WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Text concatenation fault with Chrome and the "clip" technique for hiding text

for

From: Murphy, Sean
Date: Mar 26, 2020 8:53PM

I think I have, but wrote it up as a Jaws issue. A real good finding.

-----Original Message-----
From: WebAIM-Forum < <EMAIL REMOVED> > On Behalf Of Steve Green
Sent: Friday, 27 March 2020 1:51 PM
To: WebAIM Discussion List < <EMAIL REMOVED> >
Subject: Re: [WebAIM] Text concatenation fault with Chrome and the "clip" technique for hiding text

[External Email] This email was sent from outside the organisation – be cautious, particularly with links and attachments.

I agree that it's a Chrome issue. Our software library contains all the old versions of Chrome, so I will see if I can find out when this issue was introduced.

Given that this is such a common way of hiding text and that Chrome has pretty much become the default browser for testing, I am very surprised that neither we nor anyone else appear to have encountered this before.

Steve


-----Original Message-----
From: WebAIM-Forum < <EMAIL REMOVED> > On Behalf Of Murphy, Sean
Sent: 27 March 2020 02:44
To: WebAIM Discussion List < <EMAIL REMOVED> >
Subject: Re: [WebAIM] Text concatenation fault with Chrome and the "clip" technique for hiding text

Steve,

This looks like a Chrome issue, not a screen reader issue. As I tested in NVDA and we saw the issue before you made the changes. If I get time today I will check again.

Sean

-----Original Message-----
From: WebAIM-Forum < <EMAIL REMOVED> > On Behalf Of Steve Green
Sent: Friday, 27 March 2020 1:42 PM
To: WebAIM Discussion List < <EMAIL REMOVED> >
Subject: Re: [WebAIM] Text concatenation fault with Chrome and the "clip" technique for hiding text

[External Email] This email was sent from outside the organisation – be cautious, particularly with links and attachments.

Thanks for checking that out. Changepassword is one of the concatenated pairs of words. Ham and sandwich are also concatenated, but it is less obvious.

I have updated the test page not only to make the fault more obvious, but with examples showing the surprising finding that the concatenation occurs when the hidden text is in a <p> element, but it does not occur when identical content occurs in <li> elements.

Steve


-----Original Message-----
From: WebAIM-Forum < <EMAIL REMOVED> > On Behalf Of Murphy, Sean
Sent: 27 March 2020 02:13
To: WebAIM Discussion List < <EMAIL REMOVED> >
Subject: Re: [WebAIM] Text concatenation fault with Chrome and the "clip" technique for hiding text

Steve,


Just tested using NVDA and Chrome. Below is what was spoken:

Changepassword
Make me a hamsandwich

I am assuming " Changepassword" is the concatenated word.

Sean
-----Original Message-----
From: WebAIM-Forum < <EMAIL REMOVED> > On Behalf Of Steve Green
Sent: Friday, 27 March 2020 1:03 PM
To: WebAIM Discussion List < <EMAIL REMOVED> >
Subject: [WebAIM] Text concatenation fault with Chrome and the "clip" technique for hiding text

[External Email] This email was sent from outside the organisation – be cautious, particularly with links and attachments.

I have just encountered a weird issue with hidden text, which I have never seen before. These days we recommend using the "clip" technique to hide text rather than hiding it off-screen, because it works with both LTR and RTL languages. However, with JAWS 2019 and 2020 and Chrome, I am getting concatenation of the hidden and visible words, which is causing mispronunciation.

This does not occur with Firefox, Internet Explorer or the old version of Edge, but it does occur with the new version of Edge based on Chromium.

I have created a test page at http://tpl1.com/hidden_text.htm if anyone wants to take a look.

You can hear the mispronunciation easily enough, but you can also check by reading word by word or letter by letter. There appears to be no space character between the concatenated words, even though there are spaces in the source code.

The "solution" is to remove the "position:absolute" style rule, but the layout then breaks, which means we can't do that.

Has anyone else encountered this, and does anyone have a solution? Has this always happened or has something changed in Chrome? I may be wrong, but I don't think this is a JAWS bug because it only happens in Chromium-based browsers.

I have seen several variants of the "clip" technique and I tested my test page with a couple of them. In both cases the only thing that made a difference was the "position:absolute" style rule, so I expect that will be the case with all the variants.

Regards,
Steve Green
Managing Director
Test Partners Ltd
020 3002 4176 (direct)
0800 612 2780 (switchboard)
07957 246 276 (mobile)
020 7692 5517 (fax)
Skype: testpartners
<EMAIL REMOVED>
www.testpartners.co.uk

Connect to me on LinkedIn - http://uk.linkedin.com/in/stevegreen2