[kartoffelsaft]

I hear Cardon get mention on rare occasion, and with how rare that is I have to assume it's been completely stagnant. Does it offer anything over C++ in current year? Seems like C++ interop begets turning your language into C++ with different syntax in a way that C interop just doesn't.

[ceteia]

The GitHub project has some activity at least, and they might be coming with some announcement later this year.

https://github.com/carbon-language/carbon-lang/

[pjmlp]

There is an announcement already planned at NDC Toronto 2026.

> Carbon: graduating from the experiment

https://ndctoronto.com/agenda/carbon-graduating-from-the-exp...

As for it being widely adopted, people keeping missing the point that Carbon is mostly for Google themselves, as means to integrate into existing C++ projects.

They are the very first ones to assert that for green field projects there are already plenty of safe languages to chose from.

[ceteia]

What concerns me is that the design of Carbon in aspects seem to have serious issues already now.

In case that you are well familiar with for instance pattern matching, might you have any opinions on the pattern matching that is currently proposed for Carbon?

https://docs.carbon-lang.dev/docs/design/pattern_matching.ht...

[pjmlp]

I am not a Google employee, as such I don't care where they take Carbon, other than being a technology nerd that had compiler design as one of the areas I majored in.

Regarding the linked pattern matching proposal, it seems alright to me, not everything has to be ML like.

[ceteia]

Are you really OK with runtime "expression patterns"?

    match (0, 1, 2) {
      case (F(), 0, G()) => ...
    }

> Here (F(), 0, G()) is not an expression, but three separate expressions in a tuple pattern. As a result, this code will call F() but not G(), because the mismatch between the middle tuple elements will cause pattern matching to fail before reaching G(). Other than this short-circuiting behavior, a tuple pattern of expression patterns behaves the same as if it were a single expression pattern.

How would that work with exhaustiveness checking? As far as I can tell, they themselves believe that Carbon's exhaustiveness checking will be very poor.

And OK with implicit conversions? Especially when combined with their way of handling templates for pattern matching?

[pjmlp]

As mentioned I have no interest in ever using Carbon, the language still isn't 1.0, and full end to end compiler is yet to be made available.