While I like frameworks like Rails and Django, Golang with AI makes me very productive. Especially because of how Golang codebases don't use too many dependencies so an LLM never recommends code that has an old dependency or mismatched version...
I'm just curious, how well does Ruby integrate with AI tools?
I don't have a lot of experience with Golang and AI, I think Rails can give you different kind of productivity.
From what I know (please correct me if I am wrong) most people use AI to create scaffolding and automate all boring and repetitive tasks in a project. So code still needs to be written, you just outsource it to AI helper.
In Rails you write less code and concentrate on business logic because everything boring like DAL, authorization, caching is already written and tested in production.
Not sure about Ruby in general, but Rails is an excellent candidate for LLM assistance. It’s very much convention over configuration, been around since forever and there isn’t a question somebody hasn’t already asked on Stack Overflow.
This is the thing when people say "LLMs write the boilerplate code for you"--why are you writing all that boilerplate? If you're dutifully cranking out boilerplate instead of creating abstractions that make it unnecessary to write boilerplate, you're a mediocre engineer at best. Add an LLM, and now you're a mediocre engineer with a mediocre simulation of a mediocre engineer.
Aren’t most business apps almost exclusively glorified CRUD? Like the sort of CRUD stuff Rails is amazing at. Those are basically all boilerplate, it’s more of a question of which flavor of boilerplate you need for a given use case.
A singular data point here, but if I ask MS Copilot to write me a unit test in python/django it will more often get it right the first time than in golang. The golang code tends to reference things that don't exist. These aren't huge codebases either.
This is a great point. I often ask LLMs which coding language they are most proficient in, and Python is the most typical response. This has makes me want to write more Python.
Keep in mind, LLMs can't introspect. They integrate your prompt with their training data.
If Python is an extremely popular programming language people like to write about using (it is), you'll expect a lot of people to state that Python is their most proficient language and for that to make its way into the training data. When you ask the LLM, you'll get Python as an answer some fraction of the time and other languages some other fraction.
Those answers are orthogonal to how good the LLM actually is though. In the training data, you'd like a sufficiently robust sample of code (which Python should have) without most of the code being terrible (which Python doesn't have in the wild).
Mind you, IME they do better with Python than very new languages, and they do better with Python tooling than whatever maven build garbage you might have to deal with, but that's faint praise.
It seems to understand the way you want your unit tests written. So if you have a particular style, it's best to write one or two tests in your style, then it will use that same style when it starts writing tests.
You can do prompts like "I need unit tests for the exceptions that are raised in the SuchAndSuch.function()", and it will do it -- particularly if you have a unit test already written similar to what you'd like.
Certain cities it's like a Rails desert. I'd love to get a job using it but it's slim pickings. It's either .NET or PHP and maybe some Python and maybe some Node most of the time for web dev.
That list is great, but it’s not guaranteed to be free of false positives. We Use Rails also offers more granularity—you can see the exact technologies people are using.
I've always been interested in rails, even though my career has gone the JS/TS route. Think i might finally take the plunge and try to build some random stuff with it. See how far i can get by looking at code examples and asking gpt "how do you do x in rails" etc.
Really love that it seems to just be a complete toolbox, especially for a solo project.
Also seeing more and more rails roles out there, but obviously don't have 5+ years with the framework
There was a period of time around the height of Rails original heyday where the idea of a very opinionated, complete framework such as Rails got a lot of blowback, which gave rise to the Javascript ecosystem and its myriad front-end frameworks.
I think the pendulum is kind of swinging back the other way now. While there are certainly advantages with the Node ecosystem, it always felt very messy to me. Rails language extensions, ORM, testing, and migrations frameworks have never been matched by any javascript libraries. Ruby and Rails have always prioritized quality of life features, which made it a very enjoyable and productive framework.
I think the reason rails has stuck around is because of it being a very opinionated framework and it dictates a method to do almost anything. This allows the docs to be very straightforward, almost as if they were written for the Rails application itself.
I also think the language itself is a big plus. I don't like the reflection system or hot patching being a feature in a language, but rails makes extensive use of it to create it's own domain specific language.
The reflection and hot patching features are often called out as negatives, and in general I agree, but for the most part, they're non-factors. All of the "find_x_by_id_and_status()" methods that rely heavily on reflection are all available with a more general, maintainable "find_by()" method which just takes a hash of parameters. And in practice, I'm not sure how much custom hot patching is actually used in production codebases. Overall it's mostly a red herring which detractors have vastly overstated.
Yeah, i mainly deal with react/next + supabase on a daily basis at work.
And do sometimes think of small apps i could build for fun, but the thought of doing react, again, is tiring in and of itself lol plus the whole setup around getting express running and picking packages.
Rails is ready out of the box, for me to not have to screw around with.
Just love the ruby syntax(along with python, which i am actively learning now with datacamp). Its cleaner, and i can probably do 90% of what i do in react there. Then for the rest throw solid js or svelte, or even just react for specific components.
Can only encourage you, I'm mostly working with Go but all my side projects are with Rails which I've only started a few years ago. It's a lot of fun.
I'm not a big fan of watching tutorials, but more of learning by doing but I signed up for https://gorails.com when I started out and it was very helpful to get an overview.
Does anyone here have any recommendations for free or at least very cheap Rails hosting for a toy app just for learning purposes? I’m looking for a PaaS offering since I don’t have the time to secure and manage a VPS.
You might also want to try Kamal, which Rails is already configured for out of the box.
You'll still be spinning up a VPS, but at least set up and management is done via a very heroku-like interface from your local machine. Just get any VPS that will be accessible via SSH, and `kamal setup` can handle the rest. [1]
The one annoying part right now, is it relies on a docker registry you have to provide. I think they're working on removing that requirement, but you can use the free github container registry until then. It's just for rollbacks AFAIK.
Does it really take any more time to secure a VPS than a docker container on a custom runtime for a hobby project? Either way you still have to run software and OS updates.
Happy render customer right here.
We’ve got multiple rails apps running on render.
Renders tech support team have been very helpful wherever needed also.
Should also point out the recently released Rails 8, has as key features focussed on making rails much easier to deploy to anywhere that supports Docker.
Rails 8 made me love Rails again after a few years working with other tools. The new direction is just perfect, specially for solo developers. The 'solid trifecta' in combination with the simplicity of using sqlite3, the new built in auth solution, Kamal... everything, so f*ing good.
I know sqlite3 is great, but I've always found it much more annoying to use it on a server for small side projects than just using Postgres where I could just connect to the prod instance from my local machine and run some queries to look at the data. With sqlite I'd always have to sync files, set up volumes, make sure permissions are correct vs. just setting an environment variable to a PG instance.
From what I understand, the sqlite3 comment was made it context of the Solid libraries. Basically common Redis-backed functionality can now be sqlite3-backed instead.
I do share your preference on using Postgres as the main database. Having that as the main db with sqlite3 used for auxiliary functionality sounds great.
I've been out of rails for more than a decade, ended up mostly with Django, but I always kinda miss rails. But Django has treated me so well, just wish async and python in general had more legs on this side
I've used nothing but Rails since about 2004. Of course I've been integrating and playing with various JS frameworks over the same time period, but the only things that's remained consistent is Rails as the base. I've gone through at least 4 or 5 different JS frameworks in that time and that aspect of development has grown rather tiresome.
Been using Rails professionally for 13 years (currently at GitLab) and I still don't think anything else beats it for developer happiness and productivity.
It cuts through all the mess of modern web development so well and makes turning an idea in to something deployable incredibly easy.
I’ve also been working with Rails for many years now but I disagree I think developer and happiness are subjective and I’ve seen great applications being built by many people in various languages. The key is picking the tool you are most proficient with and one that also aligns with your technical requirements. I certainly wouldn’t build a Discord with Rails..
Hey there's some selection bias going on there! Rails is great for consistency and even better if its your jam. Its stable and high quality and still headed in a great direction. Its a great choice. Especially if you can work in it long term, and pay its learning curve only once.
...but as someone using it in Anger for a few years now, I long for the day I can say goodbye to it and never look back, for all the usual reasons.
I should say that I've not been using Rails _exclusively_ for that period. But it's the one framework that sticks around in my toolkit no matter what. It's what I reach for in any new project by default.
I was also using it in anger until recently, and I'm really glad I got out of it.
But I disagree that it's consistent or that it leads to consistency.
The "Rails way" or "DHH way" of building apps is indeed consistent, but most big teams eschew this in the name of more complicated (and less consistently applied) patterns.
It follows the Anna Karenina principle to a T. All good codebases are alike and follow the Rails way; each bad codebase is bad in its own way and uses bullshit like Trailblazer, Active Interactors, over-engineered Service Objects or some other bullshit. Sometimes all of them at the same time.
The main application I have to work with now ended up, for various reasons, as a Rails API with a Vue client.
The Rails part has been easy to maintain and adapt to users' frequent changes, the main issues being feature requests for very complicated queries which then end up being slow. Vue has taken up a lot of time with the migration from 2 -> 3 (still not finished).
It's such a shame that a big successful Rails or Django equivalent never really materialized in the Node world. It's not like nobody tried. There were so many times it seemed like one was gaining traction, then it would be abandoned or zombified or commercialized, and it just never happened. I really can't understand why.
Indeed Next is the closest we've gotten, but see: commercialized. The default path for Rails and Django is not a vendor locked in cloud offering; they are truly FOSS projects. It seems like its that mentality which died out more than anything else, really. Also "just picking an ORM" is actually one of the key killer features of Django/Rails in general. There is no fully batteries included equivalent for JS land sadly.
Laravel still seems a lot more fleshed out on the back end. Job queues, task scheduling, notifications, migrations, email. Seems to be a lot more of a cohesive package, everything working together.
Cool site! Love seeing what others have built using rails.
I've been working on a membership management platform [1] using Rails + Hotwire Turbo & Stimulus and I can't imagine how long it would have taken me using other frameworks around.
Want a rich text editor? Just use action text. Want document storage? Active storage is easy. Job queues? Mailers? Caching? Integration testing? All easy to do.
I love that Rails is "omakase", once you lean into the ecosystem it's great that most things just... work.
Can someone provide a reason for why anyone should be using Rails? I'm always curious why people love context switching between multiple programming languages.
As opposed to using JavaScript on the front end & back end? That benefit of JS has always seemed a bit overrated to me—the context for front-end & server-side JS is pretty darn different, too.
Anyway, Ruby & Rails are such a joy to use that, at least for me, the fact that it’s written in a different language than the one we need in the browser is a non-issue.
I used to love context switching between frameworks and languages: Rails was more fun and I felt more happier and creative. Django felt more restrictive but I felt more productive. NodeJS was more chaotic but I felt more powerful.
Today I stick with Ruby and Rails. I am trying to do my context switching in the front-end JS frameworks but I just feel a bit dead doing it at the moment!
I've always loved ruby. Sinatra was my jam long ago but after awhile it felt like just remaking Rails when I could just Rails. Then datamapper project went defunct and needing to use ActiveRecord if I didn't want manually do sql seemed even more like I should just use Rails.
What’s the best way to use JavaScript on rails these days?
I’ve been using rails for over 20 years and still love it but the JavaScript story has always been in flux in bad ways. I don’t like the turbo stuff or stimulus. Basically just want to be able to add some nice charts, and some enhancements like confirmation for links. Basically just build admins with it so the caching for partial html just isn’t part of the game. Looked at upgrading to rails 8 now but the javascript in last version of rails was fragmented and complicated, so looking to find a good way to do it in a standard way now, hopefully next version of rail’s doesn’t change everything again.
- Adding npm Packages with importmap-rails
To add new packages to your import map-powered application, run the bin/importmap pin command from your terminal:
$ bin/importmap pin react react-dom
Then, import the package into application.js as usual:
It would be a valid thing if there was a documented way to convert a generic node package to import map. Not much stuff in that importmap ecosystem - seems doa.
What? No, it's just rails that has changed how it handles JS, or at least the recommended way to handle JS very often throughout the year. It's weird to blame it on the JavaScript ecosystem. Like I can't figure out what changes in the JS ecosystem you are referring to exactly.
Rails is probably one of the most complete, turnkey frameworks I’ve ever used for web development. When I first encountered it, within a day or two, I was building full stack crud apps.
However, what shocked me, having come from a C/C++ background were the thousands of dependencies for the most simple of apps. This scared me as how could I possibly understand the security holes and inner workings of all these dependencies.
Another problem I ran into was the tendency of rails developers to create monolithic applications with little separation between the UI, middle tier layer, and the data model layer. When I would be writing client side apps and interfacing with rails back ends, The rails teams would always implement new functionality into the website and then separately add functionality to an API. I think this comes down to the way people learn to develop in rails as rails teaches a monolithic method. I would always try to point out that creating an API where all of the logic was held, would make life so much easier for them, and they would agree, but continue to build The monolithic website with a separate API.
One thing I find frustrating with Django is there's not really a place to put business logic. Ask 3 Django people and you'll get 3 different answers, and often different answers from the same person! That's not a good architecture IMO. It should be obvious where stuff goes. Some say it's not Django's concern, Django is just a web framework, but Django doesn't sit nicely at the edge of an architecture where it belongs at all.
Rails definitely seems optimized for creating code, but how is it at maintaining code? I've never used it, but it seems like Ruby's dynamic types would make it really challenging to do large refactors. What techniques do rails developers use?
The key to maintaining a healthy Rails codebase is to make most of your code independent of Rails. Rails is great at the basics: Handling requests and composing views. It's okay at storage stuff. But it's not good at handling complex usecases, managing JS, etc.
This isn't to say that you can't use Rails if you're making complicated stuff, or using JS, it just means you need to know when to keep your code to Rails's structures and when to write your own structures that you then invoke from within Rails.
Honest question: Where do you deploy? What do you use for queues? Database? It's impossible to beat React for component libraries but, lemme tell you, I've struggled to be productive with Next when compared to Rails.
I don't know if I'd admit that publicly, without first reading about it and trying to understand why things like RoR and Python have huge, loyal user bases.
I'm sure they meant an actual statically typed language. I agree that dynamic languages are fun and productive ... until the codebase becomes big and complex, and then not knowing what shape any data is quickly becomes a nightmare to understand and debug.
This has been my experience as well. Rails can do a lot, but it also allows you to develop junk. Most sites built in Ruby end up in the latter category, especially if built by a team that doesn’t use basic software architecture patterns.
Well actually, if they had used Next.js deployed via Vercel or something similar, it would have been statically cached on a CDN and prevents issues like this. But because they want to use older technology that's more difficult to cache and proxy, this happens.
God, this reminds me of the worst sort of discussions I’ve had with people who don’t understand the technologies they are working with. I’m sorry for being so negative, but “older technology that's more difficult to cache and proxy” is just nonsense.
It's nonsense and worse: hype. I recently went through a bunch of Vercel apps and compared them to some of my favorite Rails/Django apps. The best performing Vercel apps were on par at best.
I don’t measure req/sec here, but at one point, I had 2.8k visitors in an hour. Before rescaling, it was running on Hetzner’s CPX11—2 vCPU cores shared with a few other Rails apps. (I didn’t expect it to go viral at all)
Last month, I built a directory of web apps built with the Rails framework.
With Rails 8 released last year, the framework is stronger than ever. Some even say it’s experiencing a renaissance.
There are already 21 submissions, and if you use Rails, you’re welcome to add your app or website!
reply