Michał Klimczak
Michał Klimczak
Oh, looks better than what we finally merged, thank you! Still, could be even simpler if you allowed for nullable type inside MemberName. Although I must admit, that I feel...
In my own project I will be generating ios source sets from common code so I won't need that, but I noticed that when playing with different source set configurations....
I believe so, I even have my processor defined as a jvm module, but I'm not sure if it's a problem. My case was: 1. Have annotation on commonMain source...
Here's the repro :) https://github.com/micHar/kotlinpoet-ksp-repro
It turns out that simply changing the order of checks is enough to fix the issue. PR awaiting review (you can check that it works with the reproduction repository attached...
It's explained a bit here https://github.com/google/ksp/issues/965, though I'm not sure what will be the final form of multiplatform KSP. In general, imagine a processor, that takes common code as input...
Sure, it's [here](https://github.com/google/ksp/issues/1042).
Tbh, I would go as far as to include an option to opt out of using `@Preview` annotations altogether. If I can achieve all the same things with a dedicated...
Hmm, I'm not a big fan of this solution. It's like the library hijacks the more general annotation for its specific purpose. Would be more elegant to at least ignore...
Yes, seems like a better idea to warn than to crash. Generally I would agree with you when it comes to the fail fast principle, but here it feels like...