Rails 5.0 release topic on HN 50% are comments about Elixir/Phoenix

So seems like Elixir/Phoenix are really gaining mindshare :grin:


I tried RoR about two years ago and it just didn’t really do it for me. I didn’t “get it” like I have with Elixir and Phoenix. Thus I’m not surprised that the conversation is going in this direction.


Is this the discussion you’re referring to, @andre1sk?


The Rails 5 post on HN is this: https://news.ycombinator.com/item?id=12012048


Wow, that thread is pretty funny to watch as the Rails folks end up “defending” their tool rather than just accepting the idea that it’s in many ways the Elixir/Phoenix ecosystem might just be a more efficient environment.

"Leave my tool alone. And now get off my lawn. "

1 Like

Things are looking good for new green field projects to choose Elixir/Phoenix for their stack.

I’ve been a heavy user of Ruby and Rails since 2007. I feel Elxir/Phoenix is in a good spot to adopt for startups or new projects right now (whereas I thought it was a bit early a year ago). I think there is an exodus from Ruby/Rails into Elixir/Phoenix, but I don’t feel like that’s the main thing mopping up Ruby/Rails. Instead I think it’s Node.js cutting into Ruby/Rails and nearly every other dynamic language ecosystem used for web/api development.

I live in Los Angeles, and as far as startups go over the last two years, it seems a vast majority have opted for Node. I know of a fair number that picked Go, two that picked Ruby/Rails and one that has picked Elixir/Phoenix. Yes, anecdotal, but it’s the dynamic as I see it from my part of the world.

At any rate I hope Elixir/Phoenix can thrive under the Node developer tractor beam and I hope Ruby/Rails continues to thrive as well. Choice is good and it’s always fun to have 15 different languages running systems at your company forcing dozens of devops jobs :wink:


It might be a bit of a tangent here, but why is Node.js now considered to be more useful by many people than before? I have used it in the past, and I found it not very pleasant to use JavaScript for server-side things, but that is of course my very personal opinion from a few years back. Has anything changed?

Majority of work we do is Node, but I am really hoping Elixir/Phoenix will carve out a decent size niche . If we judge by HN sentiment it will happen fairly soon :slight_smile:

Go is nice but working with DB in Go is a f$%^& pain. Obviously was designed by people that do not use RDBMS (dah google).

1 Like

It’s considered useful because of the proliferation of SPAs
with node you can share validation logic
have server side rendering (ember fastboot, Angular Universial etc)
you can also share logic between Mobile app and SPA


While I understand and appreciate the enthusiasm of people promoting Elixir, I’m not really sure that attacking other ecosystems (especially in such an amassed way) is the right way to attract more people to Elixir.

We shouldn’t be saying to people they are simply wrong. We should explain what use cases Elixir fits and what it does better and why. Saying “you’re wrong and I’m right” without giving any context is not the right way to convince people about Elixir’s awesomeness.

I know quite a few people that were heavily discouraged from giving Elixir a try exactly because of threads like this.


Eh, those happen in every community and ‘new thing’ though. Seen it way too many times over the past 30+ years.

Speaking of, I’ve been using Erlang for around ten years now off and on various projects, when I was reading that thread it just hit me that Erlang is actually popular now, this is both exciting and horrifying at the same time. I loved the tiny community around Erlang, friendly people, and those tend to get disillusioned when they get huge (here is hoping it does not happen), and yet not having to write my own libraries for every bit of extra functionality I need is going to be very nice. Elixir brought Erlang into the spotlight. :slight_smile:


Majority of comments are from former Ruby/RoR guys that are learning/using Elixir/Phoenix so to a significant degree it’s internal thing. One of the longest comment threads there was sparked by Brian Cardarella’s comment (DockYard used to be 100% Ruby/RoR shop).


Hmm, that makes sense… They learned something new and are trying to show it to their old colleagues (in better or worse ways).

Ruby/Rails is actually one of the few ‘often-used’ languages/frameworks that I never got around to learning. I wonder if any Rails5 people would be interested in making a git repository of making the same mini-servers (to potentially not ‘mini’) of comparisons of Rails and Phoenix projects built for various purposes (readability in some cases, speed in other, real-time websockets, etc… etc…). I tend to learn best by comparison and I bet others do too.

1 Like

That’s classic neophytic attitude, nothing to be happy or ashamed of.

People will ger more realistic once they have to maintain “legacy Elixir apps” and realize the language/tools is important but will not solve all your problems.

For now, it’s a honeymoon.


I’d agree with the sentiment that attacking another ecosystem or tool is a horrible idea. There are technologies I absolutely refuse to play with (cycle.js is a perfect example) because the community/authors attack everything else and say “My way is the only way because I think it’s better!”

I think a much better way to drive interest and adoption is for this amazing community to keep blogging, making videos, and building more and more compelling solutions to existing problems. Let the ecosystem stand apart on its’ strengths in a vacuum rather than making direct comparisons and risk starting a religious framework war. People testing it out will be much more open and can form their own conclusions. If you look for the bad you’ll always find it.


I would tend to agree but given it’s internal to RoR e.g. it’s former/current RoR developers having this “discussion” it doesn’t look too bad. To be honest you can not drive interest without making comparisons to existing solutions.

1 Like

I think some of it is driven by people only having being exposed to scripting languages prior to trying Elixir/Phoenix. They often had to resort to running node along side to handle websockets also if they hit performance issues the only viable option would have being to develop extension/module in C. Elixir/Phoenix let’s them build the solution using single lang./platform so it’s really exciting for them.

not sure how inserting your opinions of one technology in a thread about another technologies’ release announcement is considered “really gaining mindshare”. Seems pretty rude to me, actually.

Why does the success of one thing have to be at the expense of something else? Just because that’s how it’s done on HN/Reddit doesn’t make it right.


I’ve been advocating Elixir for 3 years among coworkers (doing Rails and Clojure) and have been the butt of jokes to prove it, but this is not the way to do it. I’m kind of disappointed in our community, to be perfectly honest. Let Rails-ists enjoy their Rails 5 release thread. Offer comparative benefits/drawbacks in other places. I have a number of connections in the Rails community and some of them are turned off to Elixir and Phoenix from this sort of fan-ism.

Yes, I take glee in how DHH has admitted to modeling Action Cable after Phoenix Channels, and I know my history enough to realize @chrismccord tried doing this sort of thing in Ruby and the Ruby interpreter just couldn’t offer the level of reliability he needed, so he decided to build Phoenix. Still, we can be happy for a new, improved release of Rails and leave it at that. Offer other blogs, lunch and learns, user group talks, etc. to do any comparison between the two environments later. Please.