That is, like, genuinely an advantage, though. At $ DAYJOB, we have a project that spans embedded, backend, web frontend and CLI, and for all of these, Rust is decent.
Like, I can see why a frontend dev would want to use HTML+CSS+JS/TS (rather than HTML+CSS+Rust), mainly because the massive ecosystem of JS components makes you more productive.
But you pretty much won’t ever develop a web frontend without an accompanying backend, and then being able to use the same language-expertise, libraries, utility functions and model types, that is also a big boost to productivity, especially if you won’t have a dedicated frontend dev anyways.
Realizing that also made me understand why people subject themselves to NodeJS for their backend, which has the same advantage, just with the big ecosystem in the frontend and the small ecosystem in the backend.
because the massive ecosystem of JS components makes you more productive.
Slightly less ironic: I question even this right now (as I have to suffer from endless “hot”-reloading and browser-crashes because of Next.js bloat).
I think the massive ecosystem has fewer high quality libraries than Rust at this point.
I use both JS/TS in frontend and Rust (either frontend more as a hobby and backend) extensively, and I very often check the dependencies-source, and even more often rewrite it (unfortunately not in Rust), because of low-quality.
And it’s sooo slow… the tooling and the frontend (albeit I think that has a very lot to do with next.js… and with how easy it is to make it slow for someone not that experienced or someone not being extremely careful).
Frontend is not yet as matured as JS/TS (whatever matured is, but the count of frontend frameworks is at least a magnitude higher in JS/TS), but I think when I would start a new company I would default to Rust now as frontend indeed, the language itself is for me reason. And I think vanilla-js (or Rust?) is not that much worse (time/effort-wise, sanity etc.) for more complex applications than what the Next.js ecosystem has produced so far.
That is, like, genuinely an advantage, though. At $ DAYJOB, we have a project that spans embedded, backend, web frontend and CLI, and for all of these, Rust is decent.
Like, I can see why a frontend dev would want to use HTML+CSS+JS/TS (rather than HTML+CSS+Rust), mainly because the massive ecosystem of JS components makes you more productive.
But you pretty much won’t ever develop a web frontend without an accompanying backend, and then being able to use the same language-expertise, libraries, utility functions and model types, that is also a big boost to productivity, especially if you won’t have a dedicated frontend dev anyways.
Realizing that also made me understand why people subject themselves to NodeJS for their backend, which has the same advantage, just with the big ecosystem in the frontend and the small ecosystem in the backend.
Slightly less ironic: I question even this right now (as I have to suffer from endless “hot”-reloading and browser-crashes because of Next.js bloat).
I think the massive ecosystem has fewer high quality libraries than Rust at this point. I use both JS/TS in frontend and Rust (either frontend more as a hobby and backend) extensively, and I very often check the dependencies-source, and even more often rewrite it (unfortunately not in Rust), because of low-quality. And it’s sooo slow… the tooling and the frontend (albeit I think that has a very lot to do with next.js… and with how easy it is to make it slow for someone not that experienced or someone not being extremely careful).
Frontend is not yet as matured as JS/TS (whatever matured is, but the count of frontend frameworks is at least a magnitude higher in JS/TS), but I think when I would start a new company I would default to Rust now as frontend indeed, the language itself is for me reason. And I think vanilla-js (or Rust?) is not that much worse (time/effort-wise, sanity etc.) for more complex applications than what the Next.js ecosystem has produced so far.