https://github.com/hsutter/cppfront

People working towards syntax improvements as well to C++ now, which they believe will pave the way for semantic improvements that break backwards compatibility with C++

Taken from the repo:

On Saturday, 17 September 2022 at 11:08:37 UTC, Tejas wrote:

That's great to hear, C++ is a giant and uncontrollable mess

Already looks better than Carbon

How funny: https://github.com/hsutter/cppfront#2021-is-as-and-pattern-matching

https://github.com/dlang/vision-document#other (in #other, unfortunately)

On Saturday, 17 September 2022 at 11:41:18 UTC, ryuukk_ wrote:

I'm just worried about D's place in this world where there's Carbon, C++26, Rust, Nim and now this 😔

On Saturday, 17 September 2022 at 11:55:16 UTC, Tejas wrote:

You forgot about:

• Val, https://www.val-lang.dev from Dave Abrahms and Sean Parent

• Verona, https://microsoft.github.io/verona/ from Microsoft Research

• Odin, https://odin-lang.org/ already used in the games industry, via EmberGen

Also how Go, Java, C#, F# are improving their low level coding features (and AOT story in Java/.NET case), so that their need for C and C++ gets reduced.

With Val, Carbon and now Cppfront coming out of the C++ community itself, we are at an inflection point, I bet C++26 might be the latest big revision.

So yeah, the competition to D is getting stiffer, and it is a question of how many of those C++ disillusioned souls might eventually find a home in D.

On Saturday, 17 September 2022 at 11:08:37 UTC, Tejas wrote:

When C++ compile times aren't long enough, let's make them longer.

My first question that comes to mind, why compile this to C++ when you have C which is much faster and you can do the same thing. Essentially this is what Nim does which has full meta programming that can be compiled to C.

C++, isn't holding up anymore and is too simple let's make C++11
Let's make C++11 modern -> C++20
C++20 has a complicated and ugly syntax, let's fix that with CppFront.
... and it just goes on, taint it up, taint it up.

Even more funny is that this is from Herb Sutter which is one of the biggest influencers of "modern C++".

On Saturday, 17 September 2022 at 11:55:16 UTC, Tejas wrote:

Indeed, simply asking for .Enum was met with negativity, so i don't see our fate improving anytime soon

On Saturday, 17 September 2022 at 17:47:57 UTC, ryuukk_ wrote:

Language design is about trade-offs, expect people to disagree about things. That doesn't mean it's not worth trying to make improvements. There is a DIP PR for $member (.member is ambiguous).

On Saturday, 17 September 2022 at 14:57:40 UTC, Paulo Pinto wrote:

Val and Verona are extremely early stage though, no? But that might not be the case ~2025...

Didn't know Odin was already getting used in industry, pretty cool...

Hope we manage to get our stuff together and not just survive, but thrive in this decade

✊

On Saturday, 17 September 2022 at 14:57:40 UTC, Paulo Pinto wrote:

I don't know. I think C++ has become fairly well-rounded now that clang is catching up on C++20. At this point it will take a while for the C++ community to make good use of C++'s take on coroutines and concepts. It would probably be a bad idea to continue to push in more big features. Some smaller ones like SIMD are missing still. But in the longer term I think we will see more standardized hardware oriented features related to parallelism, co-processors etc. I suspect Intel and AMD will have to do something to ensure their own relevance in the long term, and C/C++ is where they can make software "hardware dependent". So in that sense C/C++ has a guaranteed long life span. System programming isn't just market driven, it is also hardware driven.

On Saturday, 17 September 2022 at 11:55:16 UTC, Tejas wrote:

It'll be ironic if the thing that Walter said will kill C actually ends up killing D cause it basically put everything else on the back burner for 10 years.

Is DIP1000 the new Phobos vs Tango?