Stack Overflow Developer Survey 2020

My take on things appearing outdated is that actually with BEAM, things just work after they reach certain maturity. Resources without new updates maybe just don’t need the updates.

I published an Ecto PostgreSQL Sharding library for Elixir in 2016 and it hasn’t been updated in years for this very same reason, it still works without issues.

1 Like

That’s what I said - it’s mature, so it doesn’t need updates. But the unfortunate side effect is it looks abandoned…

Maybe you could drop a note to readme every couple of months saying “tested with Postgres 10, everything OK” :slight_smile:

5 Likes

i have never done java but i have heard its a very stable and mature language. So are its libraries left untouched just like the elixir libs?

1 Like

I think this is just what I’m getting at.

You have this feeling. Based on my recent experience, I’m surprised that you feel this way, but in the end, does it matter how we feel? Shouldn’t we instead define some metrics that we think give reasonable feedback about the state of the language?

If Elixir were in decline, we might expect:

  • Fewer production sites deployed with Elixir, year-on-year.
  • Fewer downloads of Hex packages.
  • Fewer new Hex packages, year-on-year.
  • Fewer downloads of Elixir itself.
  • Fewer downloads of Elixir-related extensions for editors/IDEs.
  • Reduced consumption of products aimed at the Elixir market, like Gigalixir.
  • A trend of reduced attendance at Elixir conferences.
  • Reduced traffic and new account creation on this forum.
  • Reduced interest in Elixir-related media (podcast downloads, blog post traffic, etc.)
  • Reduced purchasing of Elixir books and courses.

I don’t myself have these numbers. But there are people on the forum that can provide quite a lot of them. Personally, I would be very surprised if Elixir was not still growing strong, but until we have the data that sentiment doesn’t mean much. It’s particulary bad if Stack Overflow causes people to have a feeling that is out of alignment with those metrics, and worst of all if the feeling is negative when the metrics are positive. Imagine people walking away from a thriving community, because they heard on the street that it wasn’t thriving. What a shame that would be.

3 Likes

There’s certainly no reduced traffic on the forum - I will post a quick update when I get 5 mins :003:

10 Likes

Startups fail because they didn’t find product market fit. I’m not saying you are but I’d be skeptical of any startup blaming a language for their failure. You could find PMF and then rewrite it in another language if you chosed the wrong techstack.

Hiring is also hard before PMF since things are more likely to fail than not. The best thing compensation wise you have to offer is equity at that moment.

I introduced Elixir at our “startup” about for years ago, five years into the companies existence. Since then we hired many back end and front end developers who have no prior experience with Elixir but have ramped up and contributed to the codebase extensively. I’ll admit that they’re often building on a foundation that was laid by more experienced Elixir developers, but that’s how it goes with any new language or framework in my experience.

Overall the move has given us fantastic reliability, easy scaling, and made things possible that never would have worked in Ruby. The business hasn’t suffered one bit from the transition.

6 Likes

Just a thought.

Why don’t the people in charge start a survey for elixir devs? Maybe the survey could include “What language do you use at work?”, “What other language that you are interestred in?”, etc…

The survey could also post the traffic in this forum, the number of new sign ups and question asked, packages published.

2 Likes

I never said here my startup failed because of the language. Elixir actually turned out to be quite a good (or I’d say AWESOME?) choice and the reason that made the product be more than I could have ever imagined.

I was answering the question that someone asked me: “which were the pains that you faced”? So, I just described the pains we went in the beginning because of choosing an exotic language and ecosystem, in the context of 2015 and 2016. That’s very different than blaming the language or saying that we failed because of it.

Btw, I even have a talk from 2017: To Hell And Back:

Most of those points are not valid anymore :grin: (except the lack of workforce, which still holds). If I ever started from scratch, I would still pick Elixir. More than ever.

12 Likes

it looks abandoned because the creators stops making any effort to reply to the issues or merge the PR. This is so sad.

From my side, what I could do is to record and write some polished tutorials.

  1. about LiveView, and the newly introduced LiveView upload, I was thinking to make a “Build your wetransfer in 15 minutes with Phoenix LiveView”. I always loved wetransfer because it’s simple and shows beautiful pictures, so much that I five years ago I created a similar service with rails. My main issue was concurrency, which I partially solved with s3 direct upload… The second issue was the s3 download bill… but that’s another story.
    What I mean is that, if it can help, I could create a dedicated and polished site with just this super simple tutorial (maybe in collaboration with someone else? :smiley: ) to market the technology and simplicity. I’m not saying that it would change everything, just saying that it maybe could help to market these great tools.
  • second, much larger project, it could be a dedicated course, or a free website, which just targets people who want to pick their first language.

However, if the adoption rate at the end comes from companies’ decisions, and the fact that there aren’t a lot of Elixir developers etc., it risks to become a chicken and egg problem.

I was also thinking, does a massive adoption bring only good things to a community like us?

4 Likes