Dan Rusu

@blerer , I just remembered that the caching that I added before also included the relevant fields per class in addition to the shallow size to avoid having to re-traverse...

@blerer a few clarifications and then I'll include my use-case at the bottom: - The proposal doesn't require all 3 parts as any of them can provide improvements on their...

Thanks, submitted PR: https://github.com/jbellis/jamm/pull/68 The cache is temporarily created for each invocation of measureDeep so that should address the issue with my initial proposal about it growing unbounded. I also...

**JMH Benchmarks:** JMH version: 1.36 VM version: JDK 1.8.0_392, OpenJDK 64-Bit Server VM, 25.392-b08 Notice how the **BenchmarkMeasureCollections** times improved dramatically along with even tighter margins of error for these...

> @daniel-rusu Thanks for the patch. I will be curious to see the JMH result when running the benchmark with Java 17 as the numbers in general differ a lot...

@blerer I'm curious if I can do anything to help get this merged as It's been a couple of months. Since it only compiles on JDK 8, I'm not sure...

Looking at the KOperator enum, it looks like it doesn't have any representation for the index access operator forcing it to be null in MemberName thereby bypassing the check in...

The proposal to use `for { ... }` goes in the opposite direction and reduces clarity rather than improving it as `for` is usually used in the context of "for...

Kotlin 2.1.0 introduced the ability to have non-local `break` and `continue` in inlined lambdas: https://kotlinlang.org/docs/whatsnew21.html#non-local-break-and-continue Given this new ability, overloading the `repeat` function to be an infinite loop when the...

@melix , unfortunately `--rerun` also reruns the task's dependencies so it's not a good solution as it rebuilds the entire project again and re-generates the benchmarks. From a technical Gradle...