You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
of the user when presenting payment methods. It is RECOMMENDED that
the language of the user interface match the [=node/language=] of the
<adata-cite="HTML/dom.html#the-body-element-2">the body element</a>
element.
The body element's definition states "The body element of a document is the first of the html element's children that is either a body element or a frameset element, or null if there is no such element."
I'm not sure I support the use case. Other APIs, e.g., the notification API, have no such defaulting and have explicit support for language annotation. And especially for the frameset case it seems that it's highly likely it might not match the language of the application anyway.
Thus, the user interface's language should not be derived from "the body element". The spec could probably include explicit support for language annotation the way the notification API does it.
Agree, we should drop this requirement as I don't think anyone implements this. It should just be noted that, by default, payment sheet will use the browser's default language.
In future version of the spec, we should add {dir: "", lang: ""} members to either the PaymentDetailsInit or to individual PaymentItems. Not sure yet what the right level of granularity should be. I think we discussed something similar a few years ago... we should do some archeology.
From the payment request spec,
payment-request/index.html
Lines 1094 to 1097 in fe208db
The body element's definition states "The body element of a document is the first of the html element's children that is either a body element or a frameset element, or null if there is no such element."
Thus,
<body>
is a node can can have a language<frameset>
is a nodenull
is not a node and cannot have a language.In this PR, @annevk states
Thus, the user interface's language should not be derived from "the body element". The spec could probably include explicit support for language annotation the way the notification API does it.
cc @marcoscaceres
The text was updated successfully, but these errors were encountered: