Francesco Nigro

Still the most expensive number of allocations here are the substrings - which should be just views of the original string, tbh.

Let's sync today for this @Sanne - which, regardless, still contains some logic error which break something

I'm closing this since this is going soonish to change :pray:

Nope, I have to investigate

I can help in that @septatrix ; I'm one of the Netty committer, so let's sync there ;)

@He-Pin do you think this could be useful to use to try the adaptive pool with vertx?

@zekronium > Why is the heap buffer not pooled IDK I would love to know it from @vietj probably there's some historical reason for that but I would like to...

@vietj this is adding the option to enable heap pooling when SSL is enable, which can be a saviour with JDK SSL, see https://github.com/quarkusio/quarkus/issues/41880#issuecomment-2258734835

PTAL @vietj

Not good @vietj ``` Starting test: ClusteredEventBusWithSSLTest#testSendCharacter [vert.x-eventloop-thread-1] WARN io.netty.util.concurrent.AbstractEventExecutor - A task raised an exception. Task: io.vertx.core.impl.future.FutureBase$$Lambda/0x00007f59dc23dcd0@54c5bbbf java.lang.NullPointerException: Cannot invoke "io.vertx.core.net.SSLEngineOptions.usePooledHeapBuffers()" because "this.sslEngineOptions" is null at io.vertx.core.net.impl.SSLHelper.preferredNettyAllocatorWith(SSLHelper.java:141) at io.vertx.core.net.impl.TCPServerBase.lambda$listen$4(TCPServerBase.java:243)...