- From: Oda, Terri <terri.oda@intel.com>
- Date: Fri, 31 Jul 2015 11:44:32 -0700
- To: "public-webappsec@w3.org" <public-webappsec@w3.org>
- Message-ID: <CACoC0R8xqH1dxvYiEDVz3vJ_BPqc4NVBAWKV0tPQtAQkngF-Vg@mail.gmail.com>
I mentioned this in a meeting some time back and promised I'd follow up on the list, but forgot until it came up again in a discussion about Web Assembly this week. The current permissions API is a read-only way for web applications to get permission information in a consistent way. Management of these permissions by the user is currently done through the browser. For example, geolocation permissions give the user a popup, they make a decision at that time, and the user can change the decision later through the browser's tools for permission management. This is great when you have a full browser, but I'm working with some teams who are hoping to have devices, mostly embedded/Internet of Things products, where they support web apps, but have a web runtime that doesn't already include any way to alter those permissions after they're set, or even ask the user about those permissions. e.g. headless IoT devices that want to be able to execute node.js apps, car IVI systems that only have a limited number of pre-approved applications for the platform, etc. We work with enabling for a lot of diverse potential products and it's starting to become an issue I'm seeing with some frequency. I was wondering if having a standardized way to set permissions as well as query them is a thing that would be useful to others. Think of it as a sister API to the current Permissions API. It's probably not relevant to browser implementers as they already have their own ways to do this, but I'd like to know if others are starting to see potential pain points with embedded and IoT devices. Note that of course this is an API that would need some access controls itself, as exposing permission management API to all web apps would be a quick way to make permissions useless. Terri
Received on Friday, 31 July 2015 18:45:04 UTC