Thanks for this PR! I've been able to use this idea to follow the auth flow through the singlekey site and have it almost working without having to invoke the web browser (up to the point that I have an hcauth:// URL with the code). One potential downside is that sometimes after too many auth failures it will fall back to a captcha, which we will need to detect and inform the user.
We could display the captcha image for them, or just have them do a manual auth via the website which seems to clear the captcha state.
For the mqtt client config, can you separate that into a separate PR?
Thanks for this PR! I've been able to use this idea to follow the auth flow through the singlekey site and have it almost working without having to invoke the web browser (up to the point that I have an
hcauth://
URL with the code). One potential downside is that sometimes after too many auth failures it will fall back to a captcha, which we will need to detect and inform the user.We could display the captcha image for them, or just have them do a manual auth via the website which seems to clear the captcha state.
For the mqtt client config, can you separate that into a separate PR?