Implementing the board game Pandemic as a domain model.
GitHub - chriseyre2000/pandemic: Attempt to model the Pandemic board game in Elixir. · GitHub
This is not yet complete, I am still working through the role actions and other “play at any time” cards.
I created an Elixir library to use and control Toxiproxy in tests.
Toxiproxy is a resilience testing tool from Shopify that allows you to model slow or unresponsive external services.
My main focus was an easy pipe-able and understandable API, here’s an example of it:
ToxiproxyEx.get!(:mysql_master)
|> ToxiproxyEx.toxic(:latency, latency: 1000)
|> ToxiproxyEx.apply!(fn ->
Repo.all(Shop) # this took at least 1s
end)
I am currently finalizing an internal project for my company. A Kanban system exclusively for sales and service management via WhatsApp. I am using Phoenix LiveView for this application 
I’ve started working on a (rudimentary, horrible and probably super-insecure) OpenID Connect strategy for Ueberauth, here: GitHub - braunse/ueberauth_oidc: Ueberauth Strategy for generic OpenID Connect · GitHub
I’m currently reading and learning the Programming Phoenix book, a really nice intro to web programming using Elixir and Phoenix 
Dusting off all my Elixir books and tutorials and wanting to do some fun learning side-projects.
I have an over-ambitious side-project that I want to use Elixir / LiveView as the fundamental piece. But it has some very dynamic aspects that make me a little nervous. 
Meanwhile, I keep rewatching https://www.youtube.com/watch?v=JvBT4XBdoUE (GOTO 2019 • The Soul of Erlang and Elixir • Saša Jurić) to get inspiration, evening running the demo locally so I can play with it. Love this video.
Recently, I built a service that tracks (in almost “real-time”) all posts and comments on Reddit and Hacker News. I knew that Elixir (and the BEAM) is much more efficient that Ruby (my primary lang), and that’s one of the reason I decided to use Elixir for that problem. However, I was surprised how efficient Elixir is.
To give some rough context, all the traffic mentioned above (tens of millions of records per day) are processed on a $5 VPS that is loaded not more than 5-6% on average (+ I’m hosting another Rails app on the same VPS). It’s just amazing how performant and efficient it is.
On the whole, It’s a standard Phoenix app, Broadway for handling all the processing and rate-limiting and some LiveView for the demo - Apple, Google, Microsoft, Amazon, Facebook Live Monitoring | Alertcamp.
The only custom thingy is a small GenServer implementing a simplified version of the Aho-Corassick algorithm for tracking keyword matches.
Please feel free to register and try it out by monitoring your product’s brand or competitors. You can receive email notices every time someone mentions your keywords.