minutes for 6 Feb, 2015

http://www.w3.org/2015/02/06-svg-a11y-minutes.html


- DRAFT -
W3C SVG Accessibility Task Force
06 Feb 2015


See also: IRC log


Attendees
Present
Regrets
Chair
      Fred
Scribe
      Rich
Contents
      Topics
         1. What do we plan to do for blind users?
      Summary of Action Items



<richardschwerdtfeger> scribenick: Rich


What do we plan to do for blind users?


<richardschwerdtfeger> fred: I would like to scope the taskforce effort. …
the 6 disabilities under wcag2


<shepazu> AmeliaBR's email:
https://lists.w3.org/Archives/Public/public-svg-a11y/2015Jan/0065.html



<richardschwerdtfeger> doug: I have an opinion on everything


<richardschwerdtfeger> doug: from the gist of I thought there was an
interesting discussion about the scope of the group


<richardschwerdtfeger> doug: normative vs. informative


<richardschwerdtfeger> fred: I think there has been guidance put out there


<richardschwerdtfeger> doug: for each of these items it would be good for
us to say that if there something good for us to be doing best practices
vs. what ht platform implementations will support


<AmeliaBR> +1 to idea that this discussion should focus on dividing up what
can be done with author guidance, what needs browser support, what needs
browser+AT+authors working together, etc.


<richardschwerdtfeger> doug: so, authors, the best practice is for the
author to use titles to apply meta data for blind users


<richardschwerdtfeger> doug: there is no normative requirement for exposing
the requirement via the user agent


<richardschwerdtfeger> doug: rich has discussed how you discuss the
accessible name


<richardschwerdtfeger> rich: doug, do you think we can normatively expect
user agents to expose title as a tooltip


<richardschwerdtfeger> charles: I don’t think so. It is a UI choice for
user agents to expose things in a certain way


<richardschwerdtfeger> charles: so there are UI choices that we make. The
focus management is something the user agent can do


<richardschwerdtfeger> doug: there are exceptions - rendering a rectangle
is normative


<richardschwerdtfeger> doug: for example, on a mobile device you don’t have
hover


<richardschwerdtfeger> doug: on a mobile device the title could be exposed
at the bottom


<richardschwerdtfeger> doug: since user agents do render tooltips as a
title display properties can control the rendering


<richardschwerdtfeger> rich: I would not want to have title required as it
as side effects


<richardschwerdtfeger> doug: title does not allow child content so we might
want to have other content meta data that allows content


<richardschwerdtfeger> Amelia: Just following up this general conversation
would it be possible to require browser manufacturers to render title some
way?


<richardschwerdtfeger> doug: it could be a should


<richardschwerdtfeger> charles: there is no law of the universe on this


<richardschwerdtfeger> charles: if you express the requirement in a way
that makes sense there is a way to talk normative requirements into the
spec.


<shepazu> +1


<richardschwerdtfeger> charles: you can cause endless problems for very
little value, if ou do this wrong - which makes people very sceptical of
adding normative UI requirements and this is the right way for them to be


<richardschwerdtfeger> amelia: titles uses tooltips because titles are not
always the best text for a label


<richardschwerdtfeger> amelia: I agree we do have aria-label now as well.


<richardschwerdtfeger> doug: we recognize that some people want the title
to be used as tooltips


<richardschwerdtfeger> doug: there were a different set of authors back
then.


<richardschwerdtfeger> doug: there is a reason that people are warry of
specifying these things. browser vendors don’t like it when you tell them
how they must render things


<chaals> [+1]


<richardschwerdtfeger> charles: aria is not actually in my world a great
solution for a lot of things


<Zakim> chaals, you wanted to say ARIA-labels are currently not that good
either…


<richardschwerdtfeger> charles: what do we pan to do for blind users?


<richardschwerdtfeger> doug: most of what we have talked about thus far has
been to support blind users


<richardschwerdtfeger> charles: the first things we should do for users we
should identify problems face using SVG


<richardschwerdtfeger> charles: It's a bad idea to try and sort out SVG by
going one disability at a time - wee need to look at problems, and the
common threads we have. But for the specific question


<richardschwerdtfeger> charles: because we can’t see things visually we are
trying to make semantic structure available


<richardschwerdtfeger> charles: here is the solution for blind users and
others with 10% vision


<richardschwerdtfeger> charles: what we choose to do depends on how fast we
chose to do them


<richardschwerdtfeger> Amelia: I think this is a good point. We need to
think about all the use cases and all the commonality


<richardschwerdtfeger> Amelia: the blind users are the most obvious users
that need the most technological help but one thing to mention is that
there is another category of users - for tactile users


<chaals> [NB: There are significantly more colour-blind users than
zero-vision users…]


<richardschwerdtfeger> Amelia: same original probem


<chaals> [+1 to the statement that there are big gaps in WCAG 2]


<fesch> RS: role of aria -and WCAG WCAG does not address cognitive
disabilities


<fesch> RS: what aria does is work with screen readers and provide object
locations, and supports mobility impaired with SVG keyboard support


<fesch> RS: alternate views need to look at semantics, but we can't do that
if all low level graphic primitives


<fesch> RS: pan and zoom is a nice feature in user agents... don't know if
it affects us


<fesch> RS: aria does more than screen reading, when we have gaps, can
address with authoring practices


<richardschwerdtfeger> doug: you mentioned the line thickness thing. You
mentioned the non-scaling strokes


<richardschwerdtfeger> doug: If I zoomed in on a building diagam, as we
soom in 10 times the perspective stays the same


<richardschwerdtfeger> doug: I had a grid up on the screen and the
projector would not adjust to a 1 pixel widht


<richardschwerdtfeger> doug: allow the user to increase the default line
width


<richardschwerdtfeger> fred: you know what you are low vision and you wnat
lines to be this thick


<richardschwerdtfeger> doug: why can’t we have media queries but for
preferences.


<richardschwerdtfeger> doug: why can’t we have media query preferences


<richardschwerdtfeger> doug: why do we have to have one design for
everyone?


<richardschwerdtfeger> doug: why have one size fits everyone


<richardschwerdtfeger> doug: I think this is something for the CSS working
group to discuss


<richardschwerdtfeger> doug: people are concerned about exposing their
personal impairments


<richardschwerdtfeger> doug: for those people don’t set these preferences


<richardschwerdtfeger> doug: making sure the text does not goe outside a
particular boundary


<richardschwerdtfeger> doug: SVG2 will allow wrapped text


<fesch> rs: personalization in GPII is not new, for cognitive ti is not an
option - each person has different needs


<fesch> rs: you can see the work - schema.org


<fesch> rs: also working with media queries group


<fesch> rs: we had a lot of concerns - and getting info from schema.org and
we may be able to roll into media queries


<fesch> rs: WebAIM polls -


<richardschwerdtfeger>
https://dvcs.w3.org/hg/IndieUI/raw-file/default/src/indie-ui-context.html



<fesch> rs: WebAIM polls we want exposed from the user agent.


<fesch> rs: going through all user agent meta data


<fesch> rs: lots is in CSS4 with media queries need to add requirements for
other groups for our gap analysis


<fesch> rs: did a similar thing for aria - in gap analysis


<richardschwerdtfeger> doug: webaim ran an accessibility pole and they
asked people who use AT are you comfortable with letting people know that
you are blind?


<richardschwerdtfeger> doug: initially they said no.


<richardschwerdtfeger> doug: last year they had an enthusiastic yes


<richardschwerdtfeger> doug: I would like to allow a designer to have a hip
design with small fonts but with you aging users provide this alternative


<richardschwerdtfeger> doug: If you can automate things that would be good


<fesch> lol


<richardschwerdtfeger> doug: This is the best color scheme for you


<richardschwerdtfeger> leonie: I think we need to go back to being as
simple as possible


<Zakim> LJWatson, you wanted to say maybe we should start to document
problems, then look for solutions


<chaals> [+1 to Léonie - I think requiring people to identify has a long
way to go before it's a reasonable objective]


<fesch> rs: trying to expose features that don't make a declaration of
their impairments


<fesch> rs: schema.org can have settings that don't declare an impairment,
can do it different ways


<richardschwerdtfeger> charles: I am not yet on the same page


<richardschwerdtfeger> charles: the accessibility testing tool disappears
in the noise


<richardschwerdtfeger> charles: there is a lot of reasons where self
identification requirements all the way to the server is a non-starter


<richardschwerdtfeger> fred: what if it does not go back to the server?


<richardschwerdtfeger> fred: the idea of having a pre-processor that would
be fine by you


<richardschwerdtfeger> doug: the initial answer is that it can be in the
browser and they can take your computed style. We are going to collect the
information and sending it up


<richardschwerdtfeger> doug: the visited link, where you had a list of bank
sites where the visited links were colored a certain way you could use the
computed color to find out where they went


<richardschwerdtfeger> doug: so the way the browsers dealt with that, we
are not gong to tell you the computed styles.


<richardschwerdtfeger> doug: I think we are way off topic


<richardschwerdtfeger> Amelia: I wanted to say the same as doug. We can use
media queries in the browser but there ways around it.


<richardschwerdtfeger> Amelia: if you have stuff where the user and browser
do things to make things look better ...


<richardschwerdtfeger> doug: a lot of users, … I want the stroke width to
be 2 vs. 1


<richardschwerdtfeger> doug: many users won’t even now font size


<richardschwerdtfeger> fred: charles suggested identifiying problems


<richardschwerdtfeger> fred: do we do type and identify problems?


<richardschwerdtfeger> doug: address things on wikis. while I try to
address these things in presentations I don’t know all the detailed issues


<richardschwerdtfeger> doug: I don’t know how to really navigate around
with a puff sip device


<richardschwerdtfeger> doug: SVG is great for cognitive issues


<richardschwerdtfeger> doug: color blind issues need to be addressed by css


<richardschwerdtfeger> doug: we could work with color vision issues in the
CSS working group


<richardschwerdtfeger> doug: somebody said that the text is the default way
to interact with the web


<richardschwerdtfeger> doug: graphs can also be sonified


<chaals> [+1 to doug that there are approaches beyond text representation
for blind people to use the web]


<richardschwerdtfeger> doug: we should make it easy for sonification to be
done


<richardschwerdtfeger> charles: +10000 for using a wiki


<richardschwerdtfeger> charles: actually there are benefits to addressing
the reading level of users


<richardschwerdtfeger> charles: graphics do have an impact on cognitively
impaired users


<richardschwerdtfeger> charles: we are going to have gaps in our knowledge


<shepazu> https://www.w3.org/wiki/SVG_Accessibility



<richardschwerdtfeger> Amelia: we need to break down what are the main
problems and the solutions


<richardschwerdtfeger> Amelia: We need to identify the problems and how do
we make those solutions happen


<richardschwerdtfeger> fred: lets list the problems by accessibility type


<richardschwerdtfeger> fred: next week we can go from there


<shepazu> https://www.w3.org/wiki/SVG_Accessibility



<richardschwerdtfeger> doug: start on the wiki page and branch out from
there

                                                              
                                                              
                    Regards,                     Fred         
                                                              
                   Fred Esch                                  
       Accessibility, Watson Innovations                      
    AARB Complex Visualization Working Group                  
                     Chair                                    
            W3C SVG A11y Task Force                           
                IBM Watson Group