latin1StringToBytes: use standard method

[magicprinc]

1. String#getBytes(StandardCharsets.ISO_8859_1) is ~4% faster than manual copying in a for loop
2. and one line of Java code instead of 6

➕ missing @Override

[tsegismont]

Can you add the benchmark which shows this is better than the for loop?

[tsegismont]

Please don't change this, it won't change much and is unrelated to the subject of the PR

[magicprinc]

IDEA shows me a lot of recommendations and best practices 😥

[tsegismont]

Well, with Intellij's default, you get thousands of suggestions. It's not wrong to use them, but if you're sending a PR for a specific topic, please focus on it

[magicprinc]

Fixed!✅🤝

[vietj]

@magicprinc you should avoid doing too many changes in your pull requests, just change what is strictly necessary.

In practice we need to maintain the code on several branches and changing the way code looks make porting the code more difficult which means that the team porting and reviewing the code spend more time than necessary.

From now on we will close any PR that does not do the change advocated in the PR

[magicprinc]

@tsegismont I already knew the results, but I have made a benchmark with many explanations for you.
I have used your JMH defaults. I've removed Buffer from the benchmark to make the results loud and clear.

Benchmark                                 Mode  Cnt    Score   Error   Units
 baseline (almost no-op)                  thrpt  10  209,947 ± 0,449  ops/ms
 original vert.x copy-in-for-loop version thrpt  10    1,356 ± 0,005  ops/ms
 original version with micro-optimization thrpt  10    2,501 ± 0,352  ops/ms
 my JDK version 1                         thrpt  10   24,501 ± 0,357  ops/ms
 my JDK version 2                         thrpt  10   24,387 ± 0,067  ops/ms

[magicprinc]

@vietj Don't get me wrong: you ignore my PRs or take only few bits from them, and I don't have too much time to spend on this project. We are mostly a "Spring shop", but I hate WebFlux and try to remove it (we have no WebFlux experts and a lot of bug reports). With Vert.x and Virtual Threads, I am on the winning streak, but my colleagues are still very cautious.

So please 🙏 accept (with modifications)
this #4957 (I especially need this protected Object metric) and this vert-x3/vertx-web#2521

maybe these (mostly aesthetic) #4980 and public static <E extends Throwable> RuntimeException throwAsUnchecked(Throwable t) throws E {+throw Utils.throwAsUnchecked

And you won't see a PR from me 🥂 (only issues 😅)

[magicprinc]

@vietj If we speak about issues:
#4990 and #4971 would be great

I won't add it to issues, but more integration with https://github.com/smallrye/smallrye-fault-tolerance or vert.x alternative would be great.

BTW, We started to use Mutiny, but after some time returned to "bare" Vert.x with occasional Mutiny and some bits from it (mutiny-runtime and mutiny-zero). No offense to Mutiny intended: Mutiny is a breeze of fresh air compared to Spring Reactor/WebFlux.

PS: Vert.x 4.5.1 with Jackson Pool is a hot subject.

And thank you very much for such a great project!

[vietj]

@magicprinc then just provide PR that go straight to the point and don't touch code surrounded, we like contribution but we love when they are easy to review and easy to backport.

[vietj]

also please avoid remerging the main branch in your PR, rather rebase your PR otherwise again it's complicated to port.

[magicprinc]

@vietj Please forgive me (and somehow merge) these times. I am really out of new ideas and time, so no more PRs 🙏

[tsegismont]

Squashed and merged in bcddbac