Unable to use #!connect kql in Azure Machine Learning Studio

[pshelton-skype]

Describe the bug

I'm trying to get a basic notebook hosted in Azure Machine Learning studio where it can be shared with others to perform KQL queries.

I figured out the correct arcane incantation to install the correct dependencies:

sudo apt-get install -y dotnet-sdk-8.0
dotnet tool install -g Microsoft.dotnet-interactive
dotnet interactive jupyter install

I had issues connecting to the C# kernel until I ran:

sudo ln -s /home/azureuser/.dotnet/tools/dotnet-interactive /usr/local/bin/dotnet-interactive
export DOTNET_ROOT=$(dirname $(realpath $(which dotnet)))

as described in https://techcommunity.microsoft.com/blog/machinelearningblog/running-ml-net--notebooks-in-azure-machine-learning-studio/1323238.

At this point, I can connect to the C# kernel and successfully run C# code.

I then install the Microsoft.DotNet.Interactive.Kql package with:

#r "nuget: Microsoft.DotNet.Interactive.Kql, *-*"

and attempt to connect to a cluster:

#!connect kql --kernel-name samples --cluster "https://help.kusto.windows.net" --database "Samples"

which runs and never completes with no authentication prompt.

Please complete the following:

Which version of .NET Interactive are you using? (In a notebook, run the #!about magic command. ):

• OS
  • Windows 11
  • Windows 10
  • macOS
  • Linux (Please specify distro) - Ubuntu 20.04.6 LTS
  • iOS
  • Android
• Browser
  • Chrome
  • Edge
  • Firefox
  • Safari
• Frontend
  • Jupyter Notebook
  • Jupyter Lab
  • nteract
  • Visual Studio Code
  • Visual Studio Code Insiders
  • Visual Studio
  • [X ] Other (please specify) - Azure Machine Learning Studio Notebooks

Environment configuration information:

(azureml_py38) azureuser@pshelton3:~/.dotnet/tools$ dotnet --version
8.0.404
(azureml_py38) azureuser@pshelton3:~/.dotnet/tools$ jupyter kernelspec list
Available kernels:
  python3                   /anaconda/envs/azureml_py38/share/jupyter/kernels/python3
  .net-csharp               /home/azureuser/.local/share/jupyter/kernels/.net-csharp
  .net-fsharp               /home/azureuser/.local/share/jupyter/kernels/.net-fsharp
  .net-powershell           /home/azureuser/.local/share/jupyter/kernels/.net-powershell
  ir                        /usr/local/share/jupyter/kernels/ir
  python310-sdkv2           /usr/local/share/jupyter/kernels/python310-sdkv2
  python38-azureml          /usr/local/share/jupyter/kernels/python38-azureml
  python38-azureml-pt-tf    /usr/local/share/jupyter/kernels/python38-azureml-pt-tf
(azureml_py38) azureuser@pshelton3:~/.dotnet/tools$ echo $PATH
/home/azureuser/bin:/home/azureuser/.local/bin:/anaconda/envs/azureml_py38/bin:/anaconda/condabin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin:/home/azureuser/.dotnet/tools:/home/azureuser/.dotnet
(azureml_py38) azureuser@pshelton3:~/.dotnet/tools$ echo $DOTNET_ROOT
/home/azureuser/.dotnet
(azureml_py38) azureuser@pshelton3:~/.dotnet/tools$ 

Screenshots

If applicable, add screenshots to help explain your problem.

[jonsequitur]

The current KQL authentication experience opens a web page and since it's running on the remote machine in this case, that approach won't work. This is work we know we have to do to improve various auth flows (#992).