Hacker News new | ask | show | jobs
by isityettime 25 days ago
> Ruby is still a great programming language, but it really needs to intensify the effort to get out of the pit-of-decline.

The languages that have supplanted it haven't succeeded by being excellent. If excellence won't do it, what should "Ruby" do?

3 comments

We've gotten to a point now where ultimately there needs to be a very short elevator pitch for any language to gain any traction at all. Anything longer than a sentence, it generally won't go far relatively speaking. It's like a calcification/maturity thing in big sectors of software engineering. You need a VERY good reason to upset an incumbent. "Because we know it/because everyone uses it" is a powerful motivator.
Elevator pitch for the language itself or a library/library ecosystem that lets you do things better than in other languages. Ruby originally blew up because Rails was a way people enjoyed writing backend code, despite the speed issues. But the problem is other languages got good enough at writing back ends Ruby was no longer special there and didn't have anything else to back it up the way Python has such strong control of the ML library ecosystem.
> "Because we know it/because everyone uses it" is a powerful motivator.

It's fundamentally a "we want workers to be interchangeable cogs" motivator; it's a deprofessionalizing move that's against the interests of all programmers per se. Managers can repeat it, but there's no good reason for developers to do anything but resist it where we can.

The other side of this for developers is that the places where "everyone knows X" precludes all non-X options are actually signaling loudly that they don't much value developer autonomy. That's a useful signal worth attending to when applying for jobs.

At this point, we should just appreciate Ruby and move on. In the AI age, other languages are better choices. Ruby is my favorite language, but I build with Go now. Or rather, I guide my minions to build with Go. They write Go better than they would write Ruby (or Python... please die, Python).
Ruby is a great utility language target for LLM-generated tools.
I wonder if Google was the X factor why Python edged out Ruby.