I mean, the type system isn’t very strict at all to begin with in some . It’s trivially breakable when performing common operations like type punning in C.

[–] [email protected] 16 points 2 months ago

Depends on the programming language

[–] GetOffMyLan 3 points 2 months ago (1 children)

That's why we don't like c

[–] embed_me 3 points 2 months ago

(Loads revolver)

[–] onlinepersona 11 points 2 months ago (1 children)

What did you actually do?

Anti Commercial-AI license

[–] [email protected] 27 points 2 months ago (2 children)

I broke the type system in Kotlin and ended up with a null value in a non-null (not lateinit) variable.

[–] [email protected] 6 points 2 months ago

How convoluted was it?

[–] [email protected] 1 points 2 months ago (1 children)

With no platform types involved?

[–] [email protected] 1 points 2 months ago (1 children)

Yes

[–] [email protected] 1 points 2 months ago (1 children)

Interesting. Care to share? I think they fixed a few of those issues with K2. Did you find something new or is it known?

[–] [email protected] 8 points 2 months ago* (last edited 2 months ago) (1 children)

It was about 3 years ago, and I haven't been able to find my old code again. I probably discarded it after finding out it didn't work well. I was able to find this screenshot though:

[–] [email protected] 3 points 2 months ago

Uh. Probably not exactly your case, but I managed to "reproduce" them.

The first case is easy if you just override the setter of a property, and is more a feature of the language rather than a bug.

The second case is more tricky, but by using a generic delegate you can manage to trick it by doing an unchecked cast. Got the same warning in IntelliJ as well.

Link: https://pl.kotl.in/pIt-uHw7d

[–] kryllic 2 points 2 months ago

Heinz Doofenshmirtz as senior dev