Elected

I was elected onto the board of the Erlang Ecosystem Foundation. In many ways it is a small thing. I don't expect I'll be able to move mountains. 3 years where I put my few cents into board meetings about funding stipends and maybe tip the focus of direction in various ways. But fundamentally it is a volunteer organisation and the community does most of the work and that is even more free to follow their joy. I still appreciate the vote of confidence and I do feel it puts me on the hook to help things progress. The EEF board doesn't provide a ton of active direction, the work groups provide direction and express needs. They are the operating bits and beyond them sprawls the ecosystem.

The EEF is the shared forum for the BEAM and is one of the ways we can coordinate efforts around packaging, tooling, security and marketing. The BEAM world is small enough that while Erlang, Elixir and Gleam are all distinct in their ways, various strengths, weaknesses and preferences, we all depend on shared infrastructure. We don't have the muscle to go off alone and get better results.

There is one topic in the EEF and the Elixir community that dwarves everything right now. Not AI.  Adoption. Everyone is worried that the crap markets right now will almost accidentally pressure Elixir off of the map. It should be possible to make it work in our favor. We have a whole bag of case studies where Elixir replaces a gajillion servers that were doing Python or Ruby. So we might have a case. The case must be heard, seen and felt though. So we discuss adoption and outreach.

Elixir gets significant direction from José. Not explicitly, typically, but he pushes a lot of initiatives. He encourages people to do things. He has been very clear that he wants a lot of the innovation in Elixir to happen in community code, not in the language from a central source. That's why it sticks out to me when he underscores something as the current direction of Elixir. The Year of Interop. Or however he put it.

I think that is clever technical direction. And in some ways  Because Elixir needs to both solve problems of a technical nature for adoption. Zach Daniel in an after-dark discussion in Krakow brought up the typically difficult Windows experience and the lack of a ready-to-go React setup. One probably loses us people all the time without us knowing. The other one is an on-ramp we don't really have. Both these could be improved under the umbrella of interop. We also have efforts like pythonx that provide really nice Python interop with Elixir. There are libraries for running Deno (deno_ex, deno_rider), also Bun. I spoke to Chris Nelson about Wasmex just recently, should come out on BEAM Radio eventually. Erlang has had Java interop for ages which Sasa Juric actually referenced using in one of his talks. The Phoenix Sync stuff seems also intended to play well with Tanstack. Lots of stuff touches on this.

Most of the community is technically minded. We can tackle technical problems much easier than we can do outreach, marketing & evangelism. Interop is interesting because it means you now have some encouragement to make Elixir work better with your thing of interest and presumably future conferences may well pick up that as a theme and your interest in Nix, Flutter or Cobol fits in a larger narrative of the community.

Meanwhile on the less technically focused side we are looking at how we can show up more at events that aren't for the already initiated. I'm taking a hard look at what amount of conferences I can actually attend in a year and how I should mix my time.

Neither the EEF or José truly sets the direction of Elixir. Fundamentally the community does. But it is nice to see coordination and strategic ideas on how to move things forward taking hold.

What do you think should be done in Elixir for adoption or interoperability? You can reach me here on email, on the fediverse where I'm @lawik@hachyderm.io, on Bluesky or by responding to this email to lars@underjord.io.

Thank you for reading. I appreciate you.