As with any (new) language, there's no point in replacing every existing technology written before, with a new implementation written in Rust. While at times there are benefits such as with ripgrep instead of grep, this is not a generally useful strategy not a desired one.
In case you have good reasons, such as security and being able to start with a clean slate for the most common use cases, it might however make sense to write critical software in a safe language if it was previously written in a so called unsafe language. This should be done with the greatest consideration and only if worth the costs it brings with it. Our Dutch colleagues working for "de Tweede Golf" are deeply invested in Rust and have now a version of
sudo written in Rust. You can read about it at https://tweedegolf.nl/en/blog/91/reimplementing-sudo-in-rust.
Of course, starting new projects inspired by classic/foundational tooling is however an opportunity to use Rust, even if your users do not need to know or care about it. A great example of that is just,a modern alternative to Makefiles, as much as love as we have for it. You can find it at https://github.com/casey/just, basically 'Make' with variables + more.
Graphite is an in-development raster and vector graphics package that's free and open source. It is powered by a node graph compositing engine that fuses layers with nodes, providing a fully nondestructive editing experience. It's fully written in Rust, and can be used as a desktop app or in the browser. The crazy thing is that as an artist using it you are basically "painting" Rust, as Graphite documents are really rust programs. As such most of their contributors are actually people with compiler/language backgrounds. Most users however will never know this, and neither do they have to.
- In Appendix II. we already saw that Rust can be compiled to WASM, and as such you can build an entire WebApp in Rust. However instead of replacing your entire WebApp, it is also just as well possible that you only replace the components of your App where it makes sense, e.g. where you need to do a lot of computations;
- In Appendix III. we learned about Tauri to build native apps with Rust, and also here noted that it can be done while at the same time building your core frontend logic in the frameworks you are used to;
At times however you might be able to incorporate the benefits of Rust in your existing benefits without having to expose the language to anyone:
- FNM is a replacement for NVM. While the latter is more powerful the first is a lot faster, which speeds up terminal start up times by a lot, just to give one benefit;
- SWC is a speedy light alternative to babel, with the first outperforming the last with its eyes closed and 1 foot still in the bed;
These give just a small taste of what Rust can mean for Web developers even if they do not know or use Rust themselves directly.
- Pydantic is a library for Data validation using Python type hints.
- Ruff is An extremely fast Python linter, written in Rust.
Python web developers can also make use of Rust to write their Python web services, without having to know Rust:
- Granian (https://github.com/emmett-framework/granian):
- A Rust HTTP server for Python applications (written on top of Hyper);
- Supports ASGI/3, RSGI and WSGI interface applications;
- Example usage: run your FastAPI web service :)
- Robyn (https://github.com/sparckles/robyn):
- A High-Performance, Community-Driven, and Innovator Friendly Web Framework with a Rust runtime.
- In contrast to
Granianthis does mean you need to write your service using the Robyn way to do things, so not possible to use an existing service built with FastAPI.
And plenty more will follow surely.
A language is a tool, but a tool is not to be applied to any problem. Rust has its advantages and use cases, it would be foolish not to use it where desirable.