Issue summary - "Font/scaling" sort term

Bruce Bailey prepared the attached issue summary. One additional comment was added since his report, and it is appended to this message.

See Bruce's report for other proposals. 

The following proposal was sent to survey:
Add Level 2 SC: 
1.3.6: When the user configures the scale of text, information and relationships in the content are preserved. 

Add sufficient technique for 1.3.6: 
Text is specified in a way that the size can be controlled by the user. 

Add HTML technique 1.3.6 
Text size is specified using relative units that permit user control. 

Add Level 3 SC 1.3.7: 
When the user configures the width of the viewport, information and relationships in the content are preserved, and all content is displayed within the configured width without need for a horizontal scrollbar.

Comments on that proposal are at http://www.w3.org/2002/09/wbs/35422/teamb10192006/results#xlc1437. 
Highlights are summarized:

-	I really think this should remain a UA issue. UAAG checkpoint 4.1 includes checkpoint 4.1 (Configure text scale) as a user agent requirement. Users who need to adjust font size beyond the limited options in IE have lots of options (ex. Opera, FireFox -- or if they want to stick with IE as it exists today, magnification software, web-based proxies or custom CSS). We also have an (unwritten) advisory technique under guideline 1.3 titled "Providing resizable text (future link)." 

If what we want to do is underscore the UAAG requirement in WCAG 2.0, then we might consider something like, "A mechanism is available to adjust the size of text." We could list some of the technqiues these issues are concerned about missing here as sufficient, but I don't think we should go so far as to suggest that the use of points or pixels would be failures since the UA features and assistive technology options available to users today are very good.
-	It is almost impossible to make substantial increases in text size in Firefox or opera without losing information and relationships. Perhaps this should only be required for up to a certain percentage of enlargement. Say 200% . Anything larger than that would pretty well require zoomtext or Magic which have tons of horizontal scrolling anyway.
-	The sufficient techniques are much less than the 1.3.6 sc. Nothing about preserving releationships. L3 SC 1.3.7 fails with ZOOM. What are sufficient techniques for SC 1.3.7
-	Don't understand what is meant in the proposed 1.3.6, "information and relationships in the content are preserved". Does it mean, "_the visual presentation of_ information and relationships in the content are preserved"? If so, use that wording, if not, need to discuss.

The proposed SC doesn't require that text be scalable, just that things don't break when it is scaled. Therefore the two proposed sufficient techniques seem to be not applicable. On the basis of the techniques, I think this or another SC needs to say in so many words "text content can be scaled". My best proposal is a new Level 3 SC saying just that, and attaching the sufficient techniques to that. The existing proposed SC would have a general technique about making this all work.

I think it's implied in the proposed 1.3.7 that layout regions would reflow when the viewport width is changed, but that is not stated, and should be if that's what is meant. The wording as is right now also doesn't require that layout regions adapt to fit the user's viewport when the page is loaded, which I would also think is intended. Assuming that all this is what is meant, a wording for the SC to capture this would be:

"Visual layout of content adapts to the width of the viewport without need for a horizontal scrollbar and preserves visual presentation of information and relationships."

I also think, if we are to adopt these SC as I understand them, we should consider having limits. Even on a page that follows these guidelines, I don't have to hit the "font bigger" command too many times in my browser before it becomes absurdly large and you can't expect layout regions to continue to preserve visual relationships while remaining inside the viewport. When text gets that large, you really need a screen model that's larger than the viewport à la screen magnifier. That puts us into AT land, and I don't know 1) where the line between "normal" magnification and "AT" magnification should be and 2) if we're even supposed to recognize such a distinction, as AT vs. normal UA is a practical, not a theoretical, distinction.
-	Don't we mean the "presentation" of "information and relationships in the content" is preserved? For testing/quantification, do we mean "all" or "some" information and relationships? Do we mean "font-size" for text size?
-	I like MC's suggestion. A limit would need to be defined up to a 'normal' level of magnification (responsibility) if that could be a measurable and agreed upon thing.
-	I do like the *direction* this is heading. I agree that we have to find the line between UA and authors' responsibilities-- and perhaps also the line between conventional UAs (whatever *that* means!) and screen magnifiers. I switched from ZoomText to JAWS when I had to increase magnification to something like 4x-- I just felt I lost too much context to keep track of what I was reading, especially with all the automatic scrolling that ZoomText was doing. I don't know how an author could have constructed the content such that all information and relationships conveyed through presentation were preserved when the screen displayed only a few words at a time. So maybe the suggestion "Fonts can be scaled" says all we need.
-	I agree with Ben that font scaling is a UAAG issue. I am also concerned about how we could create a testable SC about scaling layout that will work in the variety of UA and AT available.

 <<font-report-22aug06.html>> Comment LC-564
Sort Terms: CONTRAST FONT
Document: WCAG 2.0 Guidelines
Submitter: Roger Hudson <rhudson@usability.com.au>
Comment Type: substantive
Location: visual-audio-contrast <http://www.w3.org/TR/WCAG20/complete.html>  
Comment:
Part of Item: 
Comment Type: TE 
Comment (including rationale for proposed change): 

Guideline 1.4 is concerned with making it easy to distinguish foreground information from the background. Many people with impaired vision, including a significant proportion of the older population, often find it hard to read the default text size. The use of absolute values for font sizes can make it very difficult for people with impaired vision to increase the size of the text on the screen. Sometime these people have the knowledge and opportunity to change the default computer settings but this is often not the case. 

Proposed Change: 

I suggest Guideline 1.4 should contain a new level 2 Success Criteria that requires the use of relative rather than absolute values to control the size of all text including headings.
Status: open
Working Group Notes: [TEAMA] [HOLD] 

Relates to LC-469 

Submitted by Team A 5/23/06 

close with 
{not accept} 
"It was determined that this was a user agent issue. Absolute font sizes are just advice to user agents who can override them and scale fonts at user request. We do however include an advisory technique "Providing resizable text (future link)" under guideline 1.3. 

Put on hold 25 May 2006 (http://www.w3.org/WAI/GL/2006/05/25-wai-wcag-minutes.html)
Resolution Working Notes - Unapproved:
Related Issues: 
Assigned To: Nobody
Last Edited: 2006-10-10 20:44:07
Edit Comment <http://www.w3.org/2006/02/lc-comments-tracker/35422/wcag20-lc/564> 

Received on Monday, 23 October 2006 22:44:01 UTC