Most of the software I write nowadays is for business web applications (with a few things going on in the background), and yes for this market I don't think there is really any better choice than Rails.

Although it may not have every new shiny bell and whistle, the ecosystem is mature enough to be stable and to have developed accepted best practices, which means you don't need to worry about whether the tool you use is still going to be supported in 6 months time.

Agree.

But the churn on gems is very painful. To be successful with Rails long term, you must make time to constantly upgrade your framework, for for too often a gem maintainer will just stop supporting older versions of the framework.

And all of a sudden the version you need won't build, or isn't compatible with something else, etc.

So Rails is great. It saves you tons of time and gives you rock solid web apps that talk to databases. As long as you don't fall behind in your gem versions. :)

That's me speaking after maintaining several apps from Rails 1 all the way to 4.2. And probably 5.

The thing about gem churn is somewhat true, BUT that is a problem in any open source eco-system that I've seen so far. And in fact I think Rails fares much better in this area than most others. There are a lot of very high quality gems out there that are actively maintained and play well together. Contrast that to the cesspool that is Wordpress plugins.

My solution for this is to never specify gem versions in my Gemfile and once a month or so do bundle update. Much easier to fix one gem update problem at a time than to try to fix 6 all at once.

The problem is when you won't notice something went wrong with the update.

You will if you use proper TDD.

What's true of all bugs? It passed all the tests

I don't here about people being bug free with TDD ( though significantly less bugful)

Obviously all about tradeoffs, but my impression is that if a third party lib needs a major version update, the time changing the version in your requirements is dwarfed by the time needed to investigate generally

You're going to love http://deppbot.com.

(Disclaimer, some friends of mine made it)

Yes my experience has been the same. It's definitely better now (i.e. maintainers seem to be caring about backwards and forward compatibility), but even 5 years ago there were a lot of popular gems that have now been abandoned.

It's still better than JavaScript, where whatever popular build tool you use today will be replaced in 6 months :D

Having used Rails seriously since it was originally made available as an open source project, I find myself agreeing with you less and less. It has really failed to stand the test of time, in my opinion.

I agree that it nailed the needs of web applications in 2004. If you are still building web applications like they did then, perhaps it is still the best tool. In the circles I find myself in, there is a push for much more Javascript heavy applications and Rails starts to become a large hinderance more than a help in that environment, when compared to other tools.

We've built a log of great software together over the years, but I can honestly say that I'm not going to rush into using it in future projects.

This is a bizarre comment. Rails failed to stand the test of time because it doesn't play very nicely with monolithic JavaScript frameworks? Are you asserting that such frameworks are the future?

I think they are referring to most UI logic being on the client, not on server. That architecture does not require a monolithic JS framework, nor does Rails offer anything special in that case.

"Doesn't offer anything special" is quite different than "starts to become a large hindrance."

What are you using then and why is it better than Rails? Just curious.

webpack

This paints with an overly broad brush. It almost sounds like you're comparing Rails with the alternative of using no framework at all, but that's not the right comparison. I would probably agree with you if you had admonished people against bothering with any framework that hasn't been influenced by Rails. But most of its best ideas have filtered out into all the other ecosystems, and there are now very credible alternatives for lots of different stacks and some of those alternatives even perform better than Rails for different common use cases.

Exactly. I know Django and Rails... The Django admin is still a huge selling point for me even today. I can get some class of application ready and running in no time with Django. With Rails, with no built in supported admin app it would take me way more time.

This was a good point a few years ago. But since then there's been at least 3 good projects that allow you to easily add admin panel to rails app:

Rails Admin: https://github.com/sferik/rails_admin Active Admin: https://github.com/activeadmin/activeadmin Administrate: https://github.com/thoughtbot/administrate

I know there is many gems but I have still yet to see them used in the wild. My colleagues seems to avoid using them for some reasons and discouraged me to even try them. Maybe the strength of Django is having the admin by default so most Django app (gems) comes with a admin and those admins aggregate together in a coherent admin application.

This is true. The Django admin works out of the box. Getting the activeadmin gem to work for Rails 4? Not so much:

https://stackoverflow.com/questions/16426398/active-admin-in...

Rails has this available via a gem: Active Admin (http://activeadmin.info/).

It would take about as much time as it would take to add Active Admin or Rails Admin to the gemfile.

I submit that you have probably never maintained a huge web application with a team over the course of years with Rails, or you'd probably feel differently. All of those things that seem like a great idea in Rails... aren't.

I sort of agree with you, in that if I was going to start a company I would totally Just Use Rails.

But I think that's just where Rails is in its lifecycle. We all learned a lot from it, and we've started to build alternatives, but nothing has really coalesced yet.

Flash forward 5 years and there will be post-Rails frameworks that have worked their kinks out and are starting to enjoy the network effects that Rails has. At that point Rails will lose a lot of its advantage.

How Ruby on Rails is better than ASP.NET?

Most answers say the opposite:

https://www.quora.com/Has-C-and-ASP-NET-MVC-been-replaced-by...

I agree with a lot of this. My only complaint is the business layer has been lacking. I am using the techniques presented in the Trailblazer book and I am appreciating it.