Indicate via @autocomplete whether a password field is for new/current values

[mnoorenberghe]

This helps password managers (once #161 is fixed) along with accessibility tools.

[netlify]

Deploy Success!

Built with commit e6d7482

https://deploy-preview-275--identity.netlify.app

[erezrokah]

Thanks @mnoorenberghe,
Can you share some context/links on how password managers/accessibility tools use the string value of autocomplete?
A before/after example would be great (e.g. behaviour before and after the autocomplete change).

Seems like the available values for the field are on|off:
https://www.w3schools.com/tags/att_input_autocomplete.asp

Also, this seems relevant for other fields as well.
Instead of changing the type of password from a boolean to a string should we add new field called autocomplete?

[mnoorenberghe]

Can you share some context/links on how password managers/accessibility tools use the string value of autocomplete? A before/after example would be great (e.g. behaviour before and after the autocomplete change).

new-password can prevent autofilling of existing saved passwords and potentially provide password generation.
current-password will signal that password generation shouldn't be offered.

https://www.w3.org/WAI/WCAG21/Understanding/identify-input-purpose explains the benefits for accessibility.

Seems like the available values for the field are on|off:
https://www.w3schools.com/tags/att_input_autocomplete.asp

W3Schools isn't a trusted source and shouldn't be used. See https://www.w3fools.com/. Other values of autocomplete have been in the HTML spec for at least 4 years.

See https://developer.mozilla.org/en-US/docs/Web/HTML/Attributes/autocomplete instead.

Also, this seems relevant for other fields as well.
Instead of changing the type of password from a boolean to a string should we add new field called autocomplete?

The value needs to be set independently for each <input> so I don't think a single new field makes sense. I think changing from a boolean to take the type of password field makes sense since it forces the developer to think about which kind of password field this is.

[erezrokah]

W3Schools isn't a trusted source and shouldn't be used. See https://www.w3fools.com/. Other values of autocomplete have been in the HTML spec for at least 4 years.

Very cool thanks.

The value needs to be set independently for each <input> so I don't think a single new field makes sense. I think changing from a boolean to take the type of password field makes sense since it forces the developer to think about which kind of password field this is.

Cool, makes sense. Thanks for clarifying.

I'll do some testing then merge