Is the read out for minimum length supposed to be 0.00...01?

[ghost]

Here's what JAWS sounds like when it reads the "Slider Controls" h3:

https://drive.google.com/open?id=1-IBXvhnGrQo0n4jlHKwDEd5MomSDyFMl

For phetsims/qa#179 phetsims/qa#183.

[jessegreenberg]

That sounds incorrect, its reading the value of the "Length" slider, and it sounds like the formatting isn't right.

[jessegreenberg]

Oh, for some reason it is reading the "min" and "max" values for the slider? I have never heard that before. he min value in the PDOM is min="0.1", so I am confused about where it is getting "0.00000.....1.

[jessegreenberg]

Hearing the recording again, it is saying "min 0.100...001", so it is a precision issue maybe.

[jessegreenberg]

Likely a problem with AccessibleSlider so assigning to myself.

[jessegreenberg]

I don't see anywhere where the "min" attribute could pick up extra decimals in toFixed calls or something, inspection in the console shows that min value is 0.1.

I am not hearing this in JAWS with FF 62, but JAWS isn't reading min and max values for the slider when it receives focus. Ill see if I can figure out how to turn that on in user settings or something.

[jessegreenberg]

OK, I hear it now, only happens when reading through with the virtual cursor (not on focus). Here is the HTML for the slider:

<label tabindex="-1" id="label-247-310-321-388-387-386-346-349" for="247-310-321-388-387-386-346-349">L, Length</label>
<input style="width: 1px; height: 1px;" id="247-310-321-388-387-386-346-349" tabindex="0" role="slider" aria-orientation="vertical" step="0.1" min="0.1" max="20" aria-valuetext="10.00 centimeters" type="range">

Ill try to reproduce in a JSFiddle.

[jessegreenberg]

I was able to reproduce in this JSFiddle: https://jsfiddle.net/au7rhbg8/

Ill start removing attributes until the problem goes away. (if it goes away)

[jessegreenberg]

Seems to just happen for some "min" values for sliders. For example, this slider with min=0.3 reads the min value (only when using the virtual cursor) as "0.299999....". I think that we should submit a bug report to VFO, add to the list of AT quirks, and close this issue.

@terracoda @zepumph what do you think?

[jessegreenberg]

Here is a very basic slider where the bug is coming up, Ill probably use this one in the bug report.

[jessegreenberg]

I went ahead and reported FreedomScientific/standards-support#112. I documented in our internal tracking https://docs.google.com/document/d/1518zv6F0odExFsodShZxwTmWNQk3civuNwtQUR-rEBs/edit. OK to close this issue?

[ghost]

If you think we’re done, then it’s OK to close from QA’s standpoint. But perhaps @terracoda and @zepumph will have something to say, so I’ll leave it open for now.

[zepumph]

Sounds good to close. @jessegreenberg isn't there a place to document this sort of weirdness so qa has a list of know bugs to check before submitting issues to developers? I would recommend adding this there.

[jessegreenberg]

There is, sorry I mentioned in an edit to #177 (comment) that I added a note there.

[jessegreenberg]

Alright we have 👍 s from everyone so I think we can close!

[KatieWoe]

Does happen in phetsims/qa#476, but not always consistently.

[KatieWoe]

https://drive.google.com/file/d/1hexjqYRExi6u7Doq8CGbOSHKlDqNFM3y/view?usp=sharing
Seen with values in phetsims/qa#582 when tick marks are on.

[zepumph]

Sounds right, I don't see any progress on the Freedom Scientific issue.