🎶 Who wants to live forever? - sang Freddie. Well… #science just said: not so fast. A @PNASNews study finds that life expectancy gains are slowing down across high-income countries. For generations born after 1939, the pace of improvement has dropped by nearly half:

 https://doi.org/10.1073/pnas.2519179122

The reason? We’ve already reaped most of the benefits from reducing child mortality. So it’s no longer about living forever, but about living wisely.

#longevity #lfe #demography #mortality #lifespan

🚀 Ausschreibung: Beratung für ein Open-Source-Informationsdashboard! 🚀

Das #LFE sucht Partner:innen für die Konzeption einer Plattform, die #Walddaten in Brandenburg visualisiert.

Leistungen:
🔹 Layout-, Redaktions- & Veröffentlichungskonzept
🔹 Technisches Umsetzungskonzept
🔹 Workshops & Stakeholderbeteiligung
🔹 Beratung zu #OpenSource & #OpenData

📅 Okt–Dez 2025
📍 Brandenburg / hybrid

👉 Details: https://lnkd.in/dGqgdCTB

#OpenSource #OpenData #DigitaleSouveränität #Forst #Ausschreibung

So, a more obscure language:

I have a JSON object representing a “workflow” – a self-similar tree structure of nodes, each with a collection of predicates, a collection of actions, and a collection of child nodes. The predicates and the actions are not exactly functions; rather, they’re descriptions functions. If you squint your eyes and tilt your head, it’s a DSL. Ultimately, you apply the workflow to a state, generating a new state: w(state, workflow) -> state’.

I’ve been working this in #elixir, essentially running a workflow interpreter. Then a wild hair: homoiconicity. The JSON tree structure can be expressed in a lisp-y way fairly easily, and through some transform of sorts, that can be ultimately be expressed as a function of state – a closure, I suppose. The data (JSON object) with some transform is code.

I’m out on the edge of what I know. I can probably turn the interpreter work into a single closure using just elixir, but doesn’t it sound like a case for #lisp ?

… and what lisp runs on the BEAM? (Well, beside Clojerl :-p ) #LFE.

Do I know LFE? Absolutely not. Should that stop me? LOL…

Enter Claude Sonnet 4.0.

… and exit Claude Sonnet 4.0.

I don’t know how many ai slot machine credits I spent on the task “load a json object and tell me some stuff about it.”

Absolute torture – from getting set up to write some code to final scripts and beam code to run in lfe repl. Torture.

At least I have a sample to study… because, am I going to give up on the idea?

LOL.

Here's all the BEAM language tracks on Exercism:

#LFE: https://exercism.org/tracks/LFE
#Erlang: https://exercism.org/tracks/erlang
#Elixir: https://exercism.org/tracks/elixir
#Gleam: https://exercism.org/tracks/gleam
[Disclaimer: I never did any of those but they look like fun if you plan to learn a new language]