Hacker News

points by wseqyrku 4 days ago | 0 comments

> As I understand it, this step would mean reverting to C++23 and proceeding to C++29

What kind of design process is this

That's just a misunderstanding, probably rooted in a pre-published paper that claimed so.

In reality WG21 could just decide to start work on C++29 based on the failed C++26 draft (or the current working draft has already incorporated some C++29 changes)...

[–] locknitpicker 4 days ago

> What kind of design process is this

It is not. This is a lame attempt scaremongering, to try to inflate the consequences of contract proponents not getting their way. It's like sore losers threatening to take the ball home to prevent everyone from continuing to play the game, and claiming that taking the ball home is how games are normally played.

In practice, the standard passes the moment the sticking points are ironed out. This likely means cutting out contracts yet again, which has been done every single release since at least C++20.

[–] bluGill 4 days ago

Contracts were never in C++23 - C++26 was the goal when they cut it from C++20. The C++ cut was done by the people pushing contracts, while this is done by a different group.

[–] locknitpicker 4 days ago

> Contracts were never in C++23

Why yes, because the C++20 was so poorly thought through that in the July 2019 WG21 meeting the design was deemed irremediably broken, and consequently the committee made the unprecedented decision to just disband the group and start over by forming a new one. This happened so close to the kickoff that contracts were right there excluded from C++23.

And after all these years, the proposal remains broken.