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
When rolling out a new verification key, we print the new key to the console but the user is responsible for adding it to its code or environment.
We could simplify this process by storing the public key in the configuration file. The dispatch run command could then read it from the file and automatically set it as an environment variable.
Using a verification key in dispatch run serves no purpose, since the CLI pulls requests from an authenticated API endpoint.
We could fetch or generate a verification key in the CLI, and set DISPATCH_VERIFICATION_KEY in dispatch run, however this only adds footguns for the user. If they roll-out a new verification key through the UI, or using the CLI on a different machine, the cached verification key in the config will no longer work.
Instead of caching a verification key, and using it in dispatch run, I'll instead hide the warning about a lack of verification key in the SDK: dispatchrun/dispatch-py#152.
When rolling out a new verification key, we print the new key to the console but the user is responsible for adding it to its code or environment.
We could simplify this process by storing the public key in the configuration file. The
dispatch run
command could then read it from the file and automatically set it as an environment variable.Relates to #6
The text was updated successfully, but these errors were encountered: