Thread: RSPS Networking Library Choices

Originally Posted by hc747

Personally used Ktor and Micronaut (both in Kotlin) and have a preference for the latter. Not sure why you're comparing netty to Ktor (etc) as most of them use netty under the hood.

What's the use case for yours?
I think it's a decent comparison if you look at things such as what ktor offers and netty doesn't, and compare benchmarks [for instance].

Originally Posted by hc747

Used each in web API development at work. I used Ktor ~2 years ago - found it intuitive, but too ceremonious. Micronaut is very enjoyable to work with; it's flexible, has widespread adoption and framework support, is high level enough to keep things simple, but also allows for you to be as granular as you need to be. Unfortunately find myself using Micronaut with Groovy (instead of Kotlin) for the time being. Oh well.

Ngl I had too much fun reading that second sentence hahahaha. I'm looking it up rn.

-- edit

[Only registered and activated users can see links. ]

this is pretty juicy.

Not a single networking library can touch [Only registered and activated users can see links. ]

Leaves them all in the dust.

Originally Posted by Gluon

Not a single networking library can touch [Only registered and activated users can see links. ]

Leaves them all in the dust.

Why not?

This is interesting: http://www.coralblocks.com/index.php/coralreactor-vs-netty-performance-comparison/

Originally Posted by Tyluur

Why not?

This is interesting: [Only registered and activated users can see links. ]

Consistency, performance, simplicity. It beats them all across all categories.

Originally Posted by Gluon

Not a single networking library can touch [Only registered and activated users can see links. ]

Leaves them all in the dust.

That benchmark is bs and the company looks sus. Let me tell you why that benchmark is bad. The code comparison given doesn't even use Netty's ByteBuf but java.nio.ByteBuffer, that alone should tell you this is at best a bad comparison by someone who doesn't understand the library and at worse deliberately misleading.

Originally Posted by Fire Cape

That benchmark is bs and the company looks sus. Let me tell you why that benchmark is bad. The code comparison given doesn't even use Netty's ByteBuf but java.nio.ByteBuffer, that alone should tell you this is at best a bad comparison by someone who doesn't understand the library and at worse deliberately misleading.

Coral blocks has been around for years, in fact they've been around a decade longer than I've been on this community.

I was just never able to use it because it's not a free and open source library.