SC 1.4.11 Graphics Contrast: should exclude images of text?

[spanchang]

Is it the expectation that an essential image with text that passes 1.4.3 should be again subjected to a test under SC 1.4.11?
All contrast related SCs in WCAG 2.0 explicitly refer to images of text. Should this new SC not clearly convey whether images of text are covered or not?
I raised this in Issue 9 but the response was off the mark / inadequate.
Thanks

[johnfoliot]

Sailesh wrote:

Is it the expectation that an essential image with text that passes 1.4.3

should be again subjected to a test under SC 1.4.11? Hi Sailesh, Is there something in the current Draft Text that leads you to believe that evaluators would feel that way? The current draft text states; The visual presentation of graphical objects that are essential for understanding the content or functionality have a contrast ratio <https://www.w3.org/TR/WCAG21/#dfn-contrast-ratio> of at least 4.5:1 against the adjacent color(s), except for the following... ​When I reference the draft Understanding documentation ( https://www.w3.org/WAI/WCAG21/Understanding/21/graphics-contrast.html), it appears to me to be fairly clear that this is speaking about images that are triggers (eg: icons, etc.) as well as visible focus indication. It strikes me that while 1.4.3 speaks to legibility and readability of text burned into images, this SC is instead focused on being able to also see and understand *functionality* (which is why actionable icons and visible focus is critical here). My personal feeling is that the Understanding text makes this fairly clear, however if you feel that it is not as such, do you have any proposed text that could be added to the Understanding document to make this clearer? JF

…

On Wed, Oct 11, 2017 at 9:11 AM, spanchang ***@***.***> wrote: Is it the expectation that an essential image with text that passes 1.4.3 should be again subjected to a test under SC 1.4.11? All contrast related SCs in WCAG 2.0 explicitly refer to images of text. Should this new SC not clearly convey whether images of text are covered or not? I raised this in Issue 9 but the response was off the mark / inadequate. Thanks — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#525>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABK-c1-S-H3iweFZ3P-QmJRxfIsXEJraks5srMx5gaJpZM4P1isx> .

-- John Foliot Principal Accessibility Strategist Deque Systems Inc. john.foliot@deque.com Advancing the mission of digital accessibility and inclusion

[spanchang]

Hello John, In Issue 9 as well as in this Github issue I suggested explicitly excluding 'images of text' so there is no ambiguity and room for an alternate interpretation. I also stated other contrast SCs mention 'images of text'. One need not look at the Understanding doc to determine its scope. For instance the print image example. That image too may function like a trigger. Any objection for not making it explicit via's the SC's wording?

…

On 10/12/17, John Foliot ***@***.***> wrote: Sailesh wrote: > Is it the expectation that an essential image with text that passes 1.4.3 should be again subjected to a test under SC 1.4.11? Hi Sailesh, Is there something in the current Draft Text that leads you to believe that evaluators would feel that way? The current draft text states; The visual presentation of graphical objects that are essential for understanding the content or functionality have a contrast ratio <https://www.w3.org/TR/WCAG21/#dfn-contrast-ratio> of at least 4.5:1 against the adjacent color(s), except for the following... ​When I reference the draft Understanding documentation ( https://www.w3.org/WAI/WCAG21/Understanding/21/graphics-contrast.html), it appears to me to be fairly clear that this is speaking about images that are triggers (eg: icons, etc.) as well as visible focus indication. It strikes me that while 1.4.3 speaks to legibility and readability of text burned into images, this SC is instead focused on being able to also see and understand *functionality* (which is why actionable icons and visible focus is critical here). My personal feeling is that the Understanding text makes this fairly clear, however if you feel that it is not as such, do you have any proposed text that could be added to the Understanding document to make this clearer? JF On Wed, Oct 11, 2017 at 9:11 AM, spanchang ***@***.***> wrote: > Is it the expectation that an essential image with text that passes 1.4.3 > should be again subjected to a test under SC 1.4.11? > All contrast related SCs in WCAG 2.0 explicitly refer to images of text. > Should this new SC not clearly convey whether images of text are covered > or > not? > I raised this in Issue 9 but the response was off the mark / inadequate. > Thanks > > — > You are receiving this because you are subscribed to this thread. > Reply to this email directly, view it on GitHub > <#525>, or mute the thread > <https://github.com/notifications/unsubscribe-auth/ABK-c1-S-H3iweFZ3P-QmJRxfIsXEJraks5srMx5gaJpZM4P1isx> > . > -- John Foliot Principal Accessibility Strategist Deque Systems Inc. ***@***.*** Advancing the mission of digital accessibility and inclusion -- You are receiving this because you authored the thread. Reply to this email directly or view it on GitHub: #525 (comment)

-- Sailesh Panchang Principal Accessibility Consultant Deque Systems Inc Phone 703-225-0380 ext 105 Mobile: 571-344-1765

[johnfoliot]

Hi Sailesh, I think you are conflating issues here. Imagine the following use-case: Developer has a graphical image of a round (circular) button, and burned into that button is text. My understanding is that: - SC 1.4.3 is there to ensure that the text burned into the button image meets contrast ratios - Proposed SC 1.4.11 is there to ensure that when that button image takes visible tab focus, that the focus indication is also meeting sufficient color contrast So, yes, in my example use-case, there are two contrast requirements. In Issue 9, Alastair also respond to your other concern: The reasoning was that part of being ‘essential’, means that you have to discern the graphical object to understand the content/functionality. If there is text that says the same thing (and meets WCAG), then the graphic is not essential. Is your question then ​ ​for the use-case where the text that is part of the image file meets 1.4.3, but the same image file also features an icon that is dove gray on white (less than 3.1:1) ​? If yes, I believe Alastair's response still stands: the dove-gray part of the image (the non-text part) is no longer "essential", as the burned-in text meets the requirement. The Understanding document states: The term "graphical object" is intended to apply to stand-alone icons such as a print icon <em>(with no text)</em>, and the important parts of a more complex diagram such as each line in a graph. Not every graphical object needs to have sufficient contrast with its surroundings, only those that are required to understand what the graphic is conveying. Taking another of your concerns into consideration ("One need not look at the Understanding doc to ​ determine its scope.") ​, would providing a definition of Graphical Object in the SC and Glossary address the concern? (i.e. take the text I quoted from the Understanding document, and turn it into a Glossary entry for "Graphical Object"?) ​If yes, that should be logged as a new issue in Github. JF On Thu, Oct 12, 2017 at 11:49 AM, spanchang <notifications@github.com> wrote:

…

Hello John, In Issue 9 as well as in this Github issue I suggested explicitly excluding 'images of text' so there is no ambiguity and room for an alternate interpretation. I also stated other contrast SCs mention 'images of text'. One need not look at the Understanding doc to determine its scope. For instance the print image example. That image too may function like a trigger. Any objection for not making it explicit via's the SC's wording? On 10/12/17, John Foliot ***@***.***> wrote: > Sailesh wrote: > >> Is it the expectation that an essential image with text that passes 1.4.3 > should be again subjected to a test under SC 1.4.11? > > Hi Sailesh, > > Is there something in the current Draft Text that leads you to believe that > evaluators would feel that way? The current draft text states; > > The visual presentation of graphical objects that are essential for > understanding the content or functionality have a contrast ratio > <https://www.w3.org/TR/WCAG21/#dfn-contrast-ratio> of at least 4.5:1 > against the adjacent color(s), except for the following... > > ​When I reference the draft Understanding documentation ( > https://www.w3.org/WAI/WCAG21/Understanding/21/graphics-contrast.html), it > appears to me to be fairly clear that this is speaking about images that > are triggers (eg: icons, etc.) as well as visible focus indication. > > > It strikes me that while 1.4.3 speaks to legibility and readability of text > burned into images, this SC is instead focused on being able to also see > and understand *functionality* (which is why actionable icons and visible > focus is critical here). > > My personal feeling is that the Understanding text makes this fairly clear, > however if you feel that it is not as such, do you have any proposed text > that could be added to the Understanding document to make this clearer? > > JF > > On Wed, Oct 11, 2017 at 9:11 AM, spanchang ***@***.***> > wrote: > >> Is it the expectation that an essential image with text that passes 1.4.3 >> should be again subjected to a test under SC 1.4.11? >> All contrast related SCs in WCAG 2.0 explicitly refer to images of text. >> Should this new SC not clearly convey whether images of text are covered >> or >> not? >> I raised this in Issue 9 but the response was off the mark / inadequate. >> Thanks >> >> — >> You are receiving this because you are subscribed to this thread. >> Reply to this email directly, view it on GitHub >> <#525>, or mute the thread >> <https://github.com/notifications/unsubscribe-auth/ABK-c1-S-H3iweFZ3P- QmJRxfIsXEJraks5srMx5gaJpZM4P1isx> >> . >> > > > > -- > John Foliot > Principal Accessibility Strategist > Deque Systems Inc. > ***@***.*** > > Advancing the mission of digital accessibility and inclusion > > > -- > You are receiving this because you authored the thread. > Reply to this email directly or view it on GitHub: > #525 (comment) -- Sailesh Panchang Principal Accessibility Consultant Deque Systems Inc Phone 703-225-0380 ext 105 <(703)%20225-0380> Mobile: 571-344-1765 <(571)%20344-1765> — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#525 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABK-c4LaFAdHBZESGawRt_Sc1HgpWxI9ks5srkMqgaJpZM4P1isx> .

-- John Foliot Principal Accessibility Strategist Deque Systems Inc. john.foliot@deque.com Advancing the mission of digital accessibility and inclusion

[spanchang]

John, You are conceding that an image of text is indeed subjected only to 1.4.3 and not 1.4.11 and seem to be confusing the issue more with: <start> It appears to me to be fairly clear that this is speaking about images that are triggers (eg: icons, etc.) as well as visible focus indication. - SC 1.4.3 is there to ensure that the text burned into the button image meets contrast ratios - Proposed SC 1.4.11 is there to ensure that when that button image takes visible tab focus, that the focus indication is also meeting sufficient color contrast So, yes, in my example use-case, there are two contrast requirements. </end> You are referring to a graphical interactive control with text that is subjected to and let us say has passed 1.4.3. Now I suppose 1.4.12 deals with contrast of UI controls like this one when it gets tab focus as you state in your example. So where does 1.4.11 come in for this element? The print icon with no text in it (in the Issue 9 example)is excluded precisely because it has no text. So what's the problem in saying via 1.4.11's SC text that it does not cover images of text (a defined term in WCAG 2.0)? Why is there a need to define a new term when one can stick a defined term in the SC to exclude it from scope ... just like that term is used in 1.4.3 to include it in the scope there? Sailesh

…

On 10/12/17, John Foliot ***@***.***> wrote: Hi Sailesh, I think you are conflating issues here. Imagine the following use-case: Developer has a graphical image of a round (circular) button, and burned into that button is text. My understanding is that: - SC 1.4.3 is there to ensure that the text burned into the button image meets contrast ratios - Proposed SC 1.4.11 is there to ensure that when that button image takes visible tab focus, that the focus indication is also meeting sufficient color contrast So, yes, in my example use-case, there are two contrast requirements. In Issue 9, Alastair also respond to your other concern: The reasoning was that part of being ‘essential’, means that you have to discern the graphical object to understand the content/functionality. If there is text that says the same thing (and meets WCAG), then the graphic is not essential. Is your question then ​ ​for the use-case where the text that is part of the image file meets 1.4.3, but the same image file also features an icon that is dove gray on white (less than 3.1:1) ​? If yes, I believe Alastair's response still stands: the dove-gray part of the image (the non-text part) is no longer "essential", as the burned-in text meets the requirement. The Understanding document states: The term "graphical object" is intended to apply to stand-alone icons such as a print icon <em>(with no text)</em>, and the important parts of a more complex diagram such as each line in a graph. Not every graphical object needs to have sufficient contrast with its surroundings, only those that are required to understand what the graphic is conveying. Taking another of your concerns into consideration ("One need not look at the Understanding doc to ​ determine its scope.") ​, would providing a definition of Graphical Object in the SC and Glossary address the concern? (i.e. take the text I quoted from the Understanding document, and turn it into a Glossary entry for "Graphical Object"?) ​If yes, that should be logged as a new issue in Github. JF On Thu, Oct 12, 2017 at 11:49 AM, spanchang ***@***.***> wrote: > Hello John, > In Issue 9 as well as in this Github issue I suggested explicitly > excluding 'images of text' so there is no ambiguity and room for an > alternate interpretation. I also stated other contrast SCs mention > 'images of text'. One need not look at the Understanding doc to > determine its scope. For instance the print image example. That image > too may function like a trigger. Any objection for not making it > explicit via's the SC's wording? > > > On 10/12/17, John Foliot ***@***.***> wrote: > > Sailesh wrote: > > > >> Is it the expectation that an essential image with text that passes > 1.4.3 > > should be again subjected to a test under SC 1.4.11? > > > > Hi Sailesh, > > > > Is there something in the current Draft Text that leads you to believe > that > > evaluators would feel that way? The current draft text states; > > > > The visual presentation of graphical objects that are essential for > > understanding the content or functionality have a contrast ratio > > <https://www.w3.org/TR/WCAG21/#dfn-contrast-ratio> of at least 4.5:1 > > against the adjacent color(s), except for the following... > > > > ​When I reference the draft Understanding documentation ( > > https://www.w3.org/WAI/WCAG21/Understanding/21/graphics-contrast.html), > it > > appears to me to be fairly clear that this is speaking about images > > that > > are triggers (eg: icons, etc.) as well as visible focus indication. > > > > > > It strikes me that while 1.4.3 speaks to legibility and readability of > text > > burned into images, this SC is instead focused on being able to also > > see > > and understand *functionality* (which is why actionable icons and > > visible > > focus is critical here). > > > > My personal feeling is that the Understanding text makes this fairly > clear, > > however if you feel that it is not as such, do you have any proposed > > text > > that could be added to the Understanding document to make this clearer? > > > > JF > > > > On Wed, Oct 11, 2017 at 9:11 AM, spanchang ***@***.***> > > wrote: > > > >> Is it the expectation that an essential image with text that passes > 1.4.3 > >> should be again subjected to a test under SC 1.4.11? > >> All contrast related SCs in WCAG 2.0 explicitly refer to images of > >> text. > >> Should this new SC not clearly convey whether images of text are > >> covered > >> or > >> not? > >> I raised this in Issue 9 but the response was off the mark / > >> inadequate. > >> Thanks > >> > >> — > >> You are receiving this because you are subscribed to this thread. > >> Reply to this email directly, view it on GitHub > >> <#525>, or mute the thread > >> <https://github.com/notifications/unsubscribe-auth/ABK-c1-S-H3iweFZ3P- > QmJRxfIsXEJraks5srMx5gaJpZM4P1isx> > >> . > >> > > > > > > > > -- > > John Foliot > > Principal Accessibility Strategist > > Deque Systems Inc. > > ***@***.*** > > > > Advancing the mission of digital accessibility and inclusion > > > > > > -- > > You are receiving this because you authored the thread. > > Reply to this email directly or view it on GitHub: > > #525 (comment) > > > -- > Sailesh Panchang > Principal Accessibility Consultant > Deque Systems Inc > Phone 703-225-0380 ext 105 <(703)%20225-0380> > Mobile: 571-344-1765 <(571)%20344-1765> > > — > You are receiving this because you commented. > Reply to this email directly, view it on GitHub > <#525 (comment)>, or > mute > the thread > <https://github.com/notifications/unsubscribe-auth/ABK-c4LaFAdHBZESGawRt_Sc1HgpWxI9ks5srkMqgaJpZM4P1isx> > . > -- John Foliot Principal Accessibility Strategist Deque Systems Inc. ***@***.*** Advancing the mission of digital accessibility and inclusion -- You are receiving this because you authored the thread. Reply to this email directly or view it on GitHub: #525 (comment)

-- Sailesh Panchang Principal Accessibility Consultant Deque Systems Inc Phone 703-225-0380 ext 105 Mobile: 571-344-1765

[alastc]

An "an essential image with text" is not possible, if it has equivalent visible text, it is not essential.

You are conceding that an image of text is indeed subjected only to 1.4.3 and not 1.4.11

That is a good thing, we are supposed to avoid overlap as much as possible.

The print icon with no text in it (in the Issue 9 example)is excluded precisely because it has no text.

I don't understand what would give you that impression from the current SC text or understanding document?

Think of it this way for the print icon that is a link with no text:

• The 'content' graphic for the icon comes under graphics contrast, i.e. the outline of the printer. Note that the border of that graphical object is not the same as the border for the control.
• Any border around the control that visually indicates that it is a control comes under component contrast, as does the focus styling.

So what's the problem in saying via 1.4.11's SC text that it does not cover images of text (a defined term in WCAG 2.0)?

We have a pretty long SC text as it is, if we say everything it doesn't include (when it does say what it does include), we'll have a very long SC text.

[awkawk]

@spanchang Can you confirm if you are ok with @alastc response and that this issue is ok to close?

[spanchang]

Andrew, I responded the same day to your email of Nov 5, 2017 re. issue 525, 533, 534. Please can you please review? I resent it again today. Thanks, Sailesh

…

On 11/9/17, Andrew Kirkpatrick ***@***.***> wrote: @spanchang Can you confirm if you are ok with @alastc response and that this issue is ok to close? -- You are receiving this because you were mentioned. Reply to this email directly or view it on GitHub: #525 (comment)

-- Sailesh Panchang Principal Accessibility Consultant Deque Systems Inc Phone 703-225-0380 ext 105 Mobile: 571-344-1765

[alastc]

Hi Sailesh, could you post those here or to the list please? I can't see anything from you regarding this.

[spanchang]

The responses I got on Oct 12 seems to indicate there are others too who are not 100% sure of the scope of the SC based on the normative language and believe there is something lacking.
Are there situations when 1.4.3 needs to be applied to a graphical object (or graphical element as referred by the main intent statement in the Understanding doc) and then apply SC 1.4.11 to the same graphical object or element?
The SC's scope should be clear from the normative language and the Understanding doc should inter alia illustrate when both SCs might need to be applied.

[alastc]

Hi @spanchang, if the core comment is the same, could you close this issue and continue the discussion in #534 please?

[spanchang]

I have been urged by Andrew since Nov 5 to close issues 525, 533 and 534.
On Nov 5 I responded to him with:
I raised issue 525 hoping to start a discussion. But after the
responses, I decided to raise specific issues: one about the content
in the Understanding doc and one about the SC's wording.

On Nov 10, on learning a new SC draft is on its way, I responded with:
I would like to review the 3 issues in the light of the new SC text
... if it is on its way, and then respond.

Based on response to my comment of Nov 12 on issue 534, I am closing #525.