Comment by gmueckl
15 hours ago
There have been several talks about contracts and the somewhat hidden complexities in them. C++ contracts are not like what you'd initally expect. Compiler switches can totally alter how contracts behave from getting omitted to reporting failures to aborting the program. There is also an optional global callback for when a contract check fails.
Different TUs can be compiled with different settings for the contract behavior. But can they be binary compatible? In general, no. If a function is declared in-line in a header, the comoiler may have generated two different versions with different contract behaviors, which violates ODR.
What happens if the contract check calls a helper function that throws an exception?
The whole things is really, really complex and I don't assume that I understand it properly. But I can see that there are some valid concerns against the feature as standardized and that makes me side with the opposition here: this was not baked enough yet
That sounds like the worst kind of misfeature.
It sounds like it should solve your problem. At first it seems to work. Then you keep on finding the footguns after it is too late to change the design.
Coroutines went through the same cycle. Standardized in C++20, and I still hit compiler-specific differences in how symmetric transfer gets lowered.