-contrast mode when absolute units of measurement, such
as pixels, have been used."

The second bullet point goes on to say:

"When tables use percent font sizes, the percents are multiplied for
nested tables. This causes the text to get much larger (or smaller) as
tables are nested more and more deeply."

This is the correct behaviour for all properties that can be inherited
from parent elements. For example, the same is true for nested lists,
where 'font-size' has been specified as a percentage for 'li', as the
inherited value will be a percentage of a percentage (only works if
it's 100%). This is obviously also true for any relative unit of
measurement; for example, 0.9em is equivalent to 90%. As this is
standard behaviour, I don't think the technique should contain an
explanation of inheritance in CSS. If others disagree, then the
explanation shouldn't be in the user agent assistive support notes, as
this is a feature of CSS, not a particular user agent oddity.

Under the description for this technique, it states:

"Changes to the default font size will propagate smoothly to all the text."

I think that should be a bit more explicit; "If a font-size is
specified for the body element, all other elements inherit that value,
unless overridden by a more specific selector."

Under procedure, it states:

"Examine the font-size property of the style"

I think that might be better if it said, "Examine the font-size
property for each rule-set"


TECHNIQUE: USING NAMED FONT SIZES [2]

In the user agent and assistive support notes section, the first
bullet contains the following point:

"When font size is given in pixels, The Text Size menu commands in
Internet Explorer 6 and earlier do not resize the text"

This is true for all absolute units of measurement, such as points.

Under description, it states:

"These values provide hints so that the user agent can use fonts that
satisfy the requested size relationships as well as possible."

I don't know how to reword it, but the end of the sentence seems a bit
clumsy. Maybe,

"These values provide hints so that the user agent can choose a
font-size relative to the inherited font-size."?

Should we mention " Named Font Size in HTML"? The example does make it
clear that these elements are presentational in nature, but
personally, I think it would be better not to mention them at all.

There are a couple of grammar mistakes in the "Named Font Size in CSS" section:

"This example selects a larger font size for em element[s] so that its
text will always be larger than the surrounding text, in whatever
context [it - should be "they"] are used."


TECHNIQUE: USING EM UNITS FOR FONT SIZES [3]

In the user agent and assistive support notes section, the same issue
applies as the previous techniques.

Under description, the second sentence starts:

"Since the em square is a property of the font, it scales as the font
changes size."

This may well be correct, but I don't know what an em square is, so
can't confirm. The part about, "Changes to the default font size will
propagate smoothly to all the text", should be more explicit; "If a
font-size is specified for the body element, all other elements
inherit that value, unless overridden by a more specific selector."

Under Examples, the selector for the CSS rule-set should be 'strong':

strong {font-size: 1.6em}

Under procedure, it states:

"Examine the font-size property of the style"

This may be better if it read, "Examine the font-size property for
each rule-set"

FAILURE TECHNIQUE [4]

Under description, the second bullet point item states:

"Setting the width of a containing element but not the height, so the
container grows in length in a way that overlays containers below it"

The issue here is more about how the adjacent elements respond when a
container is resized. This point seems to suggest that restricting the
height of a container solves that, but it would only do that if the
content was clipped, which conflicts with point 1 in this list. I
would remove this point, unless there was another issue that I'm
missing.


[1] http://trace.wisc.edu/wcag_wiki/index.php?title=Using_percent_for_font_sizes

[2] http://trace.wisc.edu/wcag_wiki/index.php?title=Using_named_font_sizes

[3] http://trace.wisc.edu/wcag_wiki/index.php?title=Using_em_units_for_font_sizes

[4] http://trace.wisc.edu/wcag_wiki/index.php?title=Failure_when_resizing_visually_rendered_text_up_to_200_per_cent_causes_the_text_to_be_clipped%2C_truncated_or_obscured


-- 
_____________________________
Supplement your vitamins
http://juicystudio.com

Received on Monday, 29 January 2007 16:23:23 UTC