actix VS hatch

9. Actix

I think it's fairer to say the language got so much more powerful that there wasn't any point making actors a language feature when they can be built from existing orthogonal language features. You're probably looking for actix (not actix-web, just actix). There's also Lunatic built in Rust but supporting any actors compiled to WebAssembly.

Actix is an actor framework for developing concurrent applications built on top of the Tokio asynchronous runtime. It allows multiple actors to run on a single thread, but also allows actors to run on multiple threads via Arbiters. Actors can communicate with each other by sequentially exchanging typed messages.

I’ve seen frameworks for c++ (https://seastar.io/) and rust (https://github.com/actix/actix) which support what you’re describing out of the box.

The actix repository on github doesn't seem to be very active, and everyone seems to be focused on actix-web instead, is anyone out there using plain actix or any other actor-model implementation in Rust?

Rust is also another possibility: it's basically C++ but more modern with added features and safety. It can be tricky to write mathematical stuff in it, because you may not care too much about all the safety concerns Rust forces you to handle, but it can be useful to catch bugs ahead of times. Sadly, Rust seems to have no library for running programs on clusters of PCs, except maybe this one, which takes the Actor model implemented by Actix and runs it on a cluster. I don't know how tricky it is to use the Actor model for a scientific simulation, tho.

I'd just use Actix for that. Make your state an actor and make it a StreamHandler for each of these sources, and that's it - now you just implement the business logic for handling each message in the StreamHandler::handle methods.

However, this isn't even 50% of what's out there: Need raw parallel power (and maybe don't need an async runtime)? Checkout Rayon. Need simple Actors for concurrent processing? Checkout Actix. Need a larger Actor system for fault tolerance/CQRS messaging? Checkout Riker. Damn, I sound like a youtube advert 🤦 - For real though, this is the tip of the concurrency iceberg. There is so much more - and it's growing.

Exciting stuff! I view Hatch [1] as becoming the Cargo for Python because it's already close and has an existing (and growing) user base but I can definitely see depending on this for resolution and potentially not even using pip after it becomes more stable.

[1]: https://hatch.pypa.io/latest/

I was inspired enough by the hatch sync idea that I created a PR to add that functionality to hatch: https://github.com/pypa/hatch/pull/1094

Another concept I learned was about build backends, an import step which is used to initialize and install any dependencies of the app you're packaging. Since the tutorial went with using Hatch that is also what I went with, though it didn't provide a lot of useful details especially because it didn't show how to add any dependencies, so I took a look at the docs which were very nice and simple to follow.

Try using hatch: https://hatch.pypa.io/latest/

I'm trying to learn Hatch, I currently use [Poetry](python-poetry.org/) to manage my dependencies, and while I'm overall happy with it, I really like the features I'm reading about with Hatch. I'm also working on learning CI pipelines & Dockerizing Python applications, and Hatch seems like a really useful tool to learn for this (and just as a general use tool).

Poetry 1.2 has been a pain. Which was the dev's fault though. Switching to something new while deprecating a related feature is just plain bad. I've been looking into modern alternatives like PDM and Hatch, but haven't used them (yet).

For example, when it comes to Python, one option is to use the same packaging system that a huge number of open-source libraries and tools are published with. You can use setuptools or Hatch to build a "packaged" version of your code, and publish it to either the public PyPi repository or an internal one that you set up. Then your users can use pip to install your package, automatically fetch its dependencies, and keep it up to date, just like any other Python module.

Don't forget the new PyPa tool on the block: Hatch.

See also: https://github.com/pypa/hatch