Re:Memory Safety Without Lifetime Parameters
by einpoklum on Monday October 21, 2024 @07:16AM
(#41902915)
> which one step after another fixes the problems of C++
Somehow, multiple groups of people have recently decided they want to "fix the problem": Carbon, Circle, CppFront2. But apparently, it's different fixes, and the formulation of the "the problem" is different.
Re:Memory Safety Without Lifetime Parameters
by flohofwoe on Monday October 21, 2024 @10:54AM
(#41904855)
AFAIK the big difference of Circle to both Carbon and CppFront2 is that Circle is (mostly?) a superset of C++17, so it doesn't change the basic syntax, just adds things on top.
I don't care much either way tbh, just watching from the sidelines while happily churning out C and Zig code ;)
But IMHO Sean Baxter does more for C++ than the entire committee ever did in the last three decades - if C++ would have moved into a similar direction I probably wouldn't have dropped it as my main language).
Re:Memory Safety Without Lifetime Parameters
by fithisux on Monday October 21, 2024 @03:04PM
(#41907298)
It fixed the problem with C++ and in an exceptional way.
Now with @live, even when GC is on, they are ahead of Rust.
I love it.
Re:Memory Safety Without Lifetime Parameters
by einpoklum on Monday October 21, 2024 @05:04PM
(#41908517)
An annoying but rather valid quip by Bjarne Stroustrup is that "there are only two kinds of languages: the ones people complain about, and the ones nobody uses." D has, statistically, not been attactive enough for people to switch to it from C++. Perhaps this is unfortuante... ? I wouldn't mind a link to an argument in favor of switching to (today's) D from (today's) C++.
If the name Sean Baxter doesn't ring a bell, he's creating the Circle compiler which one step after another fixes the problems of C++ (while not trailing off into a completely new language):
https://www.circle-lang.org/quickref.html