remote provisioning over RNS / LoRa not working

[unosonic]

tried a 100 times, disabled chrome local network checks, verified backend urls of meshchatx, checked paths and connectivity (can load the nomadnetwork node on the device, can rnprobe the probe_destination) etc., but it failes, rns.link.open timeout after 30000ms. any ideas?

[warrenjmcdonald]

I now have this working well with the latest Webconsole index.html and a new websocket to RNS proxy at https://github.com/attermann/ReticulumAPI

This avoids using Meshchatx altogether and is easy to install on linux.

Important steps include making sure you copy your identity file from rns nomadnet, as per ReticulumAPI install instructions, and adding the identity hash to the rnodes list of allowed management identities via a serial connection from the Rnode_console.

This also allows the retrieval of rnode status and metrics via the microreticulum nomadnet node page in Meshchatx or Columba once you have identified to the page.

I am comfortable enough with this setup to now deploy a few microreticulum_firmware rnodes in the wild.

[attermann]

Another thing to note...
There is discussion in the Provisioning doc about Chrome's "Local Network Access Checks". If you have tried that to no avail, I suggest just browsing to the ReticulumAPI REST endpoint (eg, https://127.0.0.1:8000/health) first before attempting to connect to over WSS from the web console. This should warn you about the self-signed cert but allow you to proceed anyway. This seems to trigger a security exception in Chrome that also extends to the WSS connection.
I know it's annoying, but short of acquiring and installing your own TLS certs into your ReticulumAPI instance, it's a necessary evil to circumvent Chrome security.