Nit: this makes it sounds like the BEAM was designed for web dev, which it was not. Erlang came out of Ericsson and was built for telecoms (OTP stands for Open Telecom Platform), which is where its unique set of trade-offs comes from. Many of those trade-offs make a ton of sense for web but that's not because it was designed for web, it's because there's overlap between the fields.

One way to see the difference between telecoms and web is to ask yourself when was the last time that you were working on a project with an availability SLA of 9 nines (1/3 of a second of downtime per year) or even 6 nines (32s per year). Some web dev has that, but most doesn't come close, and if that's not you then Erlang wasn't built for you (though you may still find you like it for other reasons!).

Whatsapp and telecoms have a lot in common, so no one questions that they benefited a ton from the BEAM.

Airbnb, though? The main similarity is that they both send large quantities of signal over wires.

Again, none of this is to stop you from liking the BEAM, but when we're talking about professional software engineering it pays to be explicit about what the design constraints were for the products that you're using so that you can make sure that your own design constraints are not in conflict with theirs.

since BEAM gives you smart disconnection handling, web stuff built in elixir gives you the ability to build on client-server distributed without too much headache and with good defaults.

but look, if you want a concrete example of why this sucks. how much do you hate it when you push changes to your PR on github and the CI checks on browser tab are still not updated with the new CI that has been triggered? you've got to refresh first.

if they had built github in elixir instead of ruby would almost certainly have this sync isdur solved. in maybe two or three lines of code.

I'm not cautioning against making the calculated decision that realtime is a core requirement and choosing the BEAM accordingly. I'm cautioning against positioning the BEAM as being designed for web use cases in general, which it's not.

Many projects, including GitHub, do not need that kind of immediate reactivity and would not have benefited enough from the BEAM to be worth the trade-offs involved. A single example of a UX flow that could be made slightly better by rearchitecting for realtime isn't sufficient reason to justify an entirely different architecture. Engineering is about trade-offs, and too often in our field we fall for "when all you have is a hammer". Realtime architectures are one tool in a toolbox, and they aren't even the most frequently needed tool.

what price? learning a new language that is designed to be learned from the one you already know with fewer footguns? ok fine.

but you make it seem like going to elixir is some kind of heavy lift or requires a devops team or something. the lift is low: for example i run a bespoke elixir app in my home on my local network for co2 monitoring.

and for that purpose (maybe 300 lines of code? yes, i do want reactivity. wrangling longpoll for that does not sound fun to me.

* A much smaller ecosystem of libraries to draw from.

* Much weaker editor tooling than with more established languages.

* An entirely different paradigm for deployments, monitoring, and everything else that falls under "operations" that may be incompatible with the existing infrastructure in the organization.

* When something does go wrong, using a weird stack means you have less institutional knowledge to lean on and fewer resources from people who've been doing the same thing as you.

* A whole new set of foot guns to dodge and UX problems to solve related to what happens when someone's connection is poor. This has come up repeatedly in discussions of Phoenix LiveView—what you get in reactivity comes at the expense of having to work harder to engineer for spotty connections than you would with a request/response model.

* More difficulty hiring people, and an increased tendency when hiring for selecting people who are really just obsessed with a particular tool and unwilling to see when the situation calls for something else.

There are many more, these are just the ones I can think of without having a concrete application with concrete requirements to analyze. In the end for most apps reactivity is so much a "nice to have" that it's hardly worth sacrificing the stability and predictability of the established option for moderately better support for that one aspect of UX, especially given that you can always add reactivity later if you need to at a slightly higher cost than it would have come at with Erlang.

If reactivity is a core requirement, that's a different story. If it's polish, don't choose your architecture around it.

It’s a bit like true believers switching faiths when their leader changes religions.

It’s hard to make sense of it from the outside looking in, but it’s definitely a thing that happened historically and occurs in small and big ways even today.

what is also a thing is stockholm syndrome and sunk cost fallacy.

[0] ok, now that i think about it also this was when ruby was having serious 1.84-2.0 transition troubles and i had difficulty reinstalling ruby if i needed to redo the os because some python wheel broke everything, so i had other historical reasons to leave ruby with a bad taste in my mouth. i think my gripes about activerecord are real, though.

Is this actually true though? I’d be interested if you know any data backing that perspective. I only know what I’ve worked on and my anecdotal experience doesn’t match with this statement. But I know my sphere doesn’t represent the whole. In terms of state, by now there are many ways of dealing with persistence and reconnection. Not only are most of those problems solved with existing technologies and protocols but they’re everywhere in web dev. Maybe we’re talking past each other? Did I misunderstand your point?

It's definitely better but I can definitely see why you'd still choose rails these days.

Given how rarely this comes up it feels like a tolerable problem that will only diminish as Elixir adoption continues to increase; I am aware of many rail shops that are slowly and quietly switching everything to Elixir, and it feels like that snowball continues to pick up pace as Elixir improves and those libraries are created.

In the beginning when Ruby on Rails said hello to me, I instantly fell in love with it and the simplicity and the natural semantics that flow with it. It was absurdly easy to write new features and ship them to production. As the codebase grew and the team grew we started running into situations where APIs broke, or to trace the workflow of things in terms of finding where methods came from, finding parent modules of modules, and finding their parents, configuration, and I started to note a general lack of IDE autocomplete and type-safety.

Then after a few years I jumped ship to Elixir and if felt like a breath of clean air when I had to learn FP. Everything was simple enough to understand. Performance knocks Elixir, Node, Python and any other interpreted stack out of the water. The Phoenix framework was, and is said to be thoughtfully designed and although there was no IDE support, we still had Elixir LS which was great enough to provide realtime guidance, linting and type safety during compile time. I was able to ship a very large app into production and it was bulletproof. The problem with Elixir was our other engineers struggled to shift away from Node, or any other stacks they already knew. They found the entire FP world to be weird. Hell, I found it weird too at times. Simple mutations of maps and arrays, that would be trivial in Ruby ended up being so complex in Elixir. In the end it felt like my team was not on the same page. I guess Elixir would be great if you ran a 3-person team or something, but since we were not, we got back to Ruby. In today's world though, I am largely looking at Go, for a backend system. IDE support is up there with Java, and the ecosystem is old and mature enough to find any package that you look for. Performance is C-like and learning curve is lean.

Just my 2c with all these platforms.

At this point, I putting together teams and getting new developers into Ruby on Rails. I'm also seeing companies move back to full stack RoR after the luster of React has worn off. Also, modern RoR can get you so far now with a fraction of the dual framework headaches of a RoR backend/JS frontend.

Any MVC framework with HTMX, JQuery (yes)/Hotwire/Stimulus/Turbo, etc. puts the productivity and deployment speed of the front-end setup above to extreme shame.

Try to create a way for people to upload documents like images and PDFs and documents. Okay easy enough on both platforms and I want you to generate a preview for each of those files so that people can easily find those files. Now I want you to add pagination. Now I want you to add column sorting so that people can sort by file size or by name or by upload date. Finally I want you to add a search field. Going by the way all of this stuff needs to live in the URL so that you can bookmark all the different you know choices you've made.

The stuff is pretty trivial and rails but in elixir you would have to bake it all yourself very boring code that doesn't really matter. This is why I chose to build my startups admin dashboard in rails despite our main production API being an elixir.

If you need to hire someone you'd need to train them on your system no matter what, with a framework you can use their documentation to explain where things are and how they work.

Ash does look badass though!

Elixir and Ruby really aren't that similar anyway. The syntax differences are very superficial - Elixir's a functional language with very style and semantics to Ruby, and that's even before you get into the magic of OTP and the BEAM, for which Ruby has nothing comparable.

Add in problems finding developers skilled in Elixir and Phoenix and the small available libraries.

Of course, you also have that to some degree in Rails but it is much less pronounced.

Is this actually a problem you see? I'm going on 15 years in the industry and haven't seen any issues training people up on a new language in just a couple months.

If you need an expert in some library or language to make meaningful business progress I feel like that says more about whatever tool or language you're using, and I simply don't see that with phoenix or elixir in the years I've worked with it.

I think that this is one of the reasons networking is becoming more and more important, because it lets a candidate demonstrate their generally-applicable development skills to a fellow engineer who is capable of making qualitative engineering judgements.

Some years ago the largest company using Elixir in the US, or at least on the west coast, abandoned Elixir because they couldn't find enough developers.

Yes. The adoption is poor despite the loud voices.

This issue might be partly due to the project being in a somewhat niche and conservative industry, so there are no startup vibes. However, since they started looking for someone ready to make a longer commitment than a developer who has just started their career, things have improved. But this approach also limits the pool of available developers.

It's worth noting that we also use Elixir in this project (the chief architect is quite fanboyish about it), but we have never had any new developers come in with pre-existing knowledge of Elixir.

Since the last 12-15 months, every Elixir job posted gets literal hundreds of applications. Just saying.

I like Ruby, and I feel it has significantly prettier syntax, to me, than Elixir. So that’s a big reason why I also like rails.

also activerecord doing "trust me bro" things behind the scenes (like pluralization) drove me up the wall.

to be fair ecto does a small bit of this too, but at least it doesn't change spellings (so you can global search an identifier).

In terms of successors there’s maybe Julia, or otherwise you’d have to use Python or Matlab/Octave, with all that going to a scripting language entails. In any case it doesn’t really feel like there’s been a replacement.

I did maybe 5 years of Phoenix for a customer of mine and went back to Rails for another customer. It's good enough and overall Rails is easier to deploy IMHO. Capistrano vs I don't remember what.

In that years long Phoenix project one of the developers on the team added dialyzer type annotations to the functions in the files he worked on. Everybody else did not bother. The project ended up with no type checking. The service run and the company did well.

Overall using Phoenix was a good experience. I never used Elixir in any other project and never for my own programs. I use several other languages for my own little scripts, mainly bash, Ruby, Python and Lua. I think that I really like dynamic typing.

https://www.techempower.com/benchmarks/#section=data-r23&f=z...

Ranging from 1.5-3.5x faster.

I thought to include other implementation from Ruby and Elixir. Rails has always been doing much more and on the heavier side of things. There are also many test that simply by switching server to iodine brings the performance to Elixir / Phoenix level.

IMO TechEmpower lost all credibility long time ago after it was demonstrated they do nothing against heavily gamed benchmarks where people literally do basic string matching against regexes instead of doing proper HTTP parsing. Some even relied on characters being at exact positions.

Add to this how slow they were to adopt normal production application code changes to Elixir apps where it was proven that the author of the benchmark had no idea how to code an Elixir/Phoenix app and yeah, it does not look good for TechEmpower.

All that being said, use what you like (some even expressed the confusing stance of "I like Ruby's syntax more" which, need I even comment how unprofessional that is?). But to claim Elixir is just some mere 2x times faster than Ruby is misguided. My real production experience from the last 10 years says this is bull.

I simply cannot respect that.

Having a preference though? Yeah, absolutely fine. We all do.