Well, I did a podcast back in Episode 22 (crikey! 2022) covering a few of the key points regarding Docker vs WASM.

It seems WASM (Web Assembly) and WASI (Web Assembly Infrastructure) is continuing to gain strength. WASM containers could be a more lightweight option and will run anywhere - no need to switch from Linux to Windows containers.

Anyway, take a listen to find out more! :)

Some links that can provide a deeper dive if you wish:

InfoWorld Article on WASM vs Docker

Yes! You can run WASM based Docker stuff!

If you really must support me, please do here via Patreon :)

.. and further information on the Docker infrastructure vs Containerd runtime:

Docker

Docker is a comprehensive platform for developing, shipping, and running applications inside containers. It includes several components:

Docker Engine: The core part of Docker, responsible for creating and managing containers.

Docker CLI: The command-line interface for interacting with Docker.

Docker Compose: A tool for defining and running multi-container Docker applications.

Docker Hub: A registry for storing and sharing container images.

containerd

containerd is a container runtime that focuses solely on managing the container lifecycle. It was originally part of Docker but was later separated to become an independent project. Key features include:

Container Lifecycle Management: Handles the creation, execution, and supervision of containers.

Image Management: Manages container images, including pulling from registries and storing locally.

Storage and Networking: Manages storage and networking for containers.

OCI Compliance: Uses Open Container Initiative (OCI)-compliant runtimes like runc for standardization and interoperability.

Key Differences

Scope: Docker is a full-fledged platform with tools for building, shipping, and running containers, while containerd is a more lightweight runtime focused on managing the container lifecycle2.

Components: Docker includes additional tools like Docker Compose and Docker Hub, whereas containerd is primarily concerned with the runtime aspects.

Usage: Docker is often used for development and deployment workflows, providing a user-friendly interface and additional features. containerd is used as a backend component in larger systems like Kubernetes, where it handles the low-level container operations.

In summary, Docker provides a complete ecosystem for container management, while containerd focuses on the core runtime functionalities, making it suitable for integration into other container orchestration systems.

Well, I originally took a look at Retool a while ago. It is a really rich application builder that allows self-hosting too (as a Docker container). I liked the idea of this given a company might want to have control of apps within their own infrastructure.

Retool as come along leaps and bounds and now:

* Still has an amazingly rich application building interface with lots of integrations options including your own API targets

* “Hands-Free” PostgreSQL database! Use whatever datastore you want (including the Retool internal one) or indeed PostgreSQL to have full database control either via SQL code or SQL GUI :)

* Has a really slick workflow module - you can trigger things, perform actions, and conditions and of course lots of integrations :)

* Mobile! Just investigating this, but check it out for yourself!

Hey, subscription to this podcast is FREE if you want to subscribe! I hadn’t realised before it forced you to pay ;) Of course, if you wish to or must pay, then please feel free with a paid subscription or via Patreon

Well, AI is certainly helping in all areas of life, for better or worse!

I decided to give GitHub Copilot a run for its money to attempt to create, or at least help, a command-line tool in dotnet to extract XML schema information (yes, they are still a “thing”) to generate SQL creation scripts i.e. table create / stored procedures and so on.

As a bonus, even though I know nothing of the Rust language, my good friend Tim does! (that is the symbol “0x5” formerly known as Tim Abell!)

So, I decided to ask our AI friend to convert the dotnet application to Rust… and it did!

Well, it runs and seems to provide the same output as the dotnet app however, since I don’t really know Rust (yet!), this might give me some insight at least!

Anyway, until the next time!

Well, if I could sell any software, I would be happy to pay for a Jetbrains subscription!

However, if you are like me, you can now use Jetbrains Rider (for all things .NET) and Webstorm (for all things Javascript/frontend) freeeeeeeee!

I can still seem to ramble on over 5 minutes even with something as simple as this!

Anyway, take care. Until the next time…

Well, inovation is continuous for sure. Just look at the amount of Javascript frameworks that exist at any one moment!

So, we have cursor IDE. A new (ish) IDE (integrated development environment) for coding in whatever language you want. It is built on the same platform as VS code. The key selling point is it is an AI first IDE, meaning it has AI code assistance built in from the get go. Cursor has a free plan and also paid plans that allow you more flexibility and different AI models.

Although I haven’t given cursor a good run as yet, the issue for me personally, is it is similar to VS code but not quite. I found myself doing things I would do, or look to do, in VS code but were not in the place I remembered from VS code. This is most probably down to just me, but it shows the bigger issue I guess - we all get stuck using things we know and are comfortable with. I have been trying to use Linux more and more but get frustrated by the squillion ways to do something - but again, I am not giving up - it is good to try to expand your own vision!

So, cursor seems to be a nice proposition, but I feel if I am going to use something like that, I would keep using VS Code. In addition, Codeium is pretty cool! I did a podcast about it a while ago, and it is sort of like GitHub Copilot (mandatory $10 per month) but works with whatever editor you like (they state over 40+ IDE/editors!) as well as being a simple browser extension! The great thing is with Codeium, it has a very generous free plan and on the paid version (pro version is the same as GitHub Copilot at $10 per month), you can switch between AI models. Codeium is a nice modular way to get a similar experience across a wide variety of IDE/Editors instead of putting all your focus into just one thing like cursor. Again, not putting cursor down as I haven’t tried it for long enough, but there we go!

So, take care until the next exciting episode ;)

Well, what to say… 100 episodes! yes, polluting the digital airwaves for far too long!

Taking a quick retrospective look back - he says, as if 100 episodes is an achievement!

Ulysees the Unicorn will be happy to welcome you :)

(You can subscribe free by the way!)

Well, if you have ever heard of NodeJS then this might be of interest to you :)

NodeJS has been around for quite some time, and used by a lot of large companies - but it did have some issues (not to mention the whole NPM world!).

The creator of NodeJS went away a few years ago to create DENO (yes, switch those letters around!). DENO aimed to fix some of the failings of NodeJS but didn’t support a lot of the existing NodeJS ecosystem.

DENO 2 does let you use the older NodeJS modules as well as it’s own newer modules which are, apparently, a lot safer.

Also, I didn’t quite do my research - DENO 2 is built with RUST :)

Deno is written in Rust and based on the V8 engine that powers Chromium browsers. "Rust has been crucial to Deno’s development, especially with Rusty V8, recently stabilized," Dahl told The Register. "Rust's unified build system simplifies adding third-party functionality compared to C++, keeping Deno fast and reliable for server-side apps."

Anyway, take a listen and here are a few links if you need them.

https://deno.com/

https://fresh.deno.dev/

DENO 2 Serverless hosting

Okay, take care until the next time!

Well now, are you using Docker? never heard of Docker? well perhaps you might want to ask AI to help there.

If you do use Docker to create containers of your apps/micro-services then what about Fly.io?

Fly.io is a public cloud infrastructure platform. Users deploy applications as Docker images, but Fly.io does not utilize Docker to run the applications.

Docker images are actually OCI (Open Container Initiative) images, which are essentially stacks of tarballs (are these similar to harballs that a cat has?). Fly.io unpacks these layers onto virtual machines and can retrieve the container manifest from the Docker registry.

Fly.io employs a custom registry to fetch and unpack image layers. It utilizes Firecracker microVMs for virtualization. Firecracker is a virtual machine manager that creates lightweight microVMs. Fly.io uses Firecracker to orchestrate these VMs, along with containerD for isolation. ContainerD is responsible for managing and unpacking application images, with each app instance running on its own isolated kernel.

There are some generous free options described here.