C++ was a dying language in the 2000s. But they turned it around with C++11. Now it has iterative 3 yearly updates and it's getting pretty good.

It's far too premature to declare its death. The pace of development of C++ is decently fast and there's many stakeholders with millions of lines of C++ who are interested in C++ not dying. C++ is taught in universities around the world so the chance of C++ developers becoming expensive is low.

In fact, the number of C++ programmers is growing faster than ever. This is evident in the growing number and size of C++ conventions. (Bulgaria has its own, now.) Even ISO C++ Standard committee attendance is reliably a lot more than every previous meeting, even when held in inconvenient places. Attendance will hit 500 soon. The amount achieved at each meeting grows accordingly. Proposals get more scrutiny and draw on more experience.

To suggest C++ is a dying, slowing, or stagnant language is objectively delusional. The language is healthier than ever, and going gangbusters.

The C++ that they teach you in universities barely gets you started. I wouldn’t count on those for a steady supply.

The C++ they teach in universities is either just C, or putting Java spin on C (C-with-classes). I've never seen a decent "modern C++" class taught in universities.

Sure, but that's nothing a few years in the real world can't fix. The point is that C++ programmers are not some rare breed of neckbeard that are gradually going extinct. They are being produced en mass.

I love that Java spin on C remark, given that stuff like Turbo Vision, MFC, OWL, PowerPlant, Motif++, CORBA, COM, VCL, SOM, CSet++, POET precede Java for several years, some almost a decade.

While we see some adoption of Rust from Google and Microsoft, all their new releases are heavly focused on C++.

React Native, the new Games SDK for Android, WinUI 3.0 release, lifetime analysis support for Visual C++ and clang and lets not forget they are two major contributors to ISO C++.

While the Rust/WinRT efforts from Kenny Kerr are welcomed, that it is just yet another option alongside C#/WinRT, C++/WinRT, Python/WinRT and JavaScript/WinRT.

COBOL has been dying since the early 80s yet to this day more than 80% of all daily business transactions are processed in COBOL and 5 billion lines of new COBOL are written every year. If this is what a dying language looks like then yes, C++ is a dying language.

P.S. I'm not a COBOL programmer nor have I ever been a COBOL programmer.

The only language that one can honestly say that is dying (if not already dead) is VisualBasic. Perl and Delphi are probably going to continue to decline. Objective-C will be obsoleted by Swift some day, but is still relevant. Ruby's not too hot now, but far from dead AFAIK.

Other than that it's nonsense to say that X is dying.

C++ is being blamed for allowing people to write buggy, insecure programs. In my mind, that's different to just becoming unpopular; it's dying because it's too dangerous and too complicated.

Sure, it'll stick around for some years yet but all those jobs will become maintenance jobs. Very little greenfield projects will elect to use it outside of niche areas (games and maybe embedded - UK salaries for embedded dev are terrible). Both C++ and Rust can use LLVM, so the generated code is likely to be the same.

If you were starting a greenfield project today why would you choose to do it in C++?

I forgot to mention Go too.

In fact, essentially all new development in the highest-paid financial, HPC, CAE, telecomm, aerospace, and ML fields is pure C++. Rust is not a blip in any of them, and will not be for years to come. C is dead there (except in aerospace and telecomm), Go is nonexistent, and Java is mentioned on rare occasions.

To suggest C++ usage is in decline is, objectively, nothing short of delusional. For Rust, C++ will the one to beat in ten and twenty years, provided not too many have already skipped onto some new hotness.

Rust is an objectively better design than anything else that has arisen in 30+ years. That doesn't guarantee success.

I used to hope a better language would displace C++, and for a time thought Rust might be it, but that better language turned out to be C++ itself. Nothing else is even on the horizon.

My company starts new programs in C++ because everyone knows C++ already, we write modern C++, our libraries are written in C++, and our bugs are mostly not memory safety issues.

I would choose C++ because the tooling is superb, talent is widely available, it is performant, there are libraries for pretty much everything, and it's a great language.

Oh and by the way, LLVM is 100% c++.

Saying it is too dangerous and complicated is absurd. It is easy to write safe code in C++, having been doing it for almost 30 years.

You are throwing around lots of assertions - challenge you to back them up with real data.

The world runs on C/C++ code. And new C++ projects are started every single day.

I don't understand your claim that C++ is dying. Do you have data to back this up?