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
Deno's permission sub-system is one of the features that make this project stand against other runtimes. The code to control permissions currently lives in the CLI.
I think it'd be very interesting to make them part of the core, or move them to a separated crate. That way, people working on plugins, or using Deno as an embedded runtime could control the permissions too without having to rely on the CLI.
The text was updated successfully, but these errors were encountered:
The "problem" is that cli is the only place where the operations exist that need to leverage the permissions...
What could be abstracted to the core is a permission "system" which allows registration, enablement, and assertion of permissions, but not the actual permissions themselves, because they only come into existence in the CLI. I know there is a plan to break the ops out into crates themselves, so that would seem to me to be a logical refactoring at that point.
With work on op crates in progress (#2180#6906) we had to face this issue. @kitsonk is right that permissions are cli only concept and we might end up exposing trait/interface in op crates to check permissions, the actual implementation will still live in cli/.
Deno's permission sub-system is one of the features that make this project stand against other runtimes. The code to control permissions currently lives in the CLI.
I think it'd be very interesting to make them part of the core, or move them to a separated crate. That way, people working on plugins, or using Deno as an embedded runtime could control the permissions too without having to rely on the CLI.
The text was updated successfully, but these errors were encountered: