WCAG Comments through Aug 2013
Nav: EOWG wiki main page
More recent comments are at WCAG review
How to Meet
- comment {name}
Techniques
- comment {name}
G140 delete Example 2
note: we'll probably submit this by e-mail or text fields in comment form so format as plain text, please. :-)
- comment {name}
- I don't think I'm up for rewriting a complete draft 3 - not sure if this is what's being requested. However, I will put in my 2 cents and say that a tagged pdf is a poor example for a number of reasons. One reason, as mentioned by Wayne below, is that there is not full separation between the information|structure and the presentation. Just as inline formatting or styles embedded in the header of an html file fails to fully separate content from formatting, the same is true of a tagged pdf. External style sheets allow quick and easy modification to the presentation of html content in order to accommodate the visual, perceptual, and cognitive needs of a variety of users. Internal styling or an internal style sheet does not provide this flexibility or control by the user. The presentation of a tagged pdf cannot be easily changed or customized by or for the needs of various audiences. In many ways styles in Word is a better example of separation of content|structure from presentation: the styles can at least be easily switched and modified depending on the needs of the audience.{Howard, 21 August}
- I agree that PDF is not a good example. Perhaps we could use as an example a text document created with Free Office or Word and well formatted. Or a widget which also uses CSS to separate the presentation of the structure and functionality. Emmanuelle Gutiérrez y Restrepo
Overall Draft 3
Removal of Example 2: Tagged PDF from G140 Description.
Issue: “Example 2: Tagged PDF” confuses rather than clarifies the concept of separating information and structure from presentation. The text of the example describes content that is partially separated, not fully separated. In particular, the example describes blobs of data in PDF content consisting of, “mostly of the content embedded with formatting information.” This diverges markedly from the model that most web developers associate with separating information and structure from presentation. In addition, it does not appear to support one key objective of the G140: The ability to, "Modify the presentation of content by substituting alternate presentation rules attached to structural features."
Motivation: Most web professionals know the separation principle as a foundational concept of web standards. In fact, this concept motivated the invention of markup language starting with SGML forty years ago. When web professionals think of separating information and structure from presentation their mental model looks like the CSS Zen Garden of Dave Shea —clean HTML in one file separated from CSS in another. Now, this is not the only model of data separation, but it is the model that most web professionals will have in mind when they first approach the problem.
If “Example 2: Tagged PDF” is to be a successful teaching example that is included in the WCAG 2.0 Techniques, it must bridge the gap between the objectives of the CSS Zen Garden model and the same techniques using PDF. The example must demonstrate how to modify the presentation of content by substituting alternate presentation rules attached to structural features. Until the author of "Example 2: Tagged PDF" can accomplish this task, the example should be removed from the WCAG 2.0 Techniques.
PDF has inline elements that express text level semantics. The example should show how the PDF inline elements can be used effectively to modify presentation. Web developers already know how well in-line, text-level, semantic elements work for HTML. They know how to use elements like (cite, code, dfn, em, kbd, samp, strong, sub, sup and var) to modify presentation rules. The CSS Zen Garden is full of these examples. Web developers want to know how they can do the same things with PDF. Example 2 doesn’t show how these transformations can be programmatically determined.
=====Sample Problem: The solution to a problem like this should be demonstrated in Example 2.=====
The Modern Language Association (MLA) Style Guide requires that the reference to a journal must appear in the italic variant of the running text in an article. It is well documented, that the italic format variant poses difficultly for most people with print disabilities. In HTML one can change this presentation. The running text (text in paragraphs etc.) could be changed to the APHont font from the American Printing House for the Blind. The emphasized text could be changed to a typewriter font like Courier Bold. This would be much more readable and it would give the reader with a print disability access equal access to the denotational semantics of the text and the structural semantics indicated by the formatting. How could this Zen Garden transformation be supported with PDF? A reasonable web programmer with little experience in accessibility but lots of experience in web standards would like to know how this is done. Note: This is an important example because PDF is used so extensively to transmit professional articles on the web.
Figure 1: A bibliographic item in standard MLA format
Figure 2: The same bibliographic item transformed in the Zen Garden style. Until Example 2 in G140 can explain how this kind of transformation can be performed, the example should be removed from the explanation of G140
Prepared by Wayne Dick for EOWG
Confusion Draft 2
Removal of Example 2: Tagged PDF from G140 Description.
Issue: “Example 2: Tagged PDF” confuses rather than clarifies the concept of separating information and structure from presentation. The text of the example describes content that partially separated, not fully separated. In particular, the blobs of data in PDF content are described as consisting of, “mostly of the content embedded with formatting information.” This description does not appear to support the one key objective of the G140: the ability to, "Modify the presentation of content by substituting alternate presentation rules attached to structural features."
Motivation: Most web professionals know the separation principle as a foundational concept of web standards. In fact, this concept motivated the invention of markup language starting with SGML forty years ago. When web professionals think of separatiing of information and structure from presentation their mental model looks like the "CSS Zen Garden" of David Shea, clean HTML in one file separated from CSS in another. Now, this is not the only model of data separation, but it is the model that most web professionals will have learned when they first approach the problem.
If “Example 2: Tagged PDF” is to be a successful teaching example that is worthy of inclusion in the WCAG 2.0 Techniques, it must bridge the gap between objectives of the CSS Zen Garden model and the same techniques using PDF. In the words of G140 it must demonstrate how to modify the presentation of content by substituting alternate presentation rules attached to structural features (one of three goals of G140).
Here is an example: Both HTML and PDF have a structure citing bibliographic entries. In HTML it is “cite” element and PDF it is “BibEntry” tag. Now suppose the citation is formatted in an italic variant of the running font of an article. It is well documented, that italic format variants pose difficultly for most people with low vision. In the "Zen Garden" world (HTML+CSS) one can change this presentation. The running text (text in paragraphs etc.) could be changed to the APHont font from the American Printing House for the Blind. The citation font could be changed to a typewriter font line Courier Bold with no italics. This would be much more readable and it would give the reader with low access equal access to the denotational semantics of the text (information) and the structural semantics indicated by the formatting (presentation). How could this "Zen Garden" transformation be supported with PDF? A reasonable web programmer with little experience in accessibility but lots of experience in web standards would like to know how this is done. Example 2 does not show this type of transformation and causes confusion.
Example 2 should be removed until the author can explain how PDF can fulfill this objective.
Cannot change presentation by structure
Rationale: Technique G140 <http://www.w3.org/WAI/GL/2013/WD-WCAG20-TECHS-20130711/G140.html> says: "Following this technique allows user agents to: ... Modify the presentation of content by substituting alternate presentation rules attached to structural features." Currently no PDF viewers allow users to modify the presentation by structure, e.g., users cannot change the presentation of headings. PDF is currently not sufficiently accessibility-supported given limitations in PDF viewers <http://www.tader.info/support.html#PDFisNOTaccessible>. Including PDF as an example for G140 perpetuates the myth that PDF is sufficiently accessible.
Confusion Draft 1
G140: Separating information and structure from presentation to enable different presentations
Issue: Example 2 should be clarified to explain how software can detect the semantic information conveyed in content that “consists mostly of the content embedded with formatting information.”
Note: The description about how PDF tags are used to support navigation and identification of some structures in Example 2 is quite good. The example, however, seems to be presenting a data format that is partially separated not fully separated. That is, PDF appears to be blobs of scrambled information and presentation punctuated with tag that do not appear to separate all presentation from information. Significant clarification is needed.
Basic Question: How is separation of this formatting (presentation) form text (information) achieved? The initial sentence appears to imply that mixing formatting and text in a blob supports separation of presentation from information. This example differs so radically from traditional understanding of the separation principle as demonstrated in publications like the, CSS Zen Garden, that a reasonable web developer would find it extremely confusing.
Discussion: The technique G140 is very specific. It talks about separating both “information and structure” from presentation. It does not say that this is for the purpose of supporting navigation, or specific requirements of screen reader software. It just states that information and structure should be separated from presentation. Any assistive technology can do what is needed with this separation to provide accessibility support for its target population. G140 just states that separation must exist not why.
The most important operation one undertakes on a web page is reading. Reading uses text as its base information. Formatting is presentation. Depending on its usage, formatting can augment the meaning of text. This means that content should be organized so that formatting and text can be separated without loss of the denotational meaning of text or the usage meaning of formatting.
The objective of G140 is to prevent formatting, position and layout from being used to designate information or structure that cannot be recognized by assistive technology. When formatting is present there should be semantic markers within the content so the usage of formatting can be unambiguous determined by a program. Otherwise information and presentation are inseparable.
This brings us to “Example 2: Tagged PDF”. There is apparent lack of semantic information needed to disambiguate the, “content embedded with formatting information.” Are there markers within these blobs of mixed formatting and text that define the meaning of formatting? How does software identify the meaning of formatting? The example does not explain this. In fact, this first sentence appears to be defining a format that does not separate information (text) from presentation (formatting). When an HTML author uses italics there are semantic markers (elements) to designate whether this formatting is for a citation, mathematical variable, a definition term or simple emphasis. How does PDF use tags to disambiguate this information? Example 2 does not address this.
Am I mistaken? Does the WCAG WG not consider text to be information and formatting to be presentation? In English text is a form of information and formatting is a form of presentation. Does WCAG use a special jargon where text is not information and formatting is not presentation?
If I don’t understand this subtlety, how is a programmer with limited accessibility training going to understand it? As mentioned earlier, most web programmers think of the CSS Zen Garden model when they think of separating information from presentation. Clearly PDF does not use this model. So, how does it do the same thing? With an information model that is so radically different from the normal way separation is understood, the PDF example needs to be explained more completely. WCAG WG really must explain exactly how the text / formatting blobs can be separated before they can reasonable use this as a teaching example. As it stands, it is profoundly confusing.
Question: Where does this Note fit better?
Comments below submitted 12 August 2013
Understanding Techniques for WCAG Success Criteria has a note that starts out: "Note 1: W3C cautions against requiring..." It's an important point and we want to make sure people read it. Currently it is in the Techniques are Informative section. Some think it would be better in the Sufficient Techniques section (right before the heading "Numbered Lists, "AND""). Thoughts?
- I'm one who thinks it wouod be better in the Sufficient Techniques section - they're what we're referring to {Andrew, 2/Aug}
- I agree with Andrew, it would be easier to read in the "suffiscient techniques" section. {Sylvie}
- I feel that it belongs in the "Techniques are Informative" section because it's broadly about not requiring the Techniques, rather than specifically about the sufficient techniques. (although I'm not set on this) {Shawn}
- I feel that it would sit better under the Sufficient Techniques section and most naturally just before the para starting "There may be other ways ..." and without being marked as a note. {Bim, Aug 2}
- I agree with Shawn, but it may be useful to add a reminder on Sufficient Techniques section.{Emmanuelle}
The Sufficient Techniques section currently has "(See also Techniques are Informative above.)" so it generally points to that section, though not specifically to that note. - I also agree that the information would be best if it was in the "Sufficient Techniques" section. Can we duplicate the info? I believe it wouldn't hurt to also mention it in the "Techniques are informative" section. But my first choice would be "Sufficient Techniques". {dboudreau, Aug4th.}
- I feel that for those unitiated into the special language of standards it will be a somewhat confusing and meaningless sentence. If it is intended for ordinary people it would be nice to have an ordinary language version so that they can truly understand the balance between normative and informative. Perhaps there could be a link to a plain text easy to understand version?{Suzette 5th August}
- I think the Notes break the flow but I do think a point made in both sections would carry the message forward. As such, a suggestion follows:{Vicki, August 9}
- remove the heading “Techniques are informative” (see Alan’s comment)
- remove the words “Note 1” and “Note 2” (see Catherine, Kathleen and Sailesh's comments).
- Replace the words "Note 1" with a sentence which follows from the italicized text (contributes also to - following EOWG terminology - tersification :)
- still keep info under "Sufficient Techniques" (many comments)
- Move "Note 2" as a sentence into "Sufficient Techniques" with a reminder that the techniques are for guidance:
- remove the brackets with reference to Techniques are informative,
- add a reminder as a sub-heading or an important note before "Advisory Techniques" e.g:
- Sufficient techniques are provided as guidance. A frequent misunderstanding is that they should be used for meeting conformance. The only thing that should be required is meeting the WCAG 2.0 success criteria and not the techniques which are informative. There can be negative consequences of allowing only W3C's published techniques to be used for conformance to WCAG 2.0.
- Techniques for WCAG 2.0 use the words "must" and "should" only to clarify guidance within the techniques, not to convey requirements for WCAG
- I like it where it is. Similar to other comments perhaps change "Note 1" to "Important Note".{Howard, 15 Aug}
- It is ok where it is but should be worded as "It is important to note that ...", instead of just "Note:". Notes like those are generally considered supplementary / advisory info and can be missed easily. Alternatively it should be moved up in that section nearer to the beginning and not be called a "Note". {Sailesh}
- I agree with Sailesh in that it could stay where it is but needs to stand out more. If it is moved to the sufficient techniques section, it should still be made to stand out. I think the idea that W3C cautions against something is a pretty strong statement and it is important that it not be missed. Perhaps that sentence or the words "cautions against" should be marked up in strong. {Catherine}
- I agree with Andrew that the "Techniques are Informative" section refers to "Sufficient Techniques."
Advisory Techniques and Failures are by nature not required. I assume that using a separate section is to emphasize the notes, but on first reading I found the section heading a little confusing, especially as it's followed by so many other headings, all with the word "Techniques." It rather upsets the flow of ideas to have a disclaimer as the first section.
I think it would be more coherent to make it a subsection of "Sufficient Techniques." {Alan} - "Alternatively it should be moved up in that section nearer to the beginning and not be called a "Note"." I agree. {Kathleen}
Understanding
- comment {name}