Scene failed to load and selfies fail to load since February Hubs Cloud update

[wmurphyrd]

Starting with the February hubs cloud update, myself and at least one other discord have had some issues during local dev related the hostname being used in URLS.

My setup - developing locally against my hubs cloud using npm run start

Issue 1: Scene failed to load (discord discussion). The scene model and assets are being fetched not from localhost but instead from by pc's local network ip, e.g. 192.168.1.1, on Chrome this request will be blocked with a certificate error unless you navigate directly to that ip and add another certificate exception

Issue 2: selfies fail to load. Internal domain (AWS recipe 2) is not recorgnized as a non CORS proxy domain and the local ip is being used again, so I get a url like https://192.168.1.1:8080/cors-proxy/https://hub-assets.internal-hub-domain.com/files/aaaa-bbbb.png?token=aaabbbbb and that request gets rejected with 401 Unauthorized. Note, if I edit the url in the html to remove the proxy and fetch directly from the internal domain, then the image loads

[markusTraber]

Can confirm the same behaviour. Although my server is on DO and currently working on a branch with qa-stage as base.