Re: Responsive tables and accessibility

gregg

> On Jul 12, 2016, at 11:32 AM, Jonathan Avila <jon.avila@ssbbartgroup.com> wrote:
> 
> Ø  Try taking a budget sheet and doing this — and then trying to analyze it and check it — or draw conclusions from it.    
> Gregg, thank you for your input, I totally agree that are cases when scrolling may be needed as long as it can be scrolled on mobile/all environments.  There may also be other techniques that allow users to compare two columns, etc.
>  
> Ø  I think it is FINE to linearize twice for those that feel this is better for them.  But I’m not sure that we should Require that every website behave the way instead of using horizontal scrolling.   
> When you say it is fine to linearize for those that feel it is ok – I assume you mean it’s ok if the author choose that or are you saying the end user?    ‘

End user of course.    Authors should not be determining what a user needs. 

> My main question, is it is ok for the author to force a linear layout when the viewport width is narrow and continue to use table markup even though the row markup is linearized into a column.  

I think the authors role is to make it possible for viewers to view things as they need them.  Never to force a user to view it one way or another. 

> Semantically speaking is that ok under the current WCAG 2 A/AA. 

Author forcing?    no I do not believe it is.  I think all of WCAG is about giving users options. not the reverse. 

>   I understand from a customization standpoint it forces the user into a view that he/she might not want and I think that might be an issue for future WCAG.

Don’t understand what you are referring to here.   what is IT? 

>  
> Jonathan
>  
>  
> From: Gregg Vanderheiden [mailto:gregg@raisingthefloor.org] 
> Sent: Tuesday, July 12, 2016 11:10 AM
> To: Jonathan Avila
> Cc: GLWAI Guidelines WG org
> Subject: Re: Responsive tables and accessibility
>  
> Often the purpose of a table is to allow users to compare values both vertically and horizontally.
>  
> Linearizing does not allow this — as you must choose one or the other.  You can do two linearizing but you then can do one or the other but not both.
>  
> Try taking a budget sheet and doing this — and then trying to analyze it and check it — or draw conclusions from it.    
>  
> I think it is FINE to linearize twice for those that feel this is better for them.  But I’m not sure that we should Require that every website behave the way instead of using horizontal scrolling.   
>  
>  
> This is not speaking against responsive design - which I thoroughly endorse.    Just that there are some types of information that cannot be flowed without losing information/function.   We need to figure out how to characterize the exceptions in an objective and not list-like manner. 
>  
> gregg
>  
> On Jul 12, 2016, at 10:57 AM, Jonathan Avila <jon.avila@ssbbartgroup.com <mailto:jon.avila@ssbbartgroup.com>> wrote:
>  
> There are many different methods of making tables responsive.  One technique is to turn each row into a grouped column of data by keeping the HTML table structure but using display:block so the data becomes stacked.  The column headers are positioned off-screen and the headers for the columns are placed before the data by using CSS pseudo elements.  One example of this technique can be found in a codepen by Steve Faulkner.http://codepen.io/stevef/details/myzLdr <http://codepen.io/stevef/details/myzLdr>  I like Steve’s example and this seems to be a popular technique but I have run into people who disagree on the principle that the structure doesn’t match the appearance. 
>  
> Some say that representing a row of data in a column no longer allows the semantic structure to match the visual presentation and thus there is a violation between the presentation and semantics of the structure.  Does presenting tabular data into a single column raise concern over SC 1.3.1? 
>  
> Jonathan