A lot of people hear "micro SaaS" and immediately reach for the usual stack.
Use JavaScript. Ship fast. Fix the weird production issues later. May...
For further actions, you may consider blocking this person and/or reporting abuse
Great read! Who doesn't love Rust ☺️
tried the ship-fast-in-JS approach for a similar project. maintenance overhead after 18 months was brutal. the initial velocity advantage was completely erased by then.
Can imagine, yeah.
18 months is when the shortcuts become structural. Same pattern here — velocity gains were real early on, then rewrites kept surfacing how deep the quick assumptions had gone downstream.
I've been using Rust for a lot of the same reasons
Easy solution for quick rust development is a vibe code which is also work if developer do experienced in rust: dev.to/pengeszikra/rustroke-wasm-m...
The "boring in production" framing is exactly right, and it's underrepresented in most Rust discussions which tend to focus on raw benchmarks rather than operational character.
The point about refactoring confidence is what I find most compelling for long-lived projects. Dynamic language codebases tend to accumulate "maybe this is still used somewhere" code. Rust's type system makes those decisions explicit — you either handle the type or you don't compile. For a solo operator, that's not a performance benefit, it's a cognitive load benefit. You can come back after six months and the compiler tells you everything that changed.
One thing I'd add to your list: Rust's low memory footprint has an unexpected effect on infrastructure cost at scale. Sub-10MB RSS per worker process means you can run many more instances on the same hardware before hitting limits, and you never deal with the "GC spikes are causing latency outliers at peak load" ops conversation that haunts high-traffic Node or JVM services.
What's your experience been with Rust compile times in CI? That's usually the pain point people hit once the project grows past a single crate.
"A product that feels boring in production" — that's actually the best framing I've heard for why Rust is worth the initial friction for SaaS work.
The reliability angle resonates more than the performance angle for exactly the reason you describe. When you're the only person on-call at 2 AM, you don't want to debug a garbage collector pause or a null pointer that slipped through a dynamic type system. You want the compiler to have already argued with you about that last month.
The "haunted house" metaphor is accurate too. Python/JS services have a way of accumulating subtle side effects that nobody fully understands anymore — a Rust service tends to stay legible because the type system won't let you put something in a box and forget what's in it.
How are you handling the DB layer? Did you go with Diesel, sqlx, or something else? Curious how the compile-time query checking changes your migration workflow in practice.
i am using mongodb right now for my production system and currently play with sqlx and postgress for some side projects ✌️
I am a late bloomer and I am having a really rough time with rust because I am trying to run it on ANDROID Termux to be specific and I think being poor is the reason why I went this route but hopefully that will change soon, and I can migrate some of backend to rust