Topics come off the same queue as Bruce's essays: the throw-it-away discipline in a real PR, prompt layering as software design.
Already in the subscriber archive: an alternative to stateful LiveComponents and the agent skill to match, red/green indicators for AI-generated code, how training changed in 2026.
Session 6 is free to attend. The topic goes out with the join link.
Mike Hostetler built a Slack clone with the Jido ecosystem: channels, DMs, threads, reactions. The interesting part is agents as regular participants in a room, using the same messaging and persistence paths as humans. The UI is Hologram, so it's Elixir down to the client.
When a process fires off several async requests, how does it know which reply belongs to which request? A PID tells you who, not which. Krishnadas walks through correlating them with make_ref/0 and a selective receive.
GenStage's demand is the part that trips people up, especially max_demand and min_demand. Andrea Leopardi explained it in detail and put together a visual explainer with interactive playgrounds where you can move the producer rate around and watch the queue fill.
Until now, using Tidewave meant working in the Tidewave app. Today there's a second way: using Tidewave directly with the agent you already drive from your terminal or editor. Meet Tidewave Toolbar and Tidewave Connect.
Rodolfo Carvalho dug into Hex download data to see how fast we upgrade Phoenix. Patch releases spread in two weeks, minor versions take seven to eight months to cross 50%.
Johanna Larsson makes the case that libraries shouldn't start their own supervision tree. Once they do, you can't put them under your own tree, pick a supervision strategy, or run two instances with different config.
How does a team of three engineers process two billion records a day? In this talk, you will learn how Tolemi rewrote their data platform in Elixir to achieve massive scale while keeping complexity surprisingly low.
Fuel the digital engine behind RevZilla, Cycle Gear, and J&P Cycles. We’re hiring a US-based Elixir/Phoenix engineer with a strong frontend focus to build high-octane eCommerce experiences for riders worldwide.
Note: Role requires full-time physical residence within the US; we cannot consider candidates living abroad (including dual citizens residing outside the US).
