These new features are certainly nice for die-hard C++ programmers. However, I wonder whether, given the large amount of syntax and features, it is not a miracle that there are still compilers that can parse and translate C++ properly…
I don't know if it's a miracle or just a lot of hard work by a lot of smart people, but it's pretty amazing that pretty much before papers even make it to committee vote that we usually have at least prototypes in one or both of the major, free, open source, cross-platform compilers and libraries.
There are some big things (modules) that are exceptions but most of this explosion in a punctuation factory syntax is available quickly.
Cause modules didn't go through the same process to get accepted into the standard as other features did. And C++26 won't be fully implemented until 2030 or so. Execution policies from C++17 are still not implemented in Apple Clang, and Clang only has partial support for it.
So ... modules are a notable exception and large parts (I carefully didn't say "all") are already available and usable. Got it. Totally different. :-)
I mean MSVC still doesn't implement some of C99 and it looks like there's a super obscure feature that only SunOS C compiler implements. The meter almost never goes to 100% on all the platforms everywhere - certainly not before ratification.
My underlying point was that this isn't like the early days of C++ when implementations lagged the standard by many, many years. Most of the actively-developed, big name, cross-platform compilers (I've heard that Microsoft has gotten better lately after holding back developers on its platforms for many, many years. Taking 16 years to implement snprintf just wasn't funny.) For the ratification to be still probably 18 months away and the trickle of papers only very recently shut off to turn toward stabilizing and shipping, we already have a lot of green at
https://en.cppreference.com/w/cpp/compiler_support/26 - and a lot of that came in snapshots late last year and releases early this year. So it's not like the days of TR1 boondoggle's "not a spec" that tried to paper over an eight year void.
Yes you're right, but I wasn't using C++ back then. The lack of support for execution policy really bit me hard recently when I was implementing a cross platform app that could've benefited from it.
Harzer-Zwerg@reddit
These new features are certainly nice for die-hard C++ programmers. However, I wonder whether, given the large amount of syntax and features, it is not a miracle that there are still compilers that can parse and translate C++ properly…
YetAnotherRobert@reddit
I don't know if it's a miracle or just a lot of hard work by a lot of smart people, but it's pretty amazing that pretty much before papers even make it to committee vote that we usually have at least prototypes in one or both of the major, free, open source, cross-platform compilers and libraries.
There are some big things (modules) that are exceptions but most of this explosion in a punctuation factory syntax is available quickly.
Large parts of c++26 are already usable.
CramNBL@reddit
Cause modules didn't go through the same process to get accepted into the standard as other features did. And C++26 won't be fully implemented until 2030 or so. Execution policies from C++17 are still not implemented in Apple Clang, and Clang only has partial support for it.
YetAnotherRobert@reddit
So ... modules are a notable exception and large parts (I carefully didn't say "all") are already available and usable. Got it. Totally different. :-)
I mean MSVC still doesn't implement some of C99 and it looks like there's a super obscure feature that only SunOS C compiler implements. The meter almost never goes to 100% on all the platforms everywhere - certainly not before ratification.
My underlying point was that this isn't like the early days of C++ when implementations lagged the standard by many, many years. Most of the actively-developed, big name, cross-platform compilers (I've heard that Microsoft has gotten better lately after holding back developers on its platforms for many, many years. Taking 16 years to implement snprintf just wasn't funny.) For the ratification to be still probably 18 months away and the trickle of papers only very recently shut off to turn toward stabilizing and shipping, we already have a lot of green at
https://en.cppreference.com/w/cpp/compiler_support/26 - and a lot of that came in snapshots late last year and releases early this year. So it's not like the days of TR1 boondoggle's "not a spec" that tried to paper over an eight year void.
CramNBL@reddit
Yes you're right, but I wasn't using C++ back then. The lack of support for execution policy really bit me hard recently when I was implementing a cross platform app that could've benefited from it.