I’ve been here for a while but since my introduction post, I haven’t really written anything in Elixir.
- I’ve started with Exercisms as a way of learning Elixir. It was fine, though I admit, I’ve spent like two hours simply learning how to properly iterate of string characters and general string manipulation(it’s also my first functional language). I feel like this is unnecessarily complex in Elixir, but after a while I do, kinda, get it… I think.
- For last ~year, I have for some reason stopped dabbling with Elixir. I feel like I was too burnt out to code in my own time. Just recently I got back to it after realization that I do not want to be stuck in .NET Enterprise’y world for the rest of my live. I like my current workplace, I don’t like .NET, Microsoft and their approach with it.
- For that whole time I was still listening to some podcasts(Elixir Wizards, Thinking Elixir) which kept me interested, but I guess not enough to pick it up again.
- Few days ago I’ve finally started working with Phoenix, currently learning basics of the framework, I plan on finally writing a personal project in it, to also then learn proper deployment techniques, nginx etc. .NET world spoiled me in a way where I, as a pretty experienced developer, actually feel almost disabled, that I don’t know how to properly work with Docker, SSL certs etc. In Microsoft world it’s all usually done out of the box(or rather out of your wallet).
- In general - I really like the language, Phoenix/Ecto seem nice as well, I do feel like it widened my horizons as a developer(but I guess same could be said when learning any FP language after using OO only).
- It seems that community in Elixir world is actually way better than in most other stacks(especially since .NET Community is one of the things I hate about it) which is really nice.
My ‘issues’ with Elixir:
- Would love to see it gain some more traction in Europe, feels like there should be enough remote talent to pick from but I barely see any offers.
- Techempower benchmarks. I know this has been already talked about multitude of times here, but it seems like everyone just dropped the subject a year ago. I do understand, that Techempower benchmarks in no way reflect real life perfomance and rightfully so, not everything should be optimized for them, but I do feel like they show Elixir in a bad spot. I cannot fathom how something that is this scalable, can be so low in those benchmarks. I know that all of those examples are ridiculously optimized, to the point of hardcoding static dates and headers(which we cannot really do here), sure, but in my opinion, even if most developers ‘know better’, we should still strive to be better in benchmarks, even if implementations are not realistic. Most people see those results and they don’t think about the fact that no sane person in the world would write code the way it is there. In my opinion it’s a good way of promoting technology/frameworks and to me it feels like minor things like that hurt adoption, especially since that particular benchmark seems to be the ‘go to’ in these matter these days. I wish that at some point in my journey here, I’d be able to contribute something that would help with that.
