LVTF Meeting Minutes - 9/14/17

Minutes pasted below, and at link:
http://www.w3.org/2017/09/14-lvtf-minutes.html


Low Vision Accessibility Task Force Teleconference
14 Sep 2017


See also: IRC log


Attendees
Present
      Jim, Glenda, Shawn, steverep, Erich, Laura
Regrets
      John
Chair
      Jim
Scribe
      erich
Contents
      Topics
            house keeping
            Content on hover or focus
            3:1 vs 4.5:1 contrast ratio graphics and UI
            Adapting text
            adapting text & PDF
      Summary of Action Items
      Summary of Resolutions



<scribe> Scribe: erich


house keeping


JA: need a schedule for when understanding docs will be complete


GS: if do not need written techniques, mine is done


LC: mine is pretty well done too (adapting text)


<allanj> jim put UI contrast and Adapting Text Understanding doc on Agenda
for next week


<allanj> jim put a note out to group for getting on agenda.


Content on hover or focus


<allanj> https://github.com/w3c/wcag21/issues/350#issuecomment-327589159



SR: would like to discuss some of the last-minute objections raised


<steverep> https://github.com/w3c/wcag21/issues/75#issuecomment-321443487



SR: Issue about repositioning may be easier to resolve than issue with
trigger
... repositioning was added based on an objection


<allanj> current SC


<allanj> When a user interface component which receives keyboard focus or
pointer hover causes content to become visible, the following are true:


<allanj> Visible Trigger


<allanj> Either the additional content does not obscure any essential
content within the triggering user interface component, or the additional
content can be closed or repositioned by the user;


<allanj> Hover


<allanj> If the additional content is triggered via pointer hover, then
that content remains visible when the pointer is moved over it;


<allanj> Focus


<allanj> The additional content remains visible while the triggering user
interface component has keyboard focus, unless the user dismisses the
additional content.


<allanj> Exception: The visual presentation of the content is controlled by
the user agent and is not modified by the author.


GS; feel kind of neutral. the way I read it, if the trigger is obscured,
but you can reposition, it's ok but not ideal


SR: I propose language that, if you can reposition off of the trigger
itself, then it makes sense


GS: I would agree with that


<allanj> repositioning is big loophole


SR: moving windows around is generally not something low-vision folks do


<allanj> would there even be a repositioning technique


<allanj> repositioning seems to be under user control


SR: you don't necessarily need a techniqure for it, but regardless, if you
think about the sharepoint example you would just need different on-hover
regions for the same object


JA: what should we support you on Steve?


SR: I am going to file an issue, which hopefully will be discussed and we
can take care of it
... would like to get it resolved through language that everyone is okay
with
... I am proposing language that explicitly says it can be repositioned off
of the trigger
... support could come in the form of comments on the issue I feel, or
specific resolutions we have here about what is or isn't okay in terms of
low-vision


JA: Steve could you bring this back next week, to allow time for review,
discussion and identification of proper resolutions?


SR: Sure, sounds good.


<allanj> jim put Hover on agenda for next week. discuss on list
https://github.com/w3c/wcag21/issues/75#issuecomment-321443487



3:1 vs 4.5:1 contrast ratio graphics and UI


GS: we have logical reasons for saying 4.5:1 for things that are thin, but
the complexity of testing for that may not be worth it
... if we just get 3:1, it is reasonable to ask people to test for
... majority of research leading to 4.5:1 was based on reading


<allanj> https://github.com/w3c/wcag21/issues/345



GS: color contrast need are not specific to reading, are more complex


JA: a frequent issue i've had is with form controls, even at 3:1 at times


GS: could we get examples of insufficient contrast at 3:1?
... major concern is if we fall too quickly back to 3:1 we could shoot
ourselves in the foot
... no need to exclude any possibility of using 4.5:1 for certain thinner
elements


<allanj> research this


<allanj> more than a survey, need demographics. someone who understands low
vision


<steverep> Do we have a link to his talk?


<allanj> Erich: Howard Kaplan, low vision reading, invite to a call


<Glenda> Research I found interesting:


<Glenda> http://www.vectorvision.com/contrast-sensitivity-background/



<Glenda> http://www.vectorvision.com/csv1000-norms/



<allanj> contrast vs acuity limit


<Glenda> Here is the talk that Jared did on Rethinking Color and Contrast
https://www.youtube.com/watch?v=HtUlonNewGk&index=1&list=FLio9ud-Ay0NKxGlcUZLhQVQ



<allanj> with poor contrast single pixel line width fonts vanish


SR: many people also have glare sensitive, photophobic considerations


<allanj> total luminance is an issue for reading


<Glenda> The importance of Contrast Sensitivity
https://www.youtube.com/watch?v=HtUlonNewGk&index=1&list=FLio9ud-Ay0NKxGlcUZLhQVQ



<Glenda> “Contrast Sensitivity Function (CSF)


<Glenda> Detailed contrast sensitivity measurements that include both size
(spatial frequency) and contrast are used to plot a person's contrast
sensitivity function (CSF).


<Glenda> Sine-wave grating targets with thicker bars represent low spatial
frequencies; targets with thinner bars represent higher spatial
frequencies. In this regard, determining a person's CSF is much like
evaluating the sensitivity of his or her hearing, which involves using
tones of low and high pitch as well as variations in volume.


<Glenda> Your contrast sensitivity function essentially is a plotting of
the curve that defines the lowest contrast level that you can detect for
each spatial frequency tested.


<Glenda> Generally, objects with high spatial frequencies (sine-wave
gratings with very thin bars) must have significantly higher contrast than
objects with lower spatial frequencies (gratings with medium-width bars) to
be detected by the human visual system.”


<shawn> ACTION: Shawn consider in user needs/req issue with overall
luminance (e.g., light background) SR's point in 14 Sept 2017 [recorded in
http://www.w3.org/2017/09/14-lvtf-minutes.html#action01]


<trackbot> Created ACTION-102 - Consider in user needs/req issue with
overall luminance (e.g., light background) sr's point in 14 sept 2017 [on
Shawn Henry - due 2017-09-21].


SH: Steve has a good point, we need to have reflected in our user needs
document


<Glenda> http://www.precision-vision.com/?s=contrast%20patti Look at this


<allanj> this is low contrast vs acuity limit test not with letters


<Glenda> http://www.vectorvision.com/csv1000-contrast-sensitivity/



<Glenda> Standardized Contrast Sensitivity Tests


Adapting text


<allanj> "open" captions is editorial


<allanj> discussing
https://lists.w3.org/Archives/Public/w3c-wai-gl/2017JulSep/1001.html



<laura> 5. text color to any color that the user agent allows


<laura> 6. background color to any color that the user agent allows


<laura> 7. font-family to any font family that the user agent allows and
has immediately available


<laura> NOTE: If the user is changes the text and background colors - then
the author is not responsible for meeting any contrast SC other than for
the colors specified as default by the author.


<allanj> lots of push -back on Fonts


<allanj> color/contrast alastair comment
https://github.com/w3c/wcag21/issues/348#issuecomment-329119866



<allanj> The one that comes up (occasionally) is background images. E.g.
you have a solid (or gradient) blue background with white text. The user
changes the text colour to blue, and can't read it.


SR: should focus on few things that can go wrong with color


<allanj> transparency


<allanj> it seems the group thinks these are a non-issue


<allanj> jim will revisit greg 3 bullets next week for resolution.


adapting text & PDF


<laura> https://github.com/w3c/wcag21/issues/349


<steverep> What comment are we dicsussing?


<allanj> The one that comes up (occasionally) is background images. E.g.
you have a solid (or gradient) blue background with white text. The user
changes the text colour to blue, and can't read it.


<laura> https://github.com/w3c/wcag21/issues/78#issuecomment-323662527



<allanj> PDF fails the sc. but works in browsers


<allanj> if you provide an alternative conforming version you are done


<allanj> -- shawn


<laura> Andrew’s email:
https://lists.w3.org/Archives/Public/w3c-wai-gl/2017JulSep/1009.html



<allanj> seems that PDF can't be reliably tested. would hate to restrict
the SC, so when PDF can meets the SC then it can be tested


<allanj> SR: the exception seems to cover PDF


<allanj> user agent goes back to technology support.


<allanj> because a browser doesn't read alt text to the user, the page does
not fail WCAG


<allanj> because pdf reader cannot display arabic font, the page doesn't
fail the SC


<shawn> if new technologies come up that don't support alt text, that
doens't mean it doesn't meet WCAG


<allanj> concept - accessibility support, conformance in general. supported
by user agents, widely available.


<allanj> a new user agent that is not accessibility supported is not a
worry.


<allanj> perhaps in the understanding document we say PDF user agent does
not allow these changes and so is untestable


<laura> bye


Summary of Action Items
[NEW] ACTION: Shawn consider in user needs/req issue with overall luminance
(e.g., light background) SR's point in 14 Sept 2017 [recorded in
http://www.w3.org/2017/09/14-lvtf-minutes.html#action01]

Summary of Resolutions
[End of minutes]
                                                                                        
                    Erich Manser                                                        
                    IBM                                                                 
                    Accessibility,                                                      
                    IBM Research                                                        
                    Littleton,                                                          
                    MA / tel:                                                           
                    978-696-1810                                                        
                    Search for                                                          
                    accessibility                                                       
                    answers                                                             
                                                                                        
                                                                                        
                                                                                        
                                                                                        
                                                                                        
                                                                                        
                                                                                        



You don't need eyesight to have vision.