Ruby On Rails 5.0 Released (rubyonrails.org)
steveb3210 writes: Today, Ruby On Rails released version 5.0.0 of the platform. Major new features include ActionCable which brings support for WebSockets and a slimmed-down API-only mode From the official blog post:After six months of polish, four betas, and two release candidates, Rails 5.0 is finally done! It's taken hundreds of contributors and thousands of commits to get here, but what a destination: Rails 5.0 is without a doubt the best, most complete version of Rails yet. It's incredible that this community is still going so strong after so long. Thanks to everyone who helped get us here. [...] Note: As per our maintenance policy, the release of Rails 5.0 will mean that bug fixes will only apply to 5.0.x, regular security issues to 5.0.x and 4.2.x, and severe security issues also to 5.0.x and 4.2.x (but when 5.1 drops, to 5.1.x, 5.0.x, and 4.2.x). This means 4.1.x and below will essentially be unsupported! Ruby 2.2.2+ is now also the only supported version of Rails 5.0+.
I find it hilarious how so much of Rails' initial popularity because it wasn't riddled with design patterns and the bloat of Java EE. But here we are, a decade later, and Ruby on Rails has somehow managed to surpass Java EE in terms of complexity, patterns, best practices, and all of that!
It's interesting to note that the same thing is happening with Rust. It's supposed to be better than C++, but somehow Rust ends up having a worse syntax, a worse standard library, a worse community, only one kinda-shitty implementation, godawful semantics, and a worse learning curve!
It's also interesting to note that it's the same with NoSQL databases. They're supposedly better than relational databases, but then those who choose to use them end up trying to reimplement basic RDBMS functionality like transactions and foreign keys using Ruby or JavaScript! Instead of using a dedicated and proven query language like SQL, they try to imitate SQL using JavaScript!
It's like Millennials/Hipsters haven't even tried to learn from the decades of experience that we acquired before they showed up. They're so sure that they're right, despite being completely wrong, that they ignore the proper way of doing things and instead do everything wrong.
I still have some hope for the next generation, though. Maybe they will be smart enough to realize how shitty the work of the Millennials/Hipsters has been, and they'll throw it all away and go back to using real, proven programming languages, frameworks and database systems, even if these proven technologies aren't "trendy" or "pretty".
Every few years a bright but not well-grounded-in-history programmer realizes he/she can query a database schema and generate a CRUD application.
One of the earlier examples was Powerbuilder. But there have been many incarnations of the idea over the years. Even Java/Hibernate attempts to provide much of the functionality (yea, I know it's an ORM, but that doesn't make it suck any less).
They all seem like a magic bullet to people who have never worked with one before, but they all share the same fatal flaw: a simple CRUD application is worthless, and once you start to modify the generated code you're locked into that version of the database schema. Maybe they've gotten better over the years, but it will never be as good as writing the application to do what you wanted in the first place.