Added ssh key parameter

[boussou]

Also:

• removed the -s parameter which is unknown to the restream command
• added -o LogLevel=ERROR to suppresses the post-quantum key exchange warning by setting the SSH log level to ERROR only

[Eeems]

Forcing the ssh key instead of allowing that to be set in the ssh config file seems like an anti-pattern to me.

[boussou]

But since we allow IP to be directly provided, it becomes usefull.

Moreover, I ended up doing this because the ssh config file is the wrong answer for the remarkable. Indeed it's IP changes depending on where you are, unless you use the USB.

This way I can easily use it with the usb cable but anytime I can also use it wireless just by providing the local WIFI IP of the remarkable.

[Eeems]

But since we allow IP to be directly provided, it becomes usefull.

It allows the host to be provided, which can just be an IP address, but ideally is just a unique host value that you configure in the ssh config file.

Moreover, I ended up doing this because the ssh config file is the wrong answer for the remarkable. Indeed it's IP changes depending on where you are, unless you use the USB.

You should be using USB if you want good performance, which always has the same IP. If you are on a network you control, which is the only place you should be using this, you would control the IP address used for wifi, which would also not change. So the config file is the correct answer for the reMarkable, just like any other ssh host.

[boussou]

I agree on both. It's just that... it is painful. I work usually in several places, I don't necessarily control the wifi network, but they are safe working places, so I often end up looking at the IP & just run it (I even have my personal version of the script with a few additions which just receives the IP for that purpose).

Regarding the perf over wifi, I find it good enough.

Btw I just stumbled upon oxide on your profile, does this work on RM2? I suppose yes?

[Eeems]

I agree on both. It's just that... it is painful. I work usually in several places, I don't necessarily control the wifi network, but they are safe working places, so I often end up looking at the IP & just run it (I even have my personal version of the script with a few additions which just receives the IP for that purpose).

The way your script solves things currently sacrifices the standard ssh config usecase for your edge case scenario. -i should only be passed to ssh if it was specified as a command line argument for the script, instead of forcing an identity file that may not exist or may be different than is expected in configuration.

Btw I just stumbled upon oxide on your profile, does this work on RM2? I suppose yes?

This is not the place to discuss this. The correct place would be the oxide repo, which already answers your question in the readme.