Launchd worker fails with "no route to host" on Sequoia

[timpeeters]

We have been running Orchard workers, configured using launchd on Sonoma for a long time already. The implementation was based on the documentation here: https://tart.run/orchard/deploying-workers/#launchd

However, after upgrading one of our hosts to Sequoia, the workers that is started via launchd no longer connects to the controller.

{"level":"warn","ts":1732044922.827034,"msg":"failed to register worker: API client failed to make a request: Post \"https://<redacted>:6120/v1/workers\": dial tcp <redacted>:6120: connect: no route to host"}

We suspect it is related to new privacy settings in Sequoia related to local network access. The dialog to allow local network access does not appear.

Not sure if this is related, but bumping Go might resolve the issue:
golang/go#68678
Any interest in a pull-request that bumps Go to test this?

It might also be related to this one:
cirruslabs/tart#921

[timpeeters]

Just tested this. Permission dialog now pops-up. After approving, the no route to host error is gone.
Thanks for the help.