I don't think node/ember will ever go away, but I'm sure that they'll lose their fashionability after a year or two much like rails has, and for that reason I'm pretty loath to chase them. I'd prefer to invest in the other type of technology instead of chasing faddish stuff.
These days the benefits of this model of development (one page js app) is also hyped enough to drown out the fairly significant drawbacks - javascript behaves unpredictably under different browsers, is weakly typed, and the whole model breaks certain abstractions. I'm betting that the hype will die down but the drawbacks aren't going away.
I also don't think that switching ecosystems is quite as easy as you make out. Each ecosystem has enormous tool chains and you quite often need years of experience to know what the best of 500 libraries/tools to use in any situation is.
Once you start building an app with a lot of interactivity, sending certain states as prerendered pages becomes a huge distraction.
No. I think customers will stop wanting extremely javascript heavy interactive apps when they start to realize that 95% of the time they can have something that is cheaper and more reliable that isn't so javascript heavy.
Are you dying for hacker news to become a one page ember app?
Didn't think so.
Users don't really care about this stuff unless all that javascript brings huge gains (e.g. for a mapping app).
At least if you want to properly learn the language...
The hard stuff is problem solving, understanding algorithms and their applicability, implementing and using data structures, how computers work, how they represent types and resources, how networks work, network protocols, databases, processes, threading, asynchronous coding etc. All of this knowledge is mostly independent from programming language and platforms and will apply from one language to another. Learning the syntax of a programming language is trivial when compared to other programming topics.
If Ruby and Ruby on Rails dies tomorrow or Python and Django or whatever, and you know of the things I just mentioned picking up the next thing should be not be a difficult task. I don't really care what language someone already knows when it comes to hiring. If you don't know JavaScript, but you know Ruby or Java or C#, and you understand Computer Science, have demonstrated being able to write working code and are not afraid to ask questions picking JavaScript on the job is OK with me. We'll do code reviews, explains things, get you going.
Yeah, and knowing which libraries to use to do all that stuff and what quirks they have.
That type of knowledge doesn't translate overnight between programming ecosystems. It takes years.
It's almost worse when the stack is new because there is so much of the wheel that has to be reinvented again.
Drats, as I was writing this, someone did:
You might not like it but you need to follow the tool chains. Specific to your situation, I do not know anyone doing Ruby anymore for new systems. I am sure some people are but it is clearly on the decline. Continuous evolution of the tool chain is the curse of our industry. Right now, Javascript is probably much more valuable than Ruby.
If you do not like trading up, some skills are more durable than others. Over the long term, I've seen C/C++, Python, and Java retain their value better than most languages over the years. I think Javascript is likely to add itself to that list.
While you might find it depressing, a lot of what you have learned is immediately applicable in these other environments. You just need to figure out how they map, which is not too difficult. Going from Ruby to Python or Javascript is not that big of a stretch. In most cases, it is a weekend learning syntax and tool chain differences.