Closed
Bug 1460378
Opened 7 years ago
Closed 4 years ago
HTML <meter> not spoken by screen readers
Categories
(Core :: Disability Access APIs, defect, P2)
Core
Disability Access APIs
Tracking
()
VERIFIED
FIXED
87 Branch
Tracking | Status | |
---|---|---|
firefox87 | --- | verified |
People
(Reporter: aaronlev, Assigned: morgan)
References
(Blocks 3 open bugs)
Details
Attachments
(5 files)
The HTML <meter> element represents either a scalar value within a known range or a fractional value.
The HTML meter element does not currently speak at all in Firefox. (In Chrome, it gets spoken as a progress meter).
CORE-AAM maps a <meter> to IA2_ROLE_LEVEL_METER. Unfortunately, NVDA + JAWS don't currently support IA2_ROLE_LEVEL_METER. I've sent them each emails to ask them to support it.
Example attached.
Comment 1•7 years ago
|
||
Taking a guess at P3 - please retriage as necessary.
Updated•7 years ago
|
Priority: -- → P3
Comment 2•7 years ago
|
||
it seems there's no role yet on os x https://bugs.webkit.org/show_bug.cgi?id=164051
Assignee: nobody → surkov.alexander
Attachment #8996798 -
Flags: review?(jteh)
Comment 3•7 years ago
|
||
Comment on attachment 8996798 [details] [diff] [review]
meter role - part1
Review of attachment 8996798 [details] [diff] [review]:
-----------------------------------------------------------------
This doesn't expose a value or the value interfaces, which it needs to do. It's not useful without this. (I realise the HTML AAM doesn't specify this. It should, and it does for ATK.) I guess the text Value should probably just be the @value attribute, rather than 70% or similar like it is for progress. FWIW, this is what Chrome does. (The spec doesn't cover this either, even for ATK.)
With this patch, there is also a div (section) accessible beneath the meter. We probably want to prune that.
Attachment #8996798 -
Flags: review?(jteh)
Comment 4•7 years ago
|
||
good point, I will update the patch and file the fix to HTML AAM.
Comment 5•7 years ago
|
||
Comment on attachment 8996798 [details] [diff] [review]
meter role - part1
Review of attachment 8996798 [details] [diff] [review]:
-----------------------------------------------------------------
let's keep the things separately, I'll put value interface stuff in follow-ups
Attachment #8996798 -
Flags: review?(jteh)
Updated•7 years ago
|
Attachment #8996798 -
Attachment description: meter → meter role - part1
Comment 6•7 years ago
|
||
Attachment #8997091 -
Flags: review?(jteh)
Comment 7•7 years ago
|
||
Thanks. However, this exposes a percentage as the value. I'm not sure if that's really what we want to do for meter, as I noted in comment 3. Aside from the way the control is described (its role), the other reason for creating this new role was because of the value:
> The element has an incorrect default ARIA role of progressbar which causes some assistive technology (I've tested in VoiceOver on OSX 10.11) to announce the element as a "progress indicator" and the value of the element as a percentage rather than the defined value (for example "80%" instead of "4 out of 5").
( https://github.com/w3c/html-aam/issues/2 )
Now, it's probably not reasonable for us to expose something like "4 out of 5"; that would be a localised string, it doesn't handle the case where the minimum isn't 0, etc. However, I tend to think we should just expose the raw value as Chrome does (350 in the attached example), rather than a percentage. The spec doesn't provide any guidance, so that's no help. :(
Comment 8•7 years ago
|
||
In some ways, I think it makes sense to use the fallback content (if specified) like aria-valuetext. There's some discussion on that here: https://github.com/w3c/html-aam/issues/27
Comment 9•7 years ago
|
||
Comment on attachment 8996798 [details] [diff] [review]
meter role - part1
Cancelling review because I don't think Alex is working on this any more and we need to re-think this based on comment 7 and bug 1499947.
Attachment #8996798 -
Flags: review?(jteh)
Comment 10•7 years ago
|
||
Comment on attachment 8997091 [details] [diff] [review]
value interface
Cancelling review because I don't think Alex is working on this any more and we need to re-think this based on comment 7 and bug 1499947. Note that bug 1499947 completely breaks this patch, since it gets rid of the base ProgressMeterAccessible template.
Attachment #8997091 -
Flags: review?(jteh)
Comment 11•7 years ago
|
||
Marco, I'd appreciate your thoughts on comment 7 and comment 8. Also, would you care to take this on?
Flags: needinfo?(mzehe)
Updated•6 years ago
|
Comment 14•5 years ago
|
||
Would be great to get support for this and also the aria role=meter that was recently added: https://w3c.github.io/aria/#meter
Assignee | ||
Updated•5 years ago
|
Assignee: nobody → mreschenberg
Assignee | ||
Comment 15•4 years ago
|
||
:Jamie, do we need to create a new type flag for this? based on the discussion in the github issue here, and the resolution that meter should be distinct from progress, it feels wrong to map mType = eProgressType, but I'm not sure we have something that fits better at the moment.
Flags: needinfo?(jteh)
Assignee | ||
Comment 16•4 years ago
|
||
Assignee | ||
Comment 17•4 years ago
|
||
Depends on D104759
Updated•4 years ago
|
Attachment #9202443 -
Attachment description: Bug 1460378: Map role description on mac for meter element r?eeejay → Bug 1460378: Verify role, role description mapping for meter element on mac r?eeejay
Comment 19•4 years ago
|
||
Pushed by mreschenberg@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/d03c36cdbc21
Introduce meter role and HTMLMeterAccessible class. r=Jamie
https://hg.mozilla.org/integration/autoland/rev/9c26b7582208
Verify role, role description mapping for meter element on mac r=eeejay
Comment 20•4 years ago
|
||
bugherder |
https://hg.mozilla.org/mozilla-central/rev/d03c36cdbc21
https://hg.mozilla.org/mozilla-central/rev/9c26b7582208
Status: NEW → RESOLVED
Closed: 4 years ago
status-firefox87:
--- → fixed
Resolution: --- → FIXED
Target Milestone: --- → 87 Branch
Updated•4 years ago
|
Flags: qe-verify+
Comment 21•4 years ago
|
||
Confirmed with ESR68 that the element is skipped from reading.
Verified with 87.0, it being read as "selected" element.
You need to log in
before you can comment on or make changes to this bug.
Description
•