Andrej Fink

Last possible optimization: Replace juicy DelegatingScheduler.ScheduledCompletableFuture with a lightweight, almost empty implementation. CompletableFuture is quite fat if used only as dumb "placeholder".

> Side note: if (delay == 0) is more of a specification problem in Scheduler, right? Should negative delay values be accepted in the first place? JDK's ScheduledThreadPoolExecutor accepts them...

I did a good job. And I hope the maintainer will find my changes interesting. It is easy to see the difference: simply generate 1 billion scheduled "events" and see.

I did a (not scientific) memory test: Old scheduled task in memory: 193 bytes New scheduled task in memory: 136 bytes 70% of original https://github.com/magicprinc/failsafe/blob/9f23989e3bf77e9b525d0863ba89b89075240355/core/src/test/java/dev/failsafe/internal/util/DelegatingSchedulerTest.java#L404

Metrics are my colleges' responsibility, so I don't have a concrete answer yet :-) But usually people like if somebody smarter than they—"an expert in the field" have already made...

Core library without dependencies is a good thing! It could be an additional (optional) artifact. And people who use let's say Prometheus directly could clone the project and make a...

> Aren't those braces ignored, anyway? It's just a single expression. (The indenting of the modified code is off, btw.) IntelliJ IDEA's work :-) As I know, the resulting byte...

Your wish is my command. Tell me if I should send an PR.

@nao-it No, I have another problem. I use some library, which depends on new h2. So, thers is a conflict between new h2 and Ignite's h2. The good solution is...

It works! :-) See more information here https://github.com/bsideup/jabel/issues/175