okio VS kotlin-jupyter

I am involved in the development of Android application which is a rather "thick" mobile client for a Web service. It heavily communicates with the server but also has a lot of inner logic too. So, I decided to use some features of Google Guava library to simplify development process. Here is a list of features I'm very interested in: immutable collections, base utils, collection extensions, functional programming sugar and idioms (common.collect and common.base), primitives utilities (common.primitives), hashing utilities (common.hash), concurrent utils (futures and AsyncFunction). Things I don't want to use in Android: common.cache (see question below), common.eventbus (we have better Android specific libs for this, such as Otto), common.io (we can use okio for Android now).

Yeah Kotlin's own standard library is a lot smaller than Java's currently so you'll need to use something third-party for this. Okio is a popular option https://square.github.io/okio/ it has a Buffer type which is pretty similar to Java's ByteBuffer

Use okio, it is Multiplatform now. I use this for my own library KStore

For networking, file IO, and streams in general, there's Korio and for Java; for just networking, there's LiteNetLib for C#; for what looks like data streams in general, there's Okio also for Java; and Tokio for multi-threaded IO in Rust.

Okio is nice for input/output streams, and sockets.

Sounds like you want https://square.github.io/okio/

On JVM, there is plenty of existing solution already on for multiplatform uses I'd suggest checking amazing Okio library by Square, that seems to cover most of basic use-cases.

I think the main multiplatform library for I/O currently is okio, https://github.com/square/okio. As for compression, you should probably create an expect class for the compressor and use platform specific calls for actual compression. Take a look at https://kotlinlang.org/docs/mpp-connect-to-apis.html.

For working with datasets (loading and processing), I use Kotlin DataFrame. It is a library designed for working with structured in-memory data, such as tabular or JSON. It offers convenient storage, manipulation, and data analysis with a convenient, typesafe, readable API. With features for data initialization and operations like filtering, sorting, and integration, Kotlin DataFrame is a powerful tool for data analytics. I also use the Kandy - Kotlin plotting library, designed specifically for full compatibility with Kotlin DataFrame. It brings many types of plots (including statistical) with rich customization options via a powerful Kotlin DSL. The best way to run all of this is Kotlin Notebook. It works out of the box, has native rendering of Kandy plots and DataFrame tables, and has IntelliJ IDEA support. It can also be run in Jupyter notebooks with a Kotlin kernel and on Datalore.

That is true for the Plugin, but you can get the Kotlin Kernel for Jupyter Notebooks separat. https://github.com/Kotlin/kotlin-jupyter

Kotlin kernel for IPython/Jupyter