True, but remember when many apps were not made for Linux, because it was not worth it? With Electron at least we have apps like VSCode for Linux too. Not ideal, but better than nothing.
My point is that we shouldn’t embrace for Electron just because we want applications in Linux.
In addition, Election is not a guarantee of a Linux port either.
In any case, I personally embrace Electron due to its development value. I can use the same codebase on all platforms, using technologies that my colleagues and I are already very familiar with (TypeScript/React/etc), with a supporting cast of a mind-shatteringly large ecosystem of FOSS packages. The speed at which I can turn out features in Electron just vastly outpaces anything else we've worked with. And that seems to be worth more to users than the 200MB footprint.
A shitty version, you mean. Same as Office 365 vs. the native Office.
The web versions of most native apps are very, very bad at the moment.
> I can use the same codebase on all platforms, using technologies that my colleagues and I are already very familiar with (TypeScript/React/etc), with a supporting cast of a mind-shatteringly large ecosystem of FOSS packages.
Everything you wrote here applies to most languages, including .NET, Java, C++, Rust and others.
The fact that you and your colleagues know other technologies does not mean they are the right approach.
> The speed at which I can turn out features in Electron
I think that is exactly what users don’t want. Users want stability and quality, not feature churn.
Indeed, hence the parenthetical.
> Everything you wrote here applies to most languages, including .NET, Java, C++, Rust and others.
Superb! Direct me to a rich Rust GUI application which targets the web, Linux, and iOS. Speaking to team tech familiarity, our front-end web engineers will be thrilled to get cracking with Rust. That'll be a sight!
> I think that is exactly what users don’t want. Users want stability and quality, not feature churn.
I'm glad you know what users want. I generally prefer them to tell us, and incidentally, "feature churn" never comes up in such discussions. What user stories do come up can be completed remarkably quickly in our Electron stack with stability and quality too.
I mean, we do already. It's called PWAs, and I have many of my favorite sites saved as an app. But it's featureless and lame.
Point being, Electron is just a stopgap while PWAs are figured out. And with the advent of WASM, we'll be able to run some pretty heavy stuff.
https://wasm.continuation-labs.com/d3demo/
My long term strategy: Learn rust for WASM, keep an eye on PWAs
The browsers also lack many APIs to compete with native applications, and those are not coming soon either.
Web apps are just fine, even nowadays, for many things. But for heavy lifting, no, they are still way, way behind.
(By the way, you can compile to Wasm with almost any popular language, not just Rust).
and yet on Linux you had already the entirety of debian already available in native ARM64 years ago, which is a lot more than what you have on windows ARM64.
Though, it is usually easy to rewrap the version from another platform using Electron on Linux. E.g. writing a Nix derivation for Notion on Linux isn't much more work than for any other Electron application:
https://github.com/danieldk/nix-packages/blob/master/pkgs/no...
(Not tested more recently though, I don't use Notion a lot.)
Add Figma to the mix. Since I moved to Linux exclusively, I've abandoned plugin authoring. Granted, there is a Linux-Figma[0] app, but every time I try it, it doesn't take long before something essential (e.g. fonts) breaks.