Education and Outreach Working Group Teleconference

12 Sep 2014

See also: IRC log


Shadi, +1.615.417.aaaa, EricE, Shawn, +1.512.731.aabb, kevin, +1.202.276.aacc, AnnaBelle, +1.562.256.aadd, Wayne, Jan, Sharron, PaulSchantz, Wilco, metzessive


<trackbot> Date: 12 September 2014

<scribe> Scribe: Sharron

<AnnaBelle> \me Yeah. Will miss you guys!

Implementing Web Accessibility

Shawn: This is ready for complete and thorough review. Use GitHub if you re comformatble with that. or email or wiki.
... anyone have discussion of that?

Kevin: Very open for any comments, suggestions, etc

Shawn: So everyone add to to-do

WCAG-EM Report Tool

Shadi: We got just a couple of comments, thanks to those of you who got people to use it, we got quite a few comments. Most are not yet implemented
... I have tried to make clear those items that have been discussed and that are scheduled for implementation. We need to discuss three comments. Link from agenda to the wiki to find those comments
... let's look at second comment,

<shawn> borders & shading https://www.w3.org/WAI/EO/wiki/WCAG-EM_Report_Tool_Comments#borders

<shadi> http://w3c.github.io/wcag-em-report-tool/dist/#/audit/scope

Shadi: OK Let's look at first comment, "borders and shading." If you look at the input area, Wilco has done one round of emphasis so now we should consider if this is good.

Shawn: The areas look kind of 3-D. I am used to the background with shading but the field itself being entirely white.
... the gradient gives it a look of a button.

Paul: None of the visual aspects of this has come up in my testing with others. In fact I got very good feedback on the overall look.

Shadi: Typically the editable aspect would be white to indicate that it can be filled

Wayne: I don't think it is the color, but the gradient

Shawn: You could have a darker border with the text box area being white

Shadi: The text area is so large that we wanted to indicate that focus is within the text box regardless of zoom
... are you then suggesting making borders larger and darker. Wilco?

Wilco: I don't see any problem.

Paul: I have a question about when focus comes into the filed, could the border not change color

Shadi: It is the way it works now with this one.

Wayne: I like how it is now

Paul: I do too, but the color change in the border can be more obvious

Wilco: Changing the background color will have an impact on more than this aspect.

Paul: The light background color was requested for an app we built at CSUN as an alternative to plain white which is too bright for some visual disabilites
... it is a subtle change but makes a great deal of difference

<shadi> http://www.w3.org/WAI/tutorials/

<metzessive> Sorry Sharron. I just saw this. It was me.

Shadi: The very faint horizontal border will suffer the most. If we can change that would it wok to make the background darker?
... one thing I notice on the tutorials is this very thin and unobtrusive red line at the top. Might that be useful?

Shawn: The report tool has no W3C logo so branding?

<Zakim> shawn, you wanted to say browser color setting

Shadi: It has W3C colors

Shawn: I was checking to see if this would accept my own application of color, not found it yet, but wanted to agree with Paul's comment about bright white

<shawn> Paul's example http://lists.w3.org/Archives/Public/w3c-wai-eo/2014JulSep/0067.html

AnnaBelle: We need to make note of the fact that we don't have clear connection to W3C. I had mentioned it before but wanted to reintroduce the fact.

<metzessive> metzessive is Jon

Shadi: I am hearing that there is a preference for a slightly shaded page background. That will allow us to have the input field become white and could darken and strengthen the borders. Will come back to the gradient question.

<paulschantz> It's a subtle change, but it really makes a big difference (our app had a white background at first).

<shawn> +1

Shadi: will add slightly darker background, field inputs white, stronger borders if needed.
... and about gradients, Wilco do you want to keep them?

Wilco: Yes I really really like them.

Shadi: Shawn, will this address yourconcern even if we keep the gradient?

Shawn: Yes I beleive with those other additions it should be OK

Shadi: Great, we will try that and is there anything else on this?

<shawn> floating save https://www.w3.org/WAI/EO/wiki/WCAG-EM_Report_Tool_Comments#floatingsave

Shadi: Next question is about the fact that the "Save" button is at the very top.
... as you make your way through the application, you must scroll back up to the nav bar to Save as you collect data. A suggestion has been to have a floating "Save"

<metzessive> Entire menu

<yatil> +1

Jon: Or float the entire menu

<kevin> +1 to perhaps, maybe it being a good idea

<metzessive> +1 from me, and the people who tested

Shadi: That would have the beenfit of having key reources handy

<paulschantz> good point, wayne

Shadi: so that no matter where you are in the process the top nav will be at teh top of your screen

<metzessive> when you say "floated" you actually mean "fixed" right?

Wayne: Big problem for magnifcation

Wilco: We can set it only for certain size screen - we can address that

<metzessive> What about a visibility when the focus shifts back up on the page? Like what happens in Android UI?

Shadi: The entire code of this is online and they are using a branch of this tool in Wilco's organization. So we know it is possible to configure the Auto-save

Wilco: and it is already in there ifyou know how to use it

Wayne: At 200% the top bar takes up about you are losing a lot of screen space
... about 1/6 of the screen and since we only have 50 words on a screen that is huge

<paulschantz> wayne are you using text only or page zoom?

<metzessive> +1 to talk about menu bar

<metzessive> oops q+

Wayne: So we need to really think about how to float and where the break points will be

Jon: When you say floating, do you mean fixed on the top? Is it possible to note when zoom is activated and get rid of icons etc to alleviate the situation? One observation that was made is the length of time it takes to use the tool. The fixed nav would be a good mediation of that.

<Wayne> I will do screen shots.

Jon: See the Android UI as you shift focus down, the navigation disappears, but if you move focus up, the nav becomes visible to the user

Shadi: I understand but I dislike that function personally. Never know when/how it shows and can be confusing

Jon: I will send some links for possible solutions

Shadi: We are running to the end of Wilco's time so we need to determine what is absolutely needed in this version.

<Zakim> shawn, you wanted to mention close & open nav at top?

<metzessive> http://stackoverflow.com/questions/1713771/how-to-detect-page-zoom-level-in-all-modern-browsers

Shawn: Brainstorm - if we have it always visible, could it inlcude a collapse option?

Shadi: Good idea but new feature...Wilco what is the effort?

Wilco: I am experimenting with the idea of float, need to consider screen height but if we inlcude a collapse that may not be an issue

Shadi: So to minimize the function, there are icons needed, placement questions, etc so I am sensing that this may be considerably more work.

Wilco: Yes it will take considerable work

Shadi: Several applications do contain this sort of thing but it is something that needs thinking, planning, testing
... we are likely to have a Control+S for some browsers but what if we don't have the fixed nav here?

Shawn: I am afraid it is significant

<metzessive> I agree with Shawn

<metzessive> yes

<metzessive> i am

<paulschantz> how quickly are you iterating? I ask only half in jest

Jon: When you first see the page, you see the Save or hear it only once with a screen reader. Many of these testing will be doen in between other projects and people don't like to spend more time than they must. Having the save more available by far outweighs the investment

Paul: I would not want to hold up the release but could it come a few weeks down the road?

Shadi: No we have no resources for further iteration, we need to get as far down the road as we can. We are past the time of funding for this but are stretching Wilco's time to make this one the best possible

Paul: It is in GitHub, we could fork it

Wayne: Why not have a save button by each input like on the wiki

Shawn: Maybe on every Guideline level?

Shadi: Will people think that is a localized save?

Shawn: Possibly

Wayne: Vertical scrolling is not as significant as horizontal scrolling but nevertheless is significant in terms of time

<shawn> subtopic: Results for the entire sample

Shadi: Staying on Step 4, the Audit the Sample, it is supposed to be flexible. Some will evaluate page by page, some will provide reports only on the sample level, not detail on every page. This is meant to support all approaches.
... if you want to evaluate on a page-by-page level, the labeling may not be as clear. There is a need to make it clear that the tool will only report what you put in, that there are functions hidden throughout the tool, and that results are customizable.

Shawn: Making sure that people will understand that they may just make a determination for all of the sample, but some will be judging page by page. The result is reported for the full sample and later after having looked at all of them, would make the determination for the whole sample.

Shadi: Step 4 is the most complex because it has so many various ways it might be used. When you are first here...there is SC to evaluate, there are Principles, Guidelines and expandble functions. At the SClevel, the expand is a button on the left hand side
... is that clear?
... if you go to Step 3 and add a couple of pages, then go to Step 4 and will ahve pages loaded
... I worry that "Show Results for Individual Pages" may not be clear and you must know that before you choose that option you are expected to choose those individual pages

Sharron: I agree that the instruction must be made more clear

Shawn: Show Results for Selected Pages?

<Zakim> kevin, you wanted to suggest higlighting

Kevin: That button only becomes useful and purposeful when you have selected something, maybe disble the button until you have chosen individual pages to look at

Wilco: perhaps it wuld be clearer if we make the "Select Sample" to "Choose or Create" sample.

Shadi: Independently of that. Say we had a better name for that...

Shawn: What about Kevin's idea of only making it visible when pages are slected

Shadi: If the button is not visible how would you find that function, how would you know you could display info for slected pages?

Shawn: We may need to have help text...one or two lines that tell you what you can do here.

Wayne: Every other control is more of a combo box, why not align with other controls?
... Just because there are only two selections is this not the same function and could you not make it more of a toggle?

Shadi: Wilco can tell me if there is an option to do something like "Show All"

<Zakim> kevin, you wanted to talk about tutorial

Wayne: To specify only certain pages in this presentation of it, is confusing

Kevin: Having an intro/help text is often done with a popup mini-tutorial that gives you an overview of the function that is available

Shadi: These have been good ideas, thanks for sharing your thoughts, Wilco and I will take this back and see what we can and cannot do.
... there were a few submitted comments that did not entirely make sense to me and I will follow up off line

Wayne: I am sending more screen shots. I think it is remarkable that this application is fully operable at 400%

Shadi: Anything else on this page?

<metzessive> My testers really REALLY liked exporting the document

Shawn: Jon, can you summarize and submit your results with your testers?

Jon: I have a very small sample, but they think the ability to export was extremely helpful. Not much detailed feedback but the length of time to do the testing was more than expected.

Shadi: We have hopes that in future we will be able to add automated test results and have it integrated into the final report


<shawn> http://www.w3.org/WAI/tutorials/tables/

Shawn: We had a comment come in about the text alternatives on the Tables Concept page. I think it is too much to read the email, but Eric will walk us through what is the issue

Eric: The comment was that the small table images have incorrect alt text. For example, the basic tables icon says "Typically you have one row or column..." Some have no alt text because the text that follows is sufficient to describe. The comment was that the approach is inconsistant

Shawn: If you don't understand the issue, notice that the table image in the first one, Basic Tables is rendered as follows:

Eric: The image is not visually where it is in the code. So the alt text in the first one specifies more about the basic tables. It is not only describing the image but is trying to communicate the same info in the flow of how a screen reader would read it.

Sharron: What is the exact objection?

<Zakim> kevin, you wanted to comment on the comment

Eric: That the alt text does not completely describe the image and the inconsistancy of sometimes using empty alt when the inline text is deemed to be sufficient

<Wayne> +1 for Kevin's comment

Kevin: I read the comment. It is not clear why the objection exists. It comes across that it is technically incorrect but it does not make clear what problem it might cause an actual user.

<Wayne> The alternative text is "equally" effective

<shawn> http://lists.w3.org/Archives/Public/wai-eo-editors/2014Sep/0006.html

Shawn: Let's try to figure this out so it does not go for another week. Please take a reading break so we can figure this out.
... There is a technical objection that the presentation does not exactly match how we decribe an approach in the images tutorial. However, there does not seem to be an example of a user who would miss information. That makes it difficult to know how to address the comment.

<metzessive> Is there a TL;DR version?

Shadi: This may be an objection to a percieved bias that WAI optimizes for screen reader users.

<Jan> I have to get on another call - have a great weekend everyone.

Shawn: However the main reason for having the icons was for users with low vision and/or dyslexia and that use case would not be dependent on the text alternative in any case.

Jon: Is the complaint that we have alt text on images that should not have alt text?

Shawn: No I think the complaint is of the fact that the alt text does not precisely describe the iamge but rather is meant to facilitate general understanding of the intent of the image and is further meant to fit into the flow of the accompanying paragraph.

Jon: We (people with cognitive disabilities that affect the ability to process information simultaneously) may be looking at images on the screen, I expect the alt text to describe the image, not to provide context. I think visually. When I talk, I translate your words into pictures, and my pictures into text.

Shawn: So you look at alt text for images?

Jon: Yes, all the time. Sometimes as a translation device to help me understand the realtion to your words and my images. It can clarify a bit more about what you are talking about.

Shadi: OK the first alt text will read "Typically will have one header..."
... so is this a case of reverse information for those who do not use screen readers? Are users of alt text getting more information that those who only see the icon. It may be seen as a decorative image by sighted users.

<Zakim> EricE, you wanted to highlight iconic use of icons

Eric: The thing with the images is that there use iconic and therefore we may not need to provide as much informaiton as we do.

Shawn: So what is the proposed change?

<paulschantz> I do not see a clear request for change in Olaf's email

Eric: The text is also iconic, it presents what was my intention of choosing that icon.

Shadi: If that was the intent, it should be the same to all

Shawn: Info in alt text is more than what is in image

Eric: Alt text tells the same basic information as what is in the image from my POV

Wayne: If I cannot percieve an icon, the alternative text that I want usually is what is the function of the icon, a plain description of the visuals is often insufficient. Second is that whatever you decide is the purpose of an iconic image, it may not be meeting the task.

<metzessive> i believe that we should use alt text to explain what the image is.

Shawn: Where is everyone else on this issue?

<kevin> +1 metzessive

Shawn: Eric do you understand why there may be an objection?

Eric: Yes

<metzessive> -1 that these are decorative.

Shadi: So my opinion is that the icon information is provided inline and that the image should have empty alt

<Wayne> +1

Kevin: I am not quite sure that providing empty alt text meets the need described by Jon

Shawn: So what is your counterproposal?

Jon: They are not decorative. 1st is to explain what's going on in the grey vs white boxes. You want to explain what is going on there. If we did not need to talk about it it would be decorative.

Shawn: The words that accompany the icons, the paragraph text is what would be a description of the boxes, would it be repeated?

Jon: I don't think it is wrong to repeat text as alt and as text. When I look at alt text I want a quick overview of what is going on.

Shawn: Can anyone stay on the call over time?
... For those that have to leave, review Forms Tutorial

Paul: In the context of this page, they are largely decorative and should probably just say "icon representing..."

Wayne: I misunderstood Jon, I change my vote and I like Jon's idea - a lot

Shawn: So you are in favor of getting the redundant information

Wayne: Not redundant so much as repeated

Sharron: I am not in favor of repeating the same alt text as the onscreen text

<metzessive> Sure. It could also be "an example of a basic table", but it could be better IMO if it clarified what was happening.

<metzessive> What about "example of basic table with a row of headers"? Short, to the point?

<shadi> http://www.w3.org/WAI/tutorials/images/decorative/#decorative-image-as-part-of-a-text-link

Shadi: Some have empty alt when the information onscreen is sufficient. We can move the rest into the parapgraph. The question really is whether the icons are illustrative or merely decorative. The reason the comment said it was contradictory is that we are not consistent with each other, we don't describe what they look like, what is the meaning of the white vs shading aspects of the icon. So

there was never an intent for this to be an informative image

Eric: Empty alt would be my preference and putting any additional info in the paragraph

Shawn: What are your thoughts about empty alt being the solution overall?

Jon: Well, you do not have to use the very same text as in the paragraph. Could use: "example of basic table with a row of headers"? and such as a short, to the point explanation of what the icon is meant to illustrate

Sharron: +1 I am OK with this

Wayne: Table for a week

Sharron: For what purpose?

Shawn: It is a challenge at this point we do not have a solution that will be accepted as the clear right answer for everyone.

Wayne: In this instance, both needs can be met. Jon and I will exchange emails. Our goal will be to find the alt text that can fill the needs and not contradict the intent of the tutorial.

<metzessive> agreed with being short and succinct

Shawn: To meet our expectations the alt will have to be very very short and succinct

Shadi: Send it to the list as soon as you have suggestions so others can weigh in

Shawn: For next week what is our action

Sharron: Leave as is until we have a solution

Shawn: We look forward to input and will continue to brainstorm.
... Shadi, Eric can you send the email for the Forms review and the type of review? Thanks all, until next week.

trackbot, end meeting

<shawn> Shawn: Any alt-text increases cognitive load for users who must read alt-text

<yatil> s/<figure><img><figcaption> Example for a basic table </figcaption></figure> ???/

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.138 (CVS log)
$Date: 2014-09-12 15:05:04 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.138  of Date: 2013-04-25 13:59:11  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/Kevin/Paul/
Succeeded: s/Wayn/Wayne/
Succeeded: s/Wilcon/Wilco/
Succeeded: s/hep/help/
Succeeded: s/and/at/
Succeeded: s/age/page/
Succeeded: s/me the same as/the same basic information as/
Succeeded: s/ojection/objection/
WARNING: Bad s/// command: s/<figure><img><figcaption> Example for a basic table </figcaption></figure> ???/
Succeeded: s/Hi sorry I have to go.//
Succeeded: s/bye//
Succeeded: s/Take care.//
Found Scribe: Sharron
Inferring ScribeNick: Sharron
Default Present: Shadi, +1.615.417.aaaa, EricE, Shawn, +1.512.731.aabb, kevin, +1.202.276.aacc, AnnaBelle, +1.562.256.aadd, Wayne, Jan, Sharron, PaulSchantz, Wilco, metzessive
Present: Shadi +1.615.417.aaaa EricE Shawn +1.512.731.aabb kevin +1.202.276.aacc AnnaBelle +1.562.256.aadd Wayne Jan Sharron PaulSchantz Wilco metzessive
Found Date: 12 Sep 2014
Guessing minutes URL: http://www.w3.org/2014/09/12-eo-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]