Re: Problem with WAI Site

Hi All,
I am carrying a discussion on with Eric. I think one could argue either way
about conformance, but I'd really like to see if Eric can find out why.
Maybe we can have a good technique that also assists evaluation. If we have
a precise way to predict problems from code its better.

Best, Wayne

On Thu, Jun 28, 2018 at 4:14 PM Alastair Campbell <acampbell@nomensa.com>
wrote:

> Hi Wayne,
>
>
>
> I think it is reasonably straightforward, the WAI site works well with
> zoom / reflow at 320px, but when text-spacing is added it over-flows the
> window-width due to the menu.
>
>
>
> Laura pointed to issue 850, where we got to the point of no-one
> disagreeing with my proposed response. I just updated it now:
>
> https://github.com/w3c/wcag21/issues/850#issuecomment-401199622
>
>
>
> So we can bring that to a group resolution soon.
>
>
>
> That basically said that SC are not cumulative, however, where there are
> variations of a page, each SC should be tested in each variation. (As-per
> the new conformance note.)
>
>
>
> That seems like a good balance, as otherwise in a responsive design which
> flexes (in a good way), you have a huge burden of testing at every possible
> width of page. Using breakpoints as ‘variations’ seems reasonable.
>
>
>
> The interesting aspect here is at what point the ‘variation’ changes. From
> the CSS It looks like there are variations at 0-35em,  47.5em, and 60em.
> (560px, 760px, 906px).
>
>
>
> So (my interpretation of) the conformance requirement is that if
> text-spacing works at under 560px, that passes, as that is a variation of
> the page.
>
>
>
> However, in the case of the WAI site I’m sure Eric (and the team) would
> want to address this. Most of the content is fine, it is just that the menu
> in the header gets pushed out which causes layout issues further down as
> well.
>
>
>
> Would you like to raise a bug?
>
> https://github.com/w3c/wai-website/issues
>
>
>
> Cheers,
>
>
>
> -Alastair
>
>
>
>
>
> *From:* Wayne Dick <wayneedick@gmail.com>
> *Sent:* 28 June 2018 20:14
> *To:* public-low-vision-a11y-tf <public-low-vision-a11y-tf@w3.org>
> *Subject:* Problem with WAI Site
>
>
>
> Dear Group,
>
> I would like to keep this within LVTF until we resolve the problem. I
> think there is a very subtle technique here.
>
>
>
> The WAI site fails at 320 CSS px. It over flows the pate. Her is how it
> works. The page works with no adjustment to letter-spacing, but with 0.12em
> letter spacing at 320 CSS px it fails.
>
>
>
> The experiment was conducted as follows. I used a user stylesheet on
> Chrome, Firefox and Safari.  * {letter-spacing: 0.12em !important}. I used
> Stylish for Chrome and Firefox and the preferences > advanced > stylesheet
> option for Safari. I tried Edge as well but it failed for other reasons, so
> I've left it out.
>
>
>
> Here is how I produced 320 CSS px (341 CSS px for Safari).
>
>
>
> Chrome allows enlargement up to 500%. So on my iMac with 1600x900 this
> gives 320 CSS px. The site fails.
>
>
>
> Safari only gives 300% enlargement. When I lower the resolution 1024x768 I
> get 341 CSS px. The site failed. (I used my Macbook Air for this last test.
> the iMac is hard to set resolution smaller than 1600x900.)
>
>
>
> Firefox gives up to 300% enlargement standard, but you can change this in
> the About:Config file. I set the last 3 enlargement values to 300%, 400%,
> 500%. With these settings I applied 500% to my iMac at 1600x900 giving 320
> CSS px. The page failed.
>
>
>
> All three browsers failed at anything smaller than 341 CSS px. The 320
> cases were worse.
>
>
>
> The only deficient browser was Edge. It just failed to split two column
> regions 300% with 1280 width.
>
>
>
> The main finding is this. When the WAI site is given 341CSS px or less, it
> fails when letter-spacing is 0.12em. The same browsers perform well on
> other sites.
>
>
>
> Alastair stated worry on this issue during our discussions. So, I think
> this problem is significant. Eric is a good programmer. That means the
> issue must be subtle. When we do, it will be a technique.
>
>
>
> I want to emphasize that Eric has worked hard on this site; he is an
> excellent developer, and he couldn't anticipate this issue. I don't want
> this to go on the general list until we have a reason why the failure
> occurred. It would be unfair to Eric.
>
>
>
> Best to All, Wayne
>
>
>
>
>
>
>