Define a PermissionStorage type for each permission to store data with. #73
Conversation
| @@ -514,43 +523,66 @@ | |||
| </dt> | |||
| </dl> | |||
| <p> | |||
| A <code>PermissionStatus</code> instance has an associated | |||
| <var>permission</var> that is a <a>PermissionDescriptor</a>. | |||
| <code>PermissionStatus</code> instances are created with the following | |||
There was a problem hiding this comment.
I'm curious why it's useful to define this as an internal slot as opposed to just saying that it has an associated PermissionDescriptor? It seems a bit more confusing to me.
There was a problem hiding this comment.
PermissionStatuses are going to have two internal slots once we get to #66, in case that makes it look a little better: https://rawgit.com/jyasskin/permissions/allow-choosers/index.html#idl-def-PermissionStatus.
I'm finding in Web Bluetooth that having a bunch of internal slots makes it easier to understand the purpose of each bit of associated data, and that foo@[[bar]] is simpler to type and link than "the bar associated with foo". The PermissionStatus slots are simple enough that I can eliminate them without needed a lot of extra wording, but I do somewhat prefer them.
@annevk and @domenic may also have opinions about which to use.
There was a problem hiding this comment.
I don't care strongly at the moment. Hopefully going forward IDL will provide formalism around this so you can define the layout of an object, including its internal slots, upfront.
This is separate from the PermissionStatus that gets returned to users and from the PermissionDescriptor that users query the permission with.
https://rawgit.com/jyasskin/permissions/storage-type/index.html#idl-def-PermissionStorage
This is separate from the PermissionStatus that gets returned to users and from
the PermissionDescriptor that users query the permission with.