Add a client-path parameter, and make the set token call vm specific

[yaacov]

Issue:
Currently a user need to specify the VNC client path and virtual machine full API URL
The VNC path is static and can be specified when starting the proxy
The virtual machine full API URL can be calculated from the vm name and namespace

What this PR do:

Breaking changes:

• proxy will no longer accept then argument
• proxy will look for vm name and namespace arguments

Example:

Deprecated:

signed_link="${proxyurl}/auth/jwt/set?token=${jwt}&then=/noVNC/vnc_lite.html?path=k8s${path}"
google-chrome "${signed_link}"

New format:

signed_link="${proxyurl}/auth/jwt/set?token=${jwt}&name=${vmname}&namespace=${vmnamespace}"
google-chrome "${signed_link}"

Running the CLI proxy:

./kube-gateway --help
  ...
  -client-path string
    	url endpoint for user client calls. (default "/noVNC/vnc_lite.html")
  ...

./kube-gateway -client-path  "/noVNC/vnc_lite.html"

Changes:

• Remove the then URL search argument from the proxy server arguments
• Add virtual machine name and namespace to the URL search argument from the proxy server arguments
• Add a -client-path CLI argument to kube-gateway CLI application
• Add -client-path CLI argument value to the example kuebernetes yaml deployment file
• Update base container image from ubi8/ubi-minimal to ubi9/ubi-minimal
• Update readme to use virtual machine name and namespace, instead of API path