Home - News ( United States | United Kingdom | Italy | Germany ) - Football scores

Showing content from https://extensionworkshop.com/documentation/develop/request-the-right-permissions/ below:

Request the right permissions | Firefox Extension Workshop

Or how to avoid permissions discouraging users from installing your extensions.

To inform users of the potential impact of installing an extension, Firefox displays a message showing what permissions an extension is requesting when it's installed. The prompt looks something like this:

Also, if an extension update requires additional permissions the user is prompted to approve the permissions before the updated version is installed:

If the user chooses not to approve the permissions and cancels the update, the previous version remains installed and available for use.

These messages could discourage some users from installing your extension, as they might suggest it's doing something “scary”. We provide users with an explanation of these permissions messages and guidance on how to judge if they're appropriate. However, there are several things you can do to reduce the likelihood of users abandoning the installation of your extension because of these messages; these include:

• Make sure you aren't requesting unnecessary permissions.
• Request permissions at runtime, which enables you to ask for the permissions in context and offer a fallback if the user doesn't grant them.
• Describe why your extension is requesting its permissions in its AMO description.

Tip: Permission messages aren't issued when you load an unpacked extension. For information on how to see the standard runtime permission flow, see Test permission requests.

Not all permissions result in advice being given to the user. The permissions that trigger the display of a message and the messages they trigger are:

The permissions that display messages and their messages may differ in other browsers. For information about viewing the permission message in Chrome, see Viewing Warnings.

These permissions don't get alerted to users:

• API permissions:
  • alarms
  • contextMenus
  • contextualIdentities
  • cookies
  • declarativeNetRequestWithHostAccess
  • identity
  • idle
  • menus
  • storage
  • theme
  • webRequest
  • webRequestBlocking
• Other permissions:
  • unlimitedStorage
  • activeTab

This section looks at situations where you might be asking for more permissions than your extension needs and what you should do about them.

This may seem obvious, but if you create an extension using an earlier example as a template or remove a feature during development or testing, you may be asking for permissions your extension doesn't need. Addressing this involves manually checking your code against the permissions (permissions and optional_permissions) that you're requesting in the extension's manifest.json.

Take an extension you're developing to help color-blind users. At the user's request, your extension looks for and updates CSS on a web page to replace colors the user may have difficulty distinguishing with safe colors. Your extension must access and update CSS on every page your user visits. You could do this by requesting the "tabs" permission and host permission for "<all_urls>".

"permissions": [
  "<all_urls>",
  "tabs"
]

Requesting these permissions results in the user getting this advice:

The alternative is to request "activeTab". This permission provides your extension with the same capabilities but only for the active tab and only when run from the extension's UI (that is, from a toolbar button, navigation bar button, context menu, or shortcut key).

Importantly, "activeTab" doesn't display a permissions message when installing the extension.

As noted in the previous example, requesting host permission "<all_urls>" results in the permissions request message "Access your data for all websites". If your extension is designed to work with one or a few websites or domains, narrow the request. Then, when installing your extension, users see details for the first four websites or domains your extension requests access to.

If your extension requests access to more than four websites or domains, the message lists the first three and indicates the number of other requests.

Only ask for "unlimitedStorage" permission if you expect your extension's local data storage to exceed 5MB; if it's not going to exceed that amount, don't ask for it.

Firefox doesn't restrict local storage size, although it does ask users to approve this permission request if you make it. Firefox may add a restriction in the future. If that happens, the limit is unlikely to be less than Chrome's current 5 MB restriction.

Users may not understand the context of permissions requested during installation. To provide the user with context, request permissions when they are needed using the Permissions API.

Permissions that can be requested a runtime are known as optional permissions. However, not all permissions can be requested at runtime. There are also a small number of optional-only permissions that can only be requested at runtime. An example of one such optional-only permission is the userScript permission in Firefox.

A typical scenario for using this approach is the "geoLocation" permission. Say you've written a note-taking extension that includes the ability to add a minimap of the note-takers location. Requesting location access during installation might leave the user unclear why the extension needs to access location, so they might not install it. However, if permission to use location is requested when the user first tries the feature to add a minimap, they'll get a clearer understanding of why the permission is needed and be more likely to grant it. Should they choose not to grant the permission, the extension can offer a graceful fallback—in this example, not adding the minimap—but the important outcome of this approach is that the user has installed and used your extension.

Making a runtime permission request is straightforward. Include any permissions you want to request under the manifest.json optional_permissions key. Then, pass the permissions you want granted to permissions.request, which prompts the user to grant the permissions. true is returned if the user grants the request, false if they don't.

You can't request all the permissions available to "permissions" using optional permissions. You can't request these API permissions:

• alarms
• background
• contentSettings
• contextualIdentities
• debugger
• downloads
• downloads.open
• find
• identity
• menus
• pageCapture
• privacy
• storage
• theme

There are a couple of things to note:

• You can only request permissions inside the handler for a user action, such as from a toolbar button (browser action), shortcut menu item, or similar.
• If you need to request an optional-only permission, it must be the only permission you request in a permissions.request call.
• If you request several permissions at once, they are either all granted or all declined; the user cannot choose to grant some and not others.

Users can also grant or revoke optional permissions anytime through an extension's option page. Your extension, therefore, needs to test to determine whether it has the necessary permissions or listen to the permissions.onAdded and permissions.onRemoved events.

For more information about optional permissions, see optional_permissions and the permissions example.

Permissions messages are most likely to prevent users from installing your extension because the user doesn't understand why permissions are being requested. While the user can get general advice on the impact of a permission, it may not be sufficient for them to understand why a permission is being requested in your extension.

To address this issue, provide information in your extension's AMO description explaining what permissions your extension requests and why.

A good example of this approach is Gesturefy, which offers users this advice:

Permissions explained:

• Access your data for all websites: This is a key permission, because the complete gesture functionality is injected in every webpage you visit (which means a part of the code is running in each tab). This is necessary, because with the new API there is no other way to track your mouse movement or draw anything on the screen. It’s also needed to perform page specific commands like scroll down or up.
• Read and modify browser settings: This is required to change the context menu behaviour for MacOS and Linux users to support the usage of the right mouse button.
• Display notifications: This is used to show a notification on Gesturefy updates or to display certain error messages.

RetroSearch is an open source project built by @garambo | Open a GitHub Issue

Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo

HTML: 3.2 | Encoding: UTF-8 | Version: 0.7.3