Notes about algorithm for input type="image"

[eps1lon]

1. 2. Otherwise use alt attribute. 3. Otherwise use title attribute.

   Is the "otherwise" in point 3 similar to 5.1.1 where we have

   If the accessible name is still empty, then: use the control's title attribute.

   ?
   Right now I can see no consistent language regarding these steps. It seems to me that every step only applies if the previous steps did not produce a name. But sometimes we have this explicitly mentioned and sometimes not.

2. the accessible name is a localized string of the word "Submit Query".

   This is followed by "If none of the above yield a usable text string there is no accessible name.". So this localization could produce an empty string? Wouldn't it be better to produce an accessible name in a foreign language rather than no name at all?

3. By spec this element cannot be labelled via <label>.
   However, a web-platform-test currently test this (https://github.com/web-platform-tests/wpt/blob/677c57c0e8816b0892cc3ae1c2772189b1bdcf65/accname/name_test_case_616-manual.html#L64-L65). Chrome 84 also allows labelling <input type="image" />. Should this test be removed or the spec updated to match implementations?

[scottaohara]

closing this as there are duplicate issues on these topics.

#423 was created to address the missing step for label elements

[eps1lon]

closing this as there are duplicate issues on these topics.

@scottaohara Could you link that? The linked PR does not address all points made in this issue.

[scottaohara]

#414

[scottaohara]

have pulled over your OP into the newer issue