First podcast since my vacation and parental leave recorded. Should turn into a great episode of BEAM Radio. Don't know the editing timeline right now but fingers crossed that it won't be very long.

If you know a company is looking to hire in Elixir, I'd love to help them. They can schedule a call here or just email me and we can figure something out.

I'm currently helping West Arete a rather special software agency doing custom software in higher education. They are currently looking for a Ruby dev but we expect them to need more Elixir devs again in a bit. They are also looking for a DevOps engineer if the right one shows up. Read the linked description for more details on what kind of company they are. US only for legal reasons. Exceptional perks for a US company.

I'm also helping a startup that's building a SaaS using some very practically minded AI/ML. All Elixir work, EU and US timezones. If that's of interest, reach out.

The next video is near and I'm going to try some humor in it! Let's see how that goes.

Does your tech have competitive advantages?

Nerdy niche frameworks and languages have their upside.

If you pick Java for your product in a Java world you are getting no competitive edge. You are also not at any particular technical back foot. Just even playing field and you just started.

I thought about this just now due to Supabase essentially shipping two well-established Phoenix features as shiny and new user-facing product features. Their blog post even says as much. This is great actually. Their customer is not the Elixir devs of the world typically. They can make a business out of selling the realtime and distributed features of Elixir and Phoenix to ecosystems that lack it.

To be fair that's not all there is to Supabase. It's a cool product and they productize a number of cool open source projects and from what I gather contribute back significantly. I enjoyed their visit on the Changelog podcast.

As an Elixir and Phoenix dev it can be easy to look at it and feel like there is nothing there. There are a lot of PaaS features in the product that are not just Phoenix in shrinkwrap but much of the pitch is the realtime functionality and that is very Phoenix. Supabase has very little competitive advantage if you are already building with Phoenix. Might save you some time but you take on a big dependency. What we should take away from it is that our stack can bring us advantages that are sellable features in their own right.

This doesn't even touch on an in-language advantage like Phoenix LiveView.

I have mixed feelings about Elm in my day-to-day but it is a really hardcore DSL for building webapps. There is certainty in the strictness and it makes certain difficult things manageable. It constrains what can be done and provides some rather unique guarantees for everything that it allows. I think compared to the much looser, less guaranteed and more watered-down combo of React + Typescript that seems to aspire to do the same thing in the end, I bet it offers a competitive advantage.

For OS choices I bet Nix and some BSDs have pretty strong selling-points.

There are databases that can also offer unique features.

Outside of the beaten path are thriving ecosystems, not as large but also not the same. If you put Python, Ruby, PHP and Node.js next to each other for a backend project the actual practical differences are quite few. Node is fastest in single-threaded performance but you better like the JS tooling. Python and Ruby are very equivalent in almost everything. Some people swear by the one or the other. PHP has some runtime differences but overall similar capabilities, abstraction levels and all dynamic. Choosing between Java and .Net for backend dev is likely going to be pretty much the same. Same OOP paradigms, same enterprise feel unless you really fight it. These can all be the right choice but they offer very little that is fundamentally different.

A lot of devs like using Go programs, especially in infrastructure? Well, static binaries are easy to deal with. No deps, no nothing. And Go has good concurrency-primitives and very reasonable performance so people have built a ton of useful infrastructure with it. The Cloud Native world is almost entirely a circle inside the Go ecosystem in the Venn diagram. Go has particular features that stick out and make it a good fit. As does Rust. As does Zig (the pragmatic Rust?).

If you can match up the tech you like with problems that would benefit from the advantages of it and not be hampered much by any disadvantages it has you can really make an outsized impact.

What's your favorite niche thing that just blows your mind? Let me know at lars@underjord.io or on Twitter where I'm @lawik.

Thank you for taking the time to read this. I appreciate it.

- Lars Wikman