Published an article about Composite – a small utility for writing dynamic queries.
fuelen
karolina.curiosum
How integrating Alpine.js with LiveView can significantly enhance the user experience in your applications.
Alpine.js with Phoenix LiveView: User Experience Guide | Curiosum
ChristianAlexander
The Elixir ecosystem has very strong abstractions. One of these is Telemetry, which decouples instrumentation from an app’s main code.
Telemetry is synchronous by default, but with a little GenServer magic it’s possible to collect analytics without a major performance hit.
sorentwo
Oban Core Team
Tips and techniques to make Oban job errors easier to identify, differentiate, and diagnose in monitoring tools like Sentry.
imkleats
After some recent experimental work with LiveView (related to what I discussed with you, @rcm765), I was inspired to write an article about distributed systems design. In doing so, I found a critical flaw in modern browser architecture that, when fixed, is sufficient to serve as a universal interface for self-executing code for AGI.
https://medium.com/@imkleats/fear-and-loathing-in-dom-vegas-abea2a21774e
If you enjoy this article, I’ve written a follow-up article. That article summarizes the findings of a thought experiment in which I asked this AGI to “maximize people’s enjoyment of apples, while minimizing wasted apples” as its kernel code, and then thought of which questions it would ask itself in response.
As I built out its physics engine using the distributed systems architecture, I found the AGI would likely wonder what this “second” thing is because time is very much important to apple production.
https://medium.com/@imkleats/a-computers-theory-of-time-84c13dceb1a5
If you can find a flaw in either of these, please let me know. I am not a physicist, and I am very startled by the conclusions my thought experiment has reached. I would appreciate someone, anyone being able to give me a sanity check.
(edit: If you can’t find an obvious flaw, please share. I don’t mind being roasted if it’s hot garbage, but the more eyes on it will lead me to knowing sooner rather than later.)
woojiahao
I recently gave a talk in school, introducing the fundamentals of functional programming and an introduction to Elixir with applications with Phoenix.
Wrote up a comprehensive guide for it that I would love to share as a learning resource for those interested in learning more about Elixir and Phoenix: https://practical-elixir.woojiahao.com/
I also have the accompanying talk on YouTube: https://youtu.be/-VHBMXxCee0?si=us88VL8HFl75fokp
Do let me know your thoughts on these as I would love to continue expanding on these resources for future uses 
I would greatly appreciate any feedback!
szsoppa
djaouen
I wrote an article on what I’m calling “The BEAM Spectrum”. Enjoy!
imkleats
I buried the lede about my first article, but I can’t figure out how to edit it again. It contains a complete solution to Fielding’s Code-on-Demand constraint for REST architecture that is arrived at via inductive proof. It’s written in gonzo style because I believe that the experience of the investigator is as essential to understanding their conclusions as the methodology itself, and we do ourselves a disservice when we are lazily spoon-fed “facts”.
The argument is actually highly structured, but you have to go along for the ride. It lays out definitions, satisfies those definitions, and then uses the newly-available properties of the thing defined to continue the inquiry. If you disagree with my initial definition, you should relax a bit because you’re probably trying to impose constraints that don’t matter; when placed in a separate layer, it becomes the rigid concept you might have had in mind.
The article is written from a front-end perspective, but when I say “DOM” in the article, you can read that to mean “any possible representation of a hypermedia resource” and “hypermedia” is in the eye of the beholder. This architecture could broadly solve network-related and distributed systems-related issues if it were used on the backend, and I imagine it could also “scale down” to orchestrate at the GPU level for parallelizable computation-intensive tasks.
Enabling CRUD on the hypermedia representation in the way proposed, while supporting the “ephemeral” networking between hypermedia representation nodes, may actually constitute a wholly new architecture that happens to satisfy REST, thus getting all the advantages of REST while being supremely flexible - the abstraction is the abstraction at that point.