<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Linux on Matthew Shields</title>
    <link>https://mshields.name/tags/linux/</link>
    <description>Recent content in Linux on Matthew Shields</description>
    <generator>Hugo -- gohugo.io</generator>
    <copyright>© 2022 - 2026 Matthew Shields</copyright>
    <lastBuildDate>Sat, 20 Jan 2024 00:00:00 +0000</lastBuildDate><atom:link href="https://mshields.name/tags/linux/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Setting Up Raspberry Pi Camera on Ubuntu 20.04 64 bit for Raspberry Pi 4</title>
      <link>https://mshields.name/blog/2024-01-20-setting-up-rapberry-pi-camera-on-ubuntu-20-04-64-bit-for-raspberry-pi-4/</link>
      <pubDate>Sat, 20 Jan 2024 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2024-01-20-setting-up-rapberry-pi-camera-on-ubuntu-20-04-64-bit-for-raspberry-pi-4/</guid>
      <description>Preamble This was a pig for no good reason, so hopefully someone else doesn’t have to go through the same pain I did.
For clarity I was using a Raspberry Pi 4 running Ubuntu 20.04 64 bit with a Pi Zero camera connected using an adapter ribbon cable. Ensure everything is properly connected before starting.
How To Edit your /boot/firmware/config.txt, adding the following… start_x=1 gpu_mem=512 disable_camera_led=1 start_file=start_4x.elf fixup_file=fixup_4x.dat sudo apt-get update sudo apt-get upgrade sudo reboot When your system comes back up you should now have your camera listed as a video device.</description>
    </item>
    
    <item>
      <title>Setting Up CUDA 11.8 and Pytorch on Ubuntu 20.04 with Secure Boot Enabled</title>
      <link>https://mshields.name/blog/2024-01-20-setting-up-cuda-11-8-and-pytorch-on-ubuntu-20-04-with-secure-boot-enabled/</link>
      <pubDate>Fri, 19 Jan 2024 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2024-01-20-setting-up-cuda-11-8-and-pytorch-on-ubuntu-20-04-with-secure-boot-enabled/</guid>
      <description>Preamble I recently got a fresh daily drive laptop which happens to have secure boot enabled. Typically this has added yet more complication to getting setup with CUDA and PyTorch. Hopefully the following reference can help someone else out or even just me if I need it again in the future.
How To 1. Purge system of NVIDIA CUDA in case of a previous failed install sudo rm -r /var/lib/dkms/nvidia sudo rm /etc/apt/sources.</description>
    </item>
    
    <item>
      <title>Using tmux to Manage a Fleet of Robots</title>
      <link>https://mshields.name/blog/2023-09-04-using-tmux-to-manage-a-fleet-of-robots/</link>
      <pubDate>Sun, 03 Sep 2023 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2023-09-04-using-tmux-to-manage-a-fleet-of-robots/</guid>
      <description>Preamble The title of this post ought to be “Using tmux to Manage a Fleet of Robots Incorrectly”. When I say manage here, I am talking about system configuration, code deployment and systems administration type tasks. Long term you really should be utilising tools like Ansible, Docker and Airbotics to keep your fleets Linux systems running as you intend. However, in reality it is often needed to do some quick maintenance operation to many Linux systems (or robots) at once even in the early days of a robotics startup when you might have a small fleet and maybe not yet settled on a setup using one of the aforementioned “proper” tools.</description>
    </item>
    
    <item>
      <title>PREEMPT RT Install for Ubuntu 20.04</title>
      <link>https://mshields.name/blog/2023-08-30-preempt-rt-install-for-ubuntu-20-04/</link>
      <pubDate>Tue, 29 Aug 2023 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2023-08-30-preempt-rt-install-for-ubuntu-20-04/</guid>
      <description>Preamble I have recently been working with robotic systems that have a greater requirement for deterministic operation. While Ubuntu will likely never be an RTOS the PREEMPT RT patch does allow for soft real-time constraints on specified threads which can be enough of a step up for many research applications.
While there are other guides on how to do this online (some linked below) this post will present as a record for the steps I needed to follow recently for my system.</description>
    </item>
    
    <item>
      <title>Using Chrony to Synchronise System Clocks for Robotic Platforms</title>
      <link>https://mshields.name/blog/2023-08-30-using-chrony-to-synchronise-system-clocks-for-robotic-platforms/</link>
      <pubDate>Tue, 29 Aug 2023 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2023-08-30-using-chrony-to-synchronise-system-clocks-for-robotic-platforms/</guid>
      <description>Preamble If you are utilising multiple Ubuntu systems in your robotic deployment it can be advantageous to have the system clocks synchronised. This is a fairly common occurrence with engineers looking to utilise the distributed nature of ROS and other middleware. If you are interested in this you might also be interested in my previous post on sharing internet connections over a LAN.
Requirements For this you’ll need at least two separate Ubuntu systems connected to each other over a LAN, VPN or similar.</description>
    </item>
    
    <item>
      <title>Network Manager Useful Commands</title>
      <link>https://mshields.name/blog/2023-08-11-network-manager-useful-commands/</link>
      <pubDate>Fri, 11 Aug 2023 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2023-08-11-network-manager-useful-commands/</guid>
      <description>Preamble Some Basics Establishing Connections with Static IPs Identify Network Interface Names Auto Connect Adding an IP Route Multiple IP Addresses Sharing an Internet Connection Hey you! Found this useful or interesting? Any question, comments, corrections or suggestions? Preamble Network Manager is great and is an unsung hero of our modern world.
Some Basics See which connections you have already setup…
sudo nmcli c s See details of a given connection…</description>
    </item>
    
    <item>
      <title>Sharing Python Code as Modules Between ROS Packages</title>
      <link>https://mshields.name/blog/2022-03-10-sharing-python-code-as-modules-between-ros-packages/</link>
      <pubDate>Wed, 23 Mar 2022 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2022-03-10-sharing-python-code-as-modules-between-ros-packages/</guid>
      <description>Preamble We can wrap our code in Python modules within our catkin work space to share with other packages in our ROS environment. Why would we do this? It can be great for code re-usability and wrapping commonly used functions and classes. This can dramatically cut down on repetition of code across your ROS packages.
How To I’m going to start with a fresh catkin work space but feel free to adapt to suit your environment, package naming and file naming conventions.</description>
    </item>
    
    <item>
      <title>Running ROS Nodes on Boot</title>
      <link>https://mshields.name/blog/2022-03-16-running-ros-nodes-on-boot/</link>
      <pubDate>Wed, 16 Mar 2022 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2022-03-16-running-ros-nodes-on-boot/</guid>
      <description>Preamble This is a reworking of a blog post that used to be on the Rover Robotics blog, which they have since taken down - sad face.
We often want our robots to be ready to use at the point of power on. We don’t want to be having to call every ROS launch in our stack every time we want to run our robot. Even if you combine all your ROS launch files into one ROS launch to rule them all the following guide will be useful to get you to a more polished robotic product, instead of a development tool.</description>
    </item>
    
    <item>
      <title>Connecting to a ROS Master Over a VPN</title>
      <link>https://mshields.name/blog/2022-03-12-connecting-ros-master-over-vpn/</link>
      <pubDate>Sat, 12 Mar 2022 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2022-03-12-connecting-ros-master-over-vpn/</guid>
      <description>Preamble This is a nice build on from my previous post on creating a VPN server. If you are using ROS and a VPN server it is possible to remotely connect to the ROS master on one of your robots from either another robot or other device. Basically the only requirement is that both systems are clients on the VPN and can run ROS. It also helps if you have provisioned each system with a static IP address, again details are in my previous post on creating a VPN server.</description>
    </item>
    
    <item>
      <title>Custom SSH Banner on Login</title>
      <link>https://mshields.name/blog/2022-03-10-custom-ssh-banner-on-login/</link>
      <pubDate>Thu, 10 Mar 2022 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2022-03-10-custom-ssh-banner-on-login/</guid>
      <description>Preamble When logging into a Linux system over ssh you can greeted with a text banner prior to entering your password. There are a number of reasons you might want to do this.
adding company branding providing a brief description about how this system is being used providing warnings and reminders to users about access How To Log into your system over ssh or get local terminal access.
sudo nano /etc/ssh/sshd_config Down near the bottom you will be able to set the Banner path.</description>
    </item>
    
    <item>
      <title>Setting Up a Robot LAN with Internet Connection Sharing</title>
      <link>https://mshields.name/blog/2022-02-26-setting-up-a-robot-lan-with-internet-vpn-connection-sharing/</link>
      <pubDate>Sat, 26 Feb 2022 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2022-02-26-setting-up-a-robot-lan-with-internet-vpn-connection-sharing/</guid>
      <description>Preamble It can be common to want to run multiple embedded computers on a robotic system. Having multiple systems on a hard LAN makes sense, to enable low latency communication between them. If the robot can connect to a local WiFi source that provides an internet connection then it can be quite easy for each embedded system to connect to the wireless network also, as most will come with a WiFi card.</description>
    </item>
    
    <item>
      <title>Setting Up an OpenVPN Server on AWS EC2</title>
      <link>https://mshields.name/blog/2022-02-26-setting-up-openvpn-server-on-aws-ec2/</link>
      <pubDate>Sat, 26 Feb 2022 00:00:00 +0000</pubDate>
      
      <guid>https://mshields.name/blog/2022-02-26-setting-up-openvpn-server-on-aws-ec2/</guid>
      <description>Preamble Using This Guide Cloud Compute Setup Creating an EC2 Instance Selecting an AMI Selecting an Instance Type Security Group Static/ Elastic IP Address Server Side Setup Connecting Using SSH Modify the Hostname Installing OpenVPN and Easy RSA Configuring the OpenVPN Server Packet Forwarding and Firewall Setting up the Server Side Keys and Certs Generating Client Keys Creating the Unified .ovpn Client File Client Setup Connecting Client Side Network Manager Options Alternatives Assigning Clients a Static IP Server Maintenance Operations Useful Commands Issuing Additional Clients the Easy Way Revoking a Client Resources and Further Reading Hey you!</description>
    </item>
    
  </channel>
</rss>
