Is this a good use case for a class instead of a struct?

[Blacksmoke16]

Dot notation just accesses a property/method on the specific object you use it on. Whether that modification edits another object somewhere else is depended on if that object is a class or struct, and where it came from. I.e. all in same scope, or passed in as a param of a method etc.

@girng

dot notation modification does work when you are accessing the “struct” itself , but not when accessing the local copy of the struct . Well, it does work… on that LOCAL copy, but not the main struct.

Correct, because they are two separte objects since it was a struct it got passed by copy, not reference. If it were a class, then it would have been passed by reference and would have been the same object.

Read through Carcin a few times. I’m hoping it’ll make things more clear on what’s going on for you.

@girng

When a developer is now modifying a property that is a local copy and not reassigning it…it now becomes an illusion to the developer and might create bugs?

Again i can’t stress it enough that dot notation just edits a specific object, what the side effects are from doing that stem from whether the object as a struct or a class. As demonstrated in the playground link.

[schovi]

@bcardiff

Maybe something that the compiler could do is that there are changes in a value type and the value is never used again after the last change.

That is more like linter job and not a compiler

[bcardiff]

@schovi

That is more like linter job and not a compiler

It requires a semantic phase run to know whether the variable is of a value type or a reference type.

I think that lints works only at a syntax level.