Getting started with PowerShell IoT on Raspbian (Raspberry Pi)

During a cleanup over the weekend I found a Raspberry Pi that wasn’t doing anything. I figured now that PowerShell Core had got to version 6.2.1 I should power it up and have a play around with maybe some sensors and outputs using PowerShell IoT. I ran into a couple of gotcha’s that took me some searching and head scratching to figure out, so I’m documenting them here for next time.

Installing PowerShell Core was straight forward. I followed the Microsoft Guide here and pretty quickly I was up and running with PowerShell Core 6.2.1.

PowerShell Core 6.2.1 on Raspbian.png

Enabling SPI and I2C Pins

By default on the latest Raspbian Stretch Lite build (that I installed – April 2019) you have to go into Preferences => Raspberry Pi Configuration and enable SPI and I2C if you are using displays etc.

GPIO Enable on Raspberry Pi for PowerShell IoT.png


Having built a number of projects with IoT Devices I thought I knew what GPIO Pins were and how to interact with them. However I quickly failed when trying to flip the outputs on GPIO pins using the Set-GPIOPin cmdlet from the Microsoft.PowerShell.IoT PowerShell Module. It turns out that the Microsoft PowerShell IoT Module utilises the Wiring Pi scheme for Pin numbering. Obviously having not done any IoT with a Raspberry Pi before I was not familiar with this underlying C Library. Therefore whilst I was trying to use the GPIO number 8 for example, what I really needed to be addressing was Pin 10.

The schematic below shows the 40 Pin GPIO connector with the WiringPi Pin numbers.


With that knowledge and a few Red Green Blue (RGB) LED’s wired to multiple GPIO outputs held high you can quickly get some nice colours.


With the SSD1306 library you can quickly and easily output some text to an OLED display.

Note: After connecting the display you will need to reboot your Raspberry Pi to get the display showing your text. I didn’t spend too much time on it but I did notice that it didn’t have many capabilities like the SSD1306 library from Adafruit does for Python like fonts, text size, text position etc.

New-OledDisplay | Set-OledText -Value "pwsh Core 6.2.1      pwsh Core 6.2.1      pwsh Core 6.2.1      pwsh Core 6.2.1      pwsh Core 6.2.1      pwsh Core 6.2.1      pwsh Core 6.2.1      pwsh Core 6.2.1"



The basics are there to set GPIO pins high and low which in turn could then trigger a relay that activates something else. The available libraries and examples only extend to a few other sensors and output devices. It was fun to have a quick mess around and understand the capabilities, and I’m sure it would be possible to develop some interesting projects with it. With Winter now here, maybe there will be a rainy weekend soon and I’ll be inspired to build something.

Windows Terminal with Tabs, on Steroids

PowerShell Cmdline Emojis Windows 10 Tabbed Terminal

At Microsoft Build last week, one of the many announcements was a new Windows Terminal.

If you spend anytime as an IT Support Person/ DevOps type role and you checkout that second link above you’ll be mightily keen for this new Terminal.

Tabs in a Terminal Window YES (heck I remember paying for a product to provide that to me in a browser) 15+ years ago; a Terminal Window that is a standard command prompt (with Unicode Support) YES; a Terminal Window for cross platform, CMD, PowerShell, PowerShell Core, Windows Subsystem for Linux DAMN YES. And of course you don’t want to have to wait for this, you want it now.

So did I, so I built the Preview Alpha Release. This post details how I did it.

Windows 10 Tabbed Terminal with icons
Windows 10 Tabbed Terminal with icons


There are a few hoops you need to jump through to get on this right now, as it isn’t available as a download. It will be coming to Windows 10 in a few months, but let’s get it now.

  • Become a Windows Insider by registering for a Windows Insider Account here
  • Have a Windows 10 v 1903 build (via registering for Windows Insiders above)
    • the process to do this I show below
  • Inside your Windows 10 machine you will then need;
    • Windows 10 SDK v 1903
    • Visual Studio 2017 (I use 2019)
      • Choose the following Workloads
        • Desktop Development with C++
        • Universal Windows Platform Development
        • For Visual Studio 2019, you’ll also need to install the “v141 Toolset” and “Visual C++ ATL for x86 and x64”
    • Git for Windows command-line

Windows 10 Test Machine Version 1903

I built a Windows 10 1709 Virtual Machine in Azure from the Azure MarketPlace. Having connected to it, I needed to enable the Windows Insider Program on it. To do that select;

Windows => Settings => Update & Security => Windows Insider Program => Get Started

Enable Windows Insider.PNG

Select Link an account and provide the account you used to sign up for Windows Insiders.

Link an Account.PNG

If, when you attempt to link an account you get a blank login window/page when being prompted for your Windows Insider Account you may need to make a couple of changes to the Windows 10 Local Security Policy Security Options. Below is the configuration of my test Windows Insider Windows 10 Virtual Machine. I’ve highlighted a few options I needed to update.

Local Security Policy.PNG

Select the Skip Ahead to the next Windows release to update Windows 10.

Skip ahead to the next Windows Release.PNG

If you are doing this like I am on a Windows 10 Virtual Machine in Azure, you’ll first go from build 1709 to 1803.

Windows 1709 to 1803.PNG

After Windows 10 has updated to 1803 log back in, go back to Windows Insider Program and chose Skip ahead to the next Windows release.

Skip ahead to the next Windows Release - 1903.PNG

Under Settings => Update & Security => Windows Update and select Check for Updates and you will see Windows 10 version 1903 become available.

1803 to 1903.PNG

Under Windows => Settings => Update & Security => Enable Developer Mode

Enable Developer Mode.PNG

Terminal Application

With the other dependencies detailed in the prerequisites above (Windows 10 1903 SDK, Visual Studio etc) downloaded and installed on your Windows 10 machine we can get on to the fun bit of building the new Terminal. Create a folder where you want to put the source to build the terminal and from a command prompt change directory into it and run the following commands;

Git clone
cd Terminal
git submodule update --init --recursive

Git Clone.PNG

Then in Visual Studio select Open a project or solution and open the Terminal Visual Studio Solution. Select SDK Version 10.0.18362.0 and No Upgrade for the Platform Toolset

Open the Solution in VS -1.PNG

Select Release and x64 and then from the Build menu Build Solution.

Build Release x64.PNG Finally, right click on CascadiaPackage and select Deploy


Terminal (Dev) will then be available through the Start Menu.

Windows Terminal Dev.PNG

Opening the Windows Terminal will give you a single Terminal Window. Press Cntrl + T to open an additional tab. 

Use the drop down menu to select Settings and you will be presented with the JSON configuration document. See (below under Icons for mine that enables CMD, PWSH, PowerShell, WSL – Ubuntu and WSL – Suse.


To have icons for your terminal tabs obtain some 32×32 pixel icons for your different terminals and drop them into the RoamingState directory under the Windows Terminal App. For me that is


Then update your profiles.json configuration file located in the same directory and add the name of the appropriate icon for each terminal.


As much as we use nice UI’s for a lot of what we do as Devs/IT Pro’s, there are still numerous tasks we perform using terminal shells. A tabbed experience for these complete with customisation brings them into the 21st century. Now the wait for another month or two to have it delivered as part of the next Windows 10 Build.

VSCode Virtual Environments using your Browser

It’s no secret I’m a huge fan of virtual environments and PowerShell. Late last year I wrote this series on Nested Virtual PowerShell Desktop Environments on Windows 10 & Windows Server 2019 in Azure A lot of the back story for that three post series was to have virtual environments for PowerShell.

Moving forward six months and I’m at the beginning of the journey towards migrating from PowerShell Desktop to PowerShell Core. The quickest way to get started with PowerShell Core is to use the Windows 10 feature of Windows Subsystem for Linux. Of course you should also have made (or be making) the migration from PowerShell ISE to VSCode.

So what if you could have Virtual VSCode environments accessible via a browser as your IDE for building and using PowerShell Core? Well you can.

Here is VSCode running a PowerShell Core command via a browser from a Ubuntu based Windows Subsystem for Linux environment in a Windows 10 Virtual Machine running in Azure.

Powershell Core on Linux on Windows via VSCode in Browser.PNG


A key component of this magic is Code Server. Code Server is an open source self-contained environment for VSCode designed to offload the IDE from lower powered environments (tablets, Chromebooks etc) and is currently supported on Linux and Mac (OSX) with Windows coming soon.

If you have an environment with either of those then all good, as you were. If you only have Windows then on Windows 10 install Windows Subsystem for Linux via the Microsoft Store App. My Windows System for Linux flavour is Ubuntu.

WSL Ubuntu.PNG

Code Server

Download Code Server from the Github Releases page here and save it to your local machine. Running Ubuntu I downloaded the Linux binary. Untar it using

tar -xf yourDownloadedversion.tar.gz

Untar Code Server

Change into the extracted directory with the code-server executable in and run code-server


Take note of the password as you will need this when you connect to Code Server via your browser.

Start Code Server

If the host you are running Code Server is remote (as mine is, in Azure) make sure you have networking configured for the default port (8443). I had to update my NSG for the inbound port of 8443.

Note: this isn’t a normal configuration, allowing source any for incoming ports. I have this locked down to the necessary source IP’s (but I’m not going to show them here).

Inbound Security Rules.PNG

Then in your browser you can hit the your host on port 8443 and provide the password that was displayed when starting Code Server. You will then have VSCode running in essentially a Virtual Environment accessible via a browser.

VSCode running in a Browser

PowerShell Core

My Windows Subsystem for Linux Ubuntu version is currently 18.04 and I already had downloaded and installed PowerShell Core for Ubuntu 18.04

lsb_release -a

Installation Instructions are here

Ubuntu Windows Subsystem for Linux Version.PNG

Install the PowerShell VSCode Extension along with any of the other extensions you regularly use. BOOM, you now have a Virtual VSCode environment for use with PowerShell Core on (and in my case on Ubuntu under Windows 10).

PowerShell VSCode Extension


What a crazy time to be in the IT industry. Nested disparate operating systems running IDE’s in browsers executing cross-platform scripting languages. Wow.

To run multiple sessions on the same host use the -p (port) command to specify a different port for each instance.

Code Server specifying Port.PNG