Occasionally, there arises a need for me to download directories stored on my remote server, often for backup purposes. Yet, on each occasion, I find myself scouring the Internet to determine the precise arguments required for the rsync utility.

Those days are now relegated to history, as I have chosen today to write down a future-oriented reminder right here. 😅

Without further ado, here is the definitive command to download a remote directory on your computer:

# Adapt the variables below
SSH_PORT=22
SSH_USER=root
SSH_SERVER=138.197.142.29
REMOTE_SOURCE_DIRECTORY=/var/opt/mealie
LOCAL_DESTINATION_DIRECTORY=.

# From your local machine
rsync \
  --archive \
  --compress \
  --human-readable \
  --partial \
  --progress \
  --rsh "ssh -p ${SSH_PORT}" \
  --stats \
  --verbose \
  "${SSH_USER}@${SSH_SERVER}:${REMOTE_SOURCE_DIRECTORY}" \
  "${LOCAL_DESTINATION_DIRECTORY}"

Explanation of the arguments

1. --archive: Preserves file permissions, ownership, timestamps, and more.
2. --compress: Enables compression of data during transfer, reducing the amount of data transmitted.
3. --human-readable: File sizes are displayed in a human-readable format (e.g., kB, Mb or gB).
4. --partial: If the transfer is interrupted, partially transferred files are keep instead of discarding them.
5. --progress: Displays a progress indicator.
6. --rsh "ssh -p ${SSH_PORT}": The communication should be done over SSH using port 22. It is useful if your server use another port for SSH.
7. --stats: Provides a summary of the transfer statistics once the process is completed.
8. --verbose: Increases the verbosity of the output, offering more detailed information.
9. "${SSH_USER}@${SSH_SERVER}:${REMOTE_SOURCE_DIRECTORY}:: The source directory on the remote server that you wish to back up.
10. "${LOCAL_DESTINATION_DIRECTORY}": The destination directory on your local machine where the backup will be saved.

Running out of server space can be a real headache, often requiring thorough investigation to pinpoint the issue. In this quick guide, I'll show you an invaluable command that can help you identify the culprits behind excessive storage consumption on your Debian server.

To uncover the largest files and directories eating up your server's precious space, you can employ the following command:

sudo du --all / | sort --numeric-sort --reverse | head --lines 15

Or alternatively:

sudo du -a / | sort -n -r | head -n 15

The output of this command provides a list of the largest files and directories, accompanied by the space they occupy:

25072264  /
20175684  /var
8949680   /var/lib
8783876   /var/lib/docker
8629384   /var/lib/docker/overlay2
6131060   /var/log
[...]

This resulted list helped me to swiftly identify the storage-intensive elements on my server. Armed with this knowledge, I implemented measures to mitigate their impact and free up valuable space.

Should you need to delve into a specific directory for a more thorough investigation, the command can be tailored accordingly. For instance, if you want to focus on the /var/log directory:

sudo du -a /var/log | sort -n -r | head -n 15

By integrating this command into your diagnostic toolkit, you can confidently tackle storage issues and prevent future space shortages on your server. The satisfaction of maintaining an organized and efficient server environment awaits—enjoy the journey! 😎

A few weeks ago, I embarked on the journey of obtaining access to the LinkedIn Community Management API for one of my application. However, I soon discovered that this privilege was exclusively reserved for verified companies. 🙈

As the automatic verification failed for my business, I had to provide some documentation that prove that Merisia was a legal business. And today, I'm excited to share my experience and the key documents I submitted, in case you find yourself in a similar situation:

• Business Registration: I provided a meticulous screenshot displaying Merisia's entry on the Registraire des entreprises.
• Domain Ownership Validation: I submitted an invoice confirming our ownership of the domain merisia.ca.
• Business Tax Identification: I also included a screenshot detailing our business tax numbers, sourced from Mon dossier - Entreprise.
• Inclusion in Business Registries: and to further substantiating our legitimacy, I included a screenshot affirming Merisia's inclusion in the public Canada's Business Registries.

With these meticulously gathered documents, I submitted them to the LinkedIn support team, who then channelled them to their legal department. The subsequent weeks were a waiting game, filled with anticipation. The culmination of this process was the exhilarating moment when I received confirmation of the granted access! 🤤

This experience stands as a testament to the efficacy of providing comprehensive documentation to establish the authenticity of a business. If you find yourself navigating the intricacies of verification, take heart from my journey – it truly works! 😎

Did you know that you can easily identify the files that have been changed within a Docker container's file system? This information can be valuable, especially when you want to ensure that all necessary files are backed up and aren't lost when the container restarts. 🙄

In this article, we will explore how to use the command made available by Docker effectively and even automate the process for multiple containers.

Listing changed files and directories

To list all the files and directories that have been added, changed, or deleted within a Docker container, you can use the "docker container diff" command. Here is the syntax:

docker container diff <CONTAINER_NAME_OR_ID>

Here is an example of its output for a MariaDB container that I use:

C /run
C /run/mysqld
A /run/mysqld/mysqld.pid
A /run/mysqld/mysqld.sock

This command will output a list of files and directories that fall into the following categories:

• A: added
• C: changed
• D: deleted

Automating the process

If you need to perform this check for multiple containers, manually finding and pasting each container's name or ID can be time-consuming. However, you can automate this process using the following script:

#!/bin/bash

# Get the names of all containers
container_names=$(docker container ls --format '{{.Names}}')

# Iterate over all container names
for container_name in $container_names
do
    # Execute `docker container diff` on the container
    docker container diff ${container_name}

    # Check if the command was successful or not 
    if [ $? -eq 0 ]
    then
        echo "The command 'docker container diff' has been executed with success on container ${container_name}."
    else
        echo "The command 'docker container diff' has failed on container ${container_name}."
    fi
done

By running this script, you will get the results for all your containers. The output will appear on the standard output, but you can easily modify the script to redirect the results to a file or multiple files according to your needs. 😉

Conclusion

By utilizing the docker container diff command and the provided script, you now have the ability to identify all the changed files within your Docker containers. This knowledge is crucial for ensuring that all necessary files are appropriately preserved, especially when working with container volumes. 🤗

If you're migrating your private GitLab repositories to a group, like I did for my personal business, you may encounter an error message like this in your pipelines:

WARNING: Failed to pull image with policy "always": Error response from daemon: pull access denied for registry.gitlab.com/merisia/tools/git-rsync, repository does not exist or may require 'docker login': denied: requested access to the resource is denied (manager.go:237:0s)

ERROR: Job failed: failed to pull image "registry.gitlab.com/merisia-inc/outils/git-rsync:1.2.0" with specified policies [always]: Error response from daemon: pull access denied for registry.gitlab.com/merisia/tools/git-rsync, repository does not exist or may require 'docker login': denied: requested access to the resource is denied (manager.go:237:0s)

Despite confirming that the URL and tag exist in the repository's Container Registry, the problem persists. 😓

After conducting some research, I discovered a related StackOverflow question where another user faced a similar issue. Although no solution was provided, I stumbled upon a helpful clue. The user encountered the same issue when moving a private repository to a group. 🤔

The thread mention the DOCKER_AUTH_CONFIG variable, which picked my curiosity. So I began to look on it and I found this documentation from a team of UToronto that helped me solve my problem. 🤩

As it's not impossible that I encountered again this problem in a distant future,I decided to share the solution I used through this blog post. I'll skip the detailed explanations, since the UToronto team has already covered the necessary commands. 🤗

Solution

Encode your GitLab credentials in base64

# The use of "-n" - prevents encoding a newline in the password.
echo -n 'GITLAB_USERNAME:GITLAB_PASSWORD' | base64

Create the DOCKER_AUTH_CONFIG variable in your CI/CD group settings

• Go to your GitLab group (or project).
• Under Settings > CI/CD > Variable, click the Add variable button.
• Set the key as DOCKER_AUTH_CONFIG and set the content with the following JSON:

{
    "auths": {
        "registry.gitlab.com": {
            "auth": "(Base64 content from above)"
        }
    }
}

• Make sure to uncheck the Protect variable and Expand variable reference options as they are optionals.
• Click the Add variable button to create this new variable.

Rerun your previously failed CI/CD pipeline

By following the previous steps, your pipeline should now be back on track! 🥳

If the issue persists, you can retry the previous steps or continue searching for an alternative solution! 😅

For a long time, I was not able to run Node.js 18 on my desktop. When I choose one of its version with nvm, I always ran into an error stating that a library was missing: node: /lib/x86_64-linux-gnu/libc.so.6: version 'GLIBC_2.28' not found (required by node). I was stuck with the latest Node.js 16 version available... 😥

But, it was something that was bothered me. I couldn't use the new features of Node.js LTS version, and I hopped that somehow, it would resolve by itself. Disclaimer: it hasn't. I tried from time to time to upgrade, but my attempts have been unsuccessful. Until now. 😮

I'm now happy to says that I manage to install and to use Node.js 18. How? By upgrading my Ubuntu version of Windows Subsystem for Linux (WSL) from 18.04 to 22.04. After a new search, I stumbled on this answer on StackOverflow from hallexcosta and it worked for me! I was able to update to Ubuntu 20.04 and with another answer by NotTheDr01ds, I updated to Ubuntu 22.04 as well. So, thanks a lot to these people! 🥰

As it's possible that I would need to update my WSL distribution on another computer, I decided to write here the commands I used for posterity. It could also be useful for you if you aren't either on the latest version of Ubuntu. 😉

Without further ado, here is the complete procedure. 🤗

From 18.04 to 20.04

Before continuing, make sure that you aren't on the latest version of Ubuntu by running the following command in your WSL terminal:

lsb_release -a

# No LSB modules are available.     
# Distributor ID: Ubuntu
# Description:    Ubuntu 18.04.6 LTS
# Release:        18.04
# Codename:       bionic

Remove the Snapd package

"For me it was necessary to remove the snapd package, because ubuntu was not allowing me to upgrade to 20.04 LTS version." – hallexcosta

As I didn't want to take any chances, I uninstall as well the Snapd package. I don't know if it would have made a difference for me, but just to be sure... 😋

sudo apt purge snapd

According to its Wikipedia page, snapd is a tool used to package and to deploy applications. The procedure doesn't reinstall it, but at least, it is noted here that it has been removed. To add it back should be simple if it is needed. 🙃

Update and upgrade packages

Always make sure that the operating system is up-to-date before taking a major upgrade! 😄

# Get updated software list for Ubuntu
sudo apt update

# Apply updates and patches
sudo apt upgrade

Install the update-manger-core package

"It’s essential to install this update manager core package this will trick the system into thinking there is a new LTS available and allow you to do an in-place upgrade." – hallexcosta

Again, I proceed exactly as I've been told. I wanted to quickly see if the procedure would work for me. 😅

sudo apt install update-manager-core

update-manager-core seems to be a simple package that manage release upgrades...

Install the new version

# Update Ubuntu
sudo do-release-upgrade

The command should take a couple of minutes to run. In my case, it has more than 600 MB to download. Go take a walk or grab a drink; once it's finished it will ask for a confirmation. 😉

Restart

Restart your computer to make sure everything has been updated and check if Ubuntu has been updated by typing the following command:

lsb_release -a

# No LSB modules are available.     
# Distributor ID: Ubuntu
# Description:    Ubuntu 20.04.5 LTS
# Release:        20.04
# Codename:       focal

Unfortunately, at the time of writing this article, Ubuntu 20.04 is not the latest LTS version. But the steps from passing to Ubuntu 22.04 are smaller than those before. Let's see them. 🤗

From 20.04 to 22.04

With the three following commands, you should be able to upgrade to the latest LTS version:

# Get updated software list for Ubuntu
sudo apt update

# Apply updates and patches
sudo apt upgrade

# Update Ubuntu
sudo do-release-upgrade

Again, restart your computer and check if your Ubuntu distribution has been upgraded:

lsb_release -a

# No LSB modules are available.     
# Distributor ID: Ubuntu
# Description:    Ubuntu 22.04.2 LTS
# Release:        22.04
# Codename:       jammy

When Ubuntu 24.04 will be released, I hope that the previous procedure remains the same, as it's really simple. 🤞

Conclusion

After upgrading to Ubuntu 22.04, you should now be able to use Node.js 18 on your computer. Install it if it's not already done and run a simple command to check if the error about the missing library is gone.

# Install the Node.js latest version available
nvm install 18.15.0

# Check if Node.js is able to run without errors
node --version
# v18.15.0

Thanks again to hallexcosta and NotTheDr01ds! 🥰

One month ago, I submitted a business plan as part of the Defi OSEntreprendre, using a new domain that I had not yet created a website for. I was worried that if a member of the jury clicked on one of my links, they would encounter a 404 error page, which would not be great for my project's credibility. 🙃

To buy myself some time, I decided to redirect the whole website to my existing website, where I could temporarily host the content for the challenge.

Initially, I tried to create a solution that would allow me to create multiple redirects, but it didn't work out. Instead, I created a quick proof-of-concept using Traefik, a tool I often use in my infrastructure. 😎

Solution

Here is the final solution I came up with using Traefik:

version: "3.8"

services:
  merisia:
    # As we are using Docker Swarm, labels need to be set inside deploy
    deploy:
      labels:
        traefik.docker.network: "traefik-net"
        traefik.enable: "true"
        traefik.http.routers.merisia.entrypoints: "https"
        traefik.http.routers.merisia.middlewares: "redirect-merisia-benjaminrancourt,default@file"
        traefik.http.routers.merisia.rule: "Host(`merisia.ca`,`www.merisia.ca`)"
        traefik.http.routers.merisia.tls.certresolver: "letsEncrypt"
        traefik.http.routers.merisia.tls.options: "intermediate@file"
        traefik.http.routers.merisia.tls: "true"
        traefik.http.services.merisia.loadbalancer.server.port: 80
        traefik.http.services.merisia.loadbalancer.sticky.cookie.httpOnly: "true"
        traefik.http.services.merisia.loadbalancer.sticky.cookie.secure: "true"
        traefik.http.middlewares.redirect-merisia-benjaminrancourt.redirectregex.regex: "^https?://(www\\.)?merisia\\.ca/(.*)"
        traefik.http.middlewares.redirect-merisia-benjaminrancourt.redirectregex.replacement: "https://www.benjaminrancourt.ca/$${2}"
        traefik.http.middlewares.redirect-merisia-benjaminrancourt.redirectregex.permanent: "false"
      resources:
        limits:
          cpus: '0.750'
          memory: 16M
        reservations:
          cpus: '0.001'
          memory: 6M
    # https://hub.docker.com/r/traefik/whoami/tags?page=1&ordering=last_updated
    image: "traefik/whoami:v1.8.7"
    networks:
      - traefik-net

networks:
  traefik-net:
    driver: overlay
    external: true

Let me explain how this solution works below. 😉

Explanations

1 – Router rule

traefik.http.routers.merisia.rule: "Host(`merisia.ca`,`www.merisia.ca`)"

This rule tells Traefik the original domain to redirect from. In this case, the domain is merisia.ca and its subdomain is www.

2 – Redirect regex middleware

 traefik.http.middlewares.redirection.redirectregex.regex: "^https?://(www\\.)?merisia\\.ca/(.*)"
 traefik.http.middlewares.redirection.redirectregex.replacement: "https://www.benjaminrancourt.ca/$${2}"
 traefik.http.middlewares.redirection.redirectregex.permanent: "false"

The regex redirects the original traffic to the replacement domain. It captures the original traffic using a regular expression (in regex) and redirects it to the replacement domain using a replacement pattern (in replacement).

The regex is designed to work with both HTTP and HTTPS (^https?://) and to capture all traffic that come from its www subdomain and its bare domain ((www\\.)?). All dots were also escaped accordingly (\\.).

The pathname is captured in the second and last capture group ($${2}).

3 – Middlewares

This label tells Traefik to use the previous middleware.

traefik.http.routers.merisia.middlewares: "redirection,default@file"

4 – Image

And finally, we also need to have a container that simply exists, as I don't think it's possible to have a docker-compose stack without a container! 😅

image: "traefik/whoami:v1.8.7"

The traefik/whoami image is a tiny web server that prints OS information and HTTP request to output. But, if you have configured everything correctly, you should never see its output, as you should be redirected to your second domain! 😉

Conclusion

In my case, the previous solution has helped me to have www.merisia.ca/linkedin temporarily redirect to www.benjaminrancourt.ca/linkedin, which finally redirect to my personal LinkedIn profile. I also had some others links, but I'm not able to disclosure yet. 😅

So, if you use Traefik, you now have a solution to temporarily redirect traffic from one domain to another. 😉

Good luck! 😎

Last week, I was looking to recreate my personal infrastructure on a new server, but somehow, I screwed something in the process and was left with a Docker container that I could not delete... 🙈

Something similar had already happened to me in the past, and the fix that I had found in that time was to completely remove and reinstall Docker. 🤯

As I was only at the beginning of my process and didn't want to lose any more time, I decided to choose again this solution. But, my previous notes were a little bit scrambled, so I decided to improve them and to share them with you. Just in case you need it! 😉

So, without further ado, here is the complete procedure with some explanations if you want to learn more! 😎

Step 1 – Remove Docker images, containers, and volumes

To completely remove Docker from your system, you'll need to remove any images, containers, and volumes that were created. To do this, use the following commands:

# Remove unused data
docker system prune --all --volumes

# Remove all services
docker service rm $(docker service ls -q)

# Remove all containers
docker rm -f $(docker ps -aq)

# Remove all images
docker rmi -f $(docker images -aq)

# Remove all volumes
docker volume rm $(docker volume ls -q)

Step 2 – Stop Docker service

Before removing Docker from your system, you should stop the Docker service by running the following command:

systemctl stop docker

This will ensure that Docker is not running in the background while you're removing it. 😉

Step 3 – Uninstall Docker

It's now time to remove Docker from the operating system. Run the followings commands:

# Uninstalls the Docker Community Edition (docker-ce) package from your Debian system
apt-get purge docker-ce -y

# Removes any remaining Docker packages and their dependencies that were not automatically removed by the previous command
apt-get autoremove --purge docker-ce -y

This will remove any packages that were installed as dependencies of Docker but are no longer needed.

Step 5 – Remove Docker group (optional)

If you had previously added users to the Docker users group to allow them to run Docker commands without sudo, you may want to remove the Docker group as well. Use the following command:

groupdel docker

Step 4 – Remove Docker directories

Next, you'll need to remove any directories associated with Docker, including its configuration files and its storage location. Use the following commands:

# Remove the Docker configuration files
rm -rf /etc/docker

# Remove the Docker Daemon's storage location
rm -rf /var/lib/docker

Conclusion

And that's it! You've now completely removed Docker from your system, along with any associated files and directories. 😎

If you want to reinstall Docker, you may now run the same commands as you typed when you installed it before! 😉

I write many user stories as part of my work and my entrepreneurial project. For each of them, I list all the acceptance criteria that will allow us to determine if the features have been successfully implemented. ☑️

To reduce the carbon footprint of the products that are developed, I have started to integrate more and more eco-responsible acceptance criteria. 🍃

These acceptance criteria stand out from the others because they focus on the environment. Specifically, they are intended to reduce the data needed to operate the website (or web application).

By minimizing this data, all the infrastructures on which they transact use less memory and electricity. ⚡

Did I pique your curiosity? 😜

Examples of sustainable acceptance criteria

If you are wondering what these famous criteria might look like, here are some real examples that I regularly put in my user stories:

• Only the data necessary for the application operation are transmitted between the different computer systems. 🖥
• Data is cached for as long as possible. 🙈
• All caches are invalidated at the appropriate time, such as when updating data or computer systems.
• SVG vector images are optimized with SVGOMG so there is no perceptible difference to the naked eye. 👁️
• Images are available in more compressed formats than traditional ones, such as WebP and AVIF. 🖼
• Resources are compressed in gzip and Brotli formats. 🗜️
• JavaScript, HTML and CSS files are minified and uglified.

Of course, these acceptance criteria are not always applicable to all user stories. Some of them are mainly intended for the design of the product itself. 😉

Conclusion

Now that you know that it is possible to write eco-responsible acceptance criteria, do you plan to add them to your next user stories? 🤗

Are you able to determine the difference between the two images below? Is it the same picture?

You probably guessed it, but it's not the same picture! The image on the right is actually an optimized version of the original image on the left.

This improvement was made manually using a tool I discovered this year in on of the latest Frontend Focus newsletter: SVGOMG.

SVGOMG is the Missing Graphical user interface of a Node.js tool specialized in optimizing SVG files, named SVGO.

SVGOMG's interface is straightforward to use and can be summarized in three steps:

• Upload your SVG image
• Play with the different controls offered to have an almost identical image, but with a lower weight than the original image
• Download the resulting SVG image

One of the advantages of using a GUI and going about it with a manual process is that we can more easily ensure the quality of the resulting image.

My favicon's image size went from 73,728 bytes to 20,480 bytes! This represents a saving of approximately 72% of the original image! Impressive, right?😮

You will now have one more tool in your toolbox! 😋

Last year, on December 6, 2021, the Ville de Sherbrooke adopted its Open Data Policy at its regular council meeting.

To celebrate its first anniversary, I decided to create this reference article on open data in Sherbrooke. 🥳

This article may also help the participants of the next HackQC, an IT competition using open data, whose theme will be "Data at the service of the ecological transition". 😎

Are you ready to dig deeper into open data in Sherbrooke? 😉

Frequently Asked Questions

What is Open Data?

Although I have already mentioned the term several times in this article, I have never defined this one. But what exactly is open data? See the definition on the linguistic showcase of the Office québécois de la langue française (OQLF):

Non-personal raw data free of rights, produced or collected by a public or private organization, and accessible to citizens via the Internet.

These data are generally delivered in one or more open formats (not requiring proprietary software), to facilitate their reuse.

Why does the Ville de Sherbrooke publish open data?

As part of its open data policy, the Ville de Sherbrooke has identified several advantages to publishing open data:

• data can be enriched and its potential increased;
• the citizens can obtain the desired information themselves;
• useful applications for the community or research can be established;
• and it encourages innovation and sustainable economic growth.
  

Where can I find the Ville de Sherbrooke's open data?

The data can be found on the Partenariat Données Québec website. As of this writing, 34 datasets have been released.

Jeu de données - Données Québec

ville-de-sherbrooke

The open data produced by the geomatics team can also be consulted via a dedicated site linked to ArcGIS, a suite of geomatics software used by the Ville de Sherbrooke.

Données ouvertes de la Ville de Sherbrooke

Portail de données ouvertes de la Ville de Sherbrooke

What open data does the Ville de Sherbrooke offer?

By combining the two previous sources, I was able to list the following 34 datasets:

1. 🌎 Developed areas
2. 🌎 Boroughs
3. 🌎 Buildings
4. ⚠️🌎 Municipal buildings and services
5. Real-time event calendar
6. 🌎 Contours and Digital Elevation Model (DEM)
7. Dataset requests
8. 🌎 Electoral districts
9. Defibrillators
10. Ecocentres - list and schedules
11. Events - SociéTIC
12. Events in the Sherbrooke region (What to do - GVQ standardization)
13. 🌎 Parking meters
14. 🌎 Public Safety Incidents
15. Login information
16. Residual materials - Materials accepted
17. Walls dedicated to temporary and permanent graffiti
18. Where to eat?
19. ⚠️🌎 Road works
20. 🌎 Parking signs
21. ⚠️🌎 Pools and beaches
22. Cycle paths
23. 🌎 Street snow removal priorities
24. Urban perimeter
25. What to do?
26. Directory of businesses in the Sherbrooke region
27. 🌎 Waste collection channels
28. 🌎 Street Segments
29. 🌎 Walking trails
30. 🌎 Landmarks
31. 🌎 Public car parks
32. Park Recreational Structures
33. Public transport
34. 🌎 Work in progress
35. ZAP: access points
36. 🌎 Landslide areas
37. 🌎 Flood zones
    

The datasets identified by a globe (🌎) are the datasets from the geomatics software ArcGIS.

As for those identified by an exclamation mark (⚠️), these datasets do not seem to be present on the Données Québec website yet.

Which Sherbrooke organizations participate in open data?

According to Données Québec, the main organizations and paramunicipal organizations are, in alphabetical order:

• Destination Sherbrooke
• Entreprendre Sherbrooke
• SociéTIC
• Société de transport de Sherbrooke (STS)
• Ville de Sherbrooke
• ZAP Sherbrooke

Who to contact to request the release of new datasets?

According to the Open Data Policy, the Communications Department is responsible for receiving requests for new datasets. This service can be reached by email at communications@sherbrooke.ca.

Under what license is the Ville de Sherbrooke's open data published?

All open data is released under the Attribution 4.0 International License (CC BY 4.0).

Creative Commons — Attribution 4.0 International — CC BY 4.0

cc.logo.white

Under the terms of the license, it is therefore permitted to:

• Share — copy, distribute and communicate the material in any format and by any means
• Adapt — transform and create from the material for any use, including commercial

under the following conditions:

• Attribution — The dataset must be credited and a link to its license must be embedded. In addition, all changes made to the data must be indicated. This information should appear, by all reasonable means, without suggesting that the dataset owner supports you in how you use their dataset.
• No Additional Restrictions — No legal requirement or technical measure can prevent others from using the same dataset.

How do I know when the Ville de Sherbrooke publishes or updates datasets?

Those interested can subscribe to the Atom feed below with their favourite feed reader, such as Feedly.

Données Québec - Custom query

Custom query

Where can I find examples of research papers using open data on Sherbrooke?

I don't know if the Ville de Sherbrooke maintains such a registry, but here is a non-exhaustive list of research reports that I found while writing this article:

• [French] Évaluation du potentiel des rues de la Ville de Sherbrooke à être transformées en rues partagées, by Alexandre Cailhier (December 29, 2020)

References

• [French] Politique de données ouvertes (ADM-2115), on the Ville de Sherbrooke website

In one of my previous posts on linting PowerShell scripts, I wasn't impressed with how GitLab CI displays PSScriptAnalyzer logs. I ended my post by asking my readers if they know how we could improve the readability of logs and one reader, Jesse Chisholm, suggested I try the script below.

function Show-LintMessages()
{
  $output = ""
  $input|ForEach-Object -Process {
    $I = $_;
    $output += ($I | Select RuleName,Severity,ScriptName,Line | Format-Table | Out-String);
    $output += ($I | Select Message | Format-Table | Out-String);
    }
  $output | Out-Host;
}

# Demonstration of function
Invoke-ScriptAnalyzer -ScriptDefinition '"b" = "b"; function eliminate-file (){}' | Show-LintMessages

After trying it, it was exactly what I was looking for! 🤩

But, since I was using PSScriptAnalyzer with Docker, I didn't want to declare this function every time I called the analyzer... So I started looking for how to make it available in my Docker container. 🤔

After a while, I finally managed to install the function and make it available in the terminal with a PowerShell module. By adding it to the /usr/local/share/powershell/Modules folder, the module is automatically imported and every user who connects to the Docker container has access to the Show-LintMessages function! 🤗

COPY Show-LintMessages.psm1 /usr/local/share/powershell/Modules/Show-LintMessages/Show-LintMessages.psm1

In GitLab CI, I also had to pipe the output of PSScriptAnalyser to the new command.

 - pwsh -Command "Invoke-ScriptAnalyzer -EnableExit -Recurse -Path . | Show-LintMessages"

Unfortunately, the message was truncated in the GitLab CI job output... 😓

'Select' is an alias of 'Select-Object'. Alias can introduce possible problems and make scripts hard to maintain. Pleas…
RuleName                  Severity ScriptName             Line
--------                  -------- ----------             ----
PSAvoidUsingCmdletAliases  Warning Show-LintMessages.psm1    6
Message
-------
'Select' is an alias of 'Select-Object'. Alias can introduce possible problems and make scripts hard to maintain. Pleas…
RuleName              Severity ScriptName Line
--------              -------- ---------- ----
PSAvoidUsingWriteHost  Warning test.ps1      1
Message
-------
File 'test.ps1' uses Write-Host. Avoid using Write-Host because it might not work in all hosts, does not work when ther…

So I tweaked the function a bit, and after a few tries, I finally ended with the script below.

function Show-LintMessages() {
  $input|ForEach-Object -Process {
    Write-Output "--------------------------------------------------------------------------------";
    $PSItem | Select-Object RuleName,Severity,ScriptName,Line | Format-Table -Wrap | Out-String -Width 80;
    Write-Output " ";
    $PSItem | Select-Object Message | Format-Table -Wrap | Out-String -Width 80;
    Write-Output "--------------------------------------------------------------------------------";
    Write-Output " ";
  }
}

And the output is now much better in GitLab CI! 😄

--------------------------------------------------------------------------------
RuleName              Severity ScriptName Line
--------              -------- ---------- ----
PSAvoidUsingWriteHost  Warning test.ps1      1

Message
-------
File 'test.ps1' uses Write-Host. Avoid using Write-Host because it might not
work in all hosts, does not work when there is no host, and (prior to PS 5.0)
cannot be suppressed, captured, or redirected. Instead, use Write-Output,
Write-Verbose, or Write-Information.
--------------------------------------------------------------------------------

It's more readable, isn't it? 😉

If you're a creator using Ghost CMS (or a Ghost theme developer), you might want your website to be fast to provide a great experience for your users. There are many ways to optimize the performance of a Ghost website. Today, I want to talk about one way that seems to be under looked according to this search: by optimizing card assets.

One thing I absolutely love about Ghost is that they offer so many tools to enhance your post content: buttons, bookmarks, galleries, headers, products, audio, video, and more.

But, if you don't use these features, chances are your website is unnecessarily loading JavaScript and/or CSS related to them. How can you quickly check if this is your case?

Open your theme's package.json file and check if you have the card_assets property in it.

{ 
  "config": {
    "card_assets": true
  }
}

If you can't find this setting inside your theme's package.json file, the default value for this property has been true since Ghost 5.0.

So, what does this imply? These settings tell Ghost to automatically provide and merge many default CSS and JavaScript assets into a cards.min.css and a cards.min.js files.

Take a look at these files here (5.0 kB) and here (2.0 kB) for the demo.ghost.io website. There's a lot of code, isn't it? 😲

But, did you know that you can exclude the features you don't use? I did, and my cards.min.css file is down to 1.6 kB, a reduction of 3.4 kB (68%)! And my cards.min.js file has been completely deleted!

It turns out there's an exclude option you can use to remove the default card implementations in Ghost.

{ 
  "config": {
    "card_assets": {
      "exclude": [
        "audio",
        "blockquote",
        "bookmark",
        "button",
        "gallery",
        "nft",
        "product",
        "toggle",
        "video"
      ]
    }
  }
}

You can also disable all cards assets by setting card_assets to false.

{ 
  "config": {
    "card_assets": false
  }
}

A simple configuration could immediately reduce your website's response time. Isn't that awesome? 🤗

FAQ

What files are automatically included in the cards.min.css and cards.min.js files?

If you look at this CSS directory and this JavaScript directory, you'll find all the files that are merged and minified by Ghost. A theme developer will probably want to exclude some of these features if they have already implemented them.

As of July 31st, 2022, the included files were:

What would a configuration that disables all default implementations look like?

If you listed all the previous card names, it would look something like below:

{ 
  "config": {
    "card_assets": {
      "exclude": [
        "audio",
        "blockquote",
        "bookmark",
        "button",
        "callout",
        "file",
        "gallery",
        "header",
        "nft",
        "product",
        "toggle",
        "video"
      ]
    }
  }
}

Can I use an allowlist (include) instead of a disallow list (exclude)?

According to the current CardAssetService implementation, you can, but I haven't been able to get it to work yet. Let me know if you succeed! 😋

// CASE: the theme has declared an include directive, we should include exactly these assets
// Include rules take precedence over exclude rules.
if (_.has(this.config, 'include')) {
  return {
    'cards.min.css': `css/(${this.config.include.join('|')}).css`,
    'cards.min.js': `js/(${this.config.include.join('|')}).js`
  };
}

Often, I end up with a list of objects and want to find some items on the list using a unique identifier. One example: enhancing other data that relates to an item on my list.

There are many ways to do this, but some are better than others, especially when there is a lot of data.

Let's see with a simple example:

const data = [{
  id: 3,
  name: "New Brunswick"
 }, {
  id: 8,
  name: "Ontario"
 }, {
  id: 10,
  name: "Quebec"
 },
];

If we want to retrieve the name of the province identified by the id 10, how do we do it? 🤔

const province10 = `?`;

Let's see some solutions we could use! 🤗

Solution #0 - Using find

A simple solution could be to use the find function to retrieve the first element that satisfies our search:

const province10 = data.find(element => element.id === 10);

This works perfectly, but it may not be quick if you need to find a lot of items and/or multiple times. Let's see if we can do better. 😋

Array.prototype.find() - JavaScript | MDN

The find() method returns the first element in the provided array that satisfies the provided testing function. If no values satisfy the testing function, undefined is returned.

MDN Web Docs.logo-m{fill:var(--text-link)}

Solution #1 - Using reduce

What if we had the following data in our hands:

{
  3: {id: 3, name: 'New Brunswick'},
  8: {id: 8, name: 'Ontario'},
  10: {id: 10, name: 'Quebec'},
}

Some people call this structure a dictionary or an index. If we had it, we could directly access the element we want using its identifier: dictionary[10]. Wouldn't that be easier? 🤔

Let's see how we can create it using one of my favourite array functions: reduce! 🥰

Array.prototype.reduce() - JavaScript | MDN

The reduce() method executes a user-supplied “reducer” callback function on each element of the array, in order, passing in the return value from the calculation on the preceding element. The final result of running the reducer across all elements of the array is a single value.

MDN Web Docs.logo-m{fill:var(--text-link)}

const dictionary = data.reduce((dictionary, province) => {
  dictionary[province.id] = province;
  return dictionary;
}, {});

const province10 = dictionary[10];
// {id: 10, name: 'Quebec'} 

The biggest advantage of this structure is that you only need to iterate on your array one time (to build the dictionary). If you have a lot of data, you should be able to get the element you want faster. 🤑

But, if you're not a fan of functional programming, I've another solution you could use. 😉

Solution #2 - Using object destructuring

While writing this post, I found another elegant solution that I like, but wouldn't use in a good codebase because it's more complex.

const dictionary = Object.assign({}, ...data.map((x) => ({[x.id]: x})));

Do you understand easily this one? Let me help you if not.

const dictionary = Object.assign({}, ...[
    {8: {id:8,name:"Ontario"}},
    {10: {id:10,name:"Quebec"}},
])

This solution use destructuring to create a dictionary with lots of smaller objects. All the generated objects are merged into a simpler one.

More complex, right? 😅

Destructuring assignment - JavaScript | MDN

The destructuring assignment syntax is a JavaScript expression that makes it possible to unpack values from arrays, or properties from objects, into distinct variables.

MDN Web Docs.logo-m{fill:var(--text-link)}

Conclusion

I hope this post introduced you to the concept of a JavaScript dictionary and that it helps you in the future! 🤗

When I decided to make full use Ghost as my CMS (Content Management System), I had to reimplement some features that I previously programmed. As I wanted to quickly ship this new version into production, I chose to skip some features that I might add again in the future. One of these features was that my website was a Progressive Web App (better known as a PWA).

PWA uses Service Workers, a small JavaScript file that provides certain features like an offline experience to users. Unfortunately for me, the service worker is cached by the browser and if I deleted that file, the service workers installed on my visitors' browsers would not be updated because the file would result in a 404 error page.

Since I didn't want to ask my visitors to manually clear site data, I had to find a way to delete it programmatically. 🤖

Fortunately, I knew there was a way to remove it using a self-destructing service worker. 🧨

Simply replace your service-worker.js (or sw.js) file and the next time a user visits your website, the service worker on its navigator will self-destruct! 💥

// Inspired from https://github.com/NekR/self-destroying-sw
self.addEventListener("install", (event) => {
  self.skipWaiting();
});

self.addEventListener("activate", (event) => {
  self.registration
    .unregister()
    .then(() => self.clients.matchAll())
    .then((clients) => {
      clients.forEach((client) => {
        if (client.url && "navigate" in client) {
          client.navigate(client.url);
        }
      });
    });
});

A useful snippet to keep on hand if you play with service workers! 🤩

And if your want to know more about removing old services workers, Chrome Developers has a great article on that:

Removing buggy service workers - Chrome Developers

How to fix a service worker that is causing problems.

Chrome Developers

Have fun playing with service workers! 🤗