When choosing a language to learn, I consider:
* what a particular industry or project of significance is using
* job availability
* compatibility with my existing toolbox
* the power/fun factor
But that's far from the whole story, because programming languages are relatively fungible and investing in domain knowledge, people skills, architecture and many other things will bring bigger benefits than learning that brand new programming language!
And yet here we are, discussing how Rust will deliver us.
> * job availability
That just boils down to popularity once again. I prefer to pick languages based on suitability for the problem domain. For me, Rust is suitable in a way that C++ is not, due to memory safety among many other reasons.
Job availability is not just about number of jobs, but also about project quality, companies and personal preference. E.g: I don't have an interest in containers, and most Go jobs involve containers, so this is a minus for Go.
This works in the other direction too. Let's say I want to work for Mozilla, I notice they use mostly C++ & Js, so I decide to invest time brushing up my skills in those areas.
I agree with that regarding learning them, but I disagree strongly when it comes to the end result.
I make a lot of my money fixing/replacing PHP projects. As has been pointed out many times before, there are issues with PHP that are guaranteed to cause bugs. The same is true of many languages.
But there is a spectrum! Some compilers/interpreters scream at you and terminate if you contradict yourself. Some of them even recognize security vulnerabilities.
I agree that people skills, domain knowledge, and architecture are hugely important, but great teams are still less effective when they use slow or confusing tools.
You might be the greatest contractor in the world, but if someone asks you to build a skyscraper with plywood and super glue, you're going to end up with something dangerous, fragile, and guaranteed to fall apart as soon as someone tries to use it.
The other narrative is also true. Using newer technologies is the best way to be a part of new projects and avoid working legacy maintenance kind of projects.
So why are you here, debating about Rust, if you already know your time would be better spent investing in domain knowledge, people skills, or architecture?
I am not really discussing Rust per se, as I am not familiar enough with it, however, when I see biased or misleading statements I do sometimes feel obliged to offer a counterpoint.
You'd have a similar problem with Rust, Ruby, Python, Java, Ada, Julia, and probably tons of other languages. It must also be really difficult to differentiate between C, C#, and C++, considering most search algorithms will ignore the # and + characters.
1. https://groups.google.com/d/msg/golang-nuts/TtPzOvhG6bM/-XCr...
Even if it remains a small niche, I think we've seen that Rust is popular enough that using it long-term is fairly safe. Maybe you'll be one of only a few thousand shops who use it -- that's enough. The benefits of writing secure code vastly outweigh to drawbacks of using something unpopular.