It is a matter of perspective. For my personal projects I decided some time ago that I would use javascript both frontend and backend. That was a world of pain, mostly attributed to the lack of static typing, asynchronous nature implemented through callbacks, and the pain of choosing alternatives in the chaotic ecosystem. It was somewhat alleviated when I switched to Typescript, started to use async and await, and settled in some choices (Bootstrap, React, webpack...). But then I tried the golang. Yes, the lack of generics hurts a bit, but otherwise it is an absolute damn pleasure to use: static typing, reliable completion in IDEs, batteries included, more clarity in choice of modules. It just felt right. Granted, it is my multi-year background in C and C++. For me multi-threading and co-routines just feel more natural way of doing concurrency than callbacks and CPS.
> In my opinion the super hype over Javascript is because of it's simplicity(compared to GO for example).
IMO, the core of modern ES6+ JavaScript is a really nice scripting language that is simpler than alternatives, but doing production JavaScript means contending with all of the other stuff. To be a working JavaScript developer, you need to know multiple ways of doing the same things, from basics like module loading and HTTP requests to UI and concurrency. Legacy and an ecosystem that is fragmented down to the standard library level make JS not the easy option.
> In my opinion the super hype over Javascript is because of it's simplicity(compared to GO for example).
Go imho is a lot simpler than Javascript. Not necessarily as a language (although Go is pretty simple, and JS has some very ugly pitfalls). Thing is, getting into the JS ecosystem is an absolute mess. With Go, I felt comfortable with it within a few days. Recently, wanting to create some sort of frontend for a REST API I'm making - I tried getting into "modern" frontend JS, having dabbled in HTML, JS and CSS a decade ago, having created a small internal bootstrap-based app targeted at mobile devices - so I understand the basics
However, the current JS ecosystem? I have no idea where to even start, I haven't even gotten to the point where I would actually have some hacky code that would at least give me an idea how to move forward...
First, it makes wonderful sense for certain projects even without productivity/learning-curve gains. E.g., anything that's not particularly supported by the HTML+CSS paradigm. Think "stuff that would today be a native application".
But let's jump into the "finding labor/business case" problem.
> It is already very hard to find a decent backend developer
And why is that?
> In my opinion the super hype over Javascript is because of it's simplicity(compared to GO for example).
Agreed! You totally hit the nail on the head! The reason frontend is easier to hire for, is because HTML+CSS+Javascript provides a relatively small minimal viable set of abstractions that allow you to be productive.
Of course, I don't want to diminish the fact that there's an infinite rabbit hole of complexity here (see e.g., wavethrough); serious engineers can definitely thrive in front-end work. But for "productive enough frontend dev", the set of things you have to learn is relatively small compared to backend.
> if the frontend is written in backend language like GO, it kind of makes it worse.
On the contrary, I think WASM could enable a fix to your problem!
The question is, could we build a WASM-backed unified development environment that allows people to learn full-stack development in the same amount of time that it currently takes to learn just the front-end stuff? And can you produce productive full-stack engineers for a particular full-stack framework while skipping a lot of the fundamental CS stuff?
If the answer to both is "yes", then coding bootcamps could step into that environment that churn out lots of relatively cheap/easy-to-find full stack development talent!
So. Who can build a single framework providing a small set of abstractions (learnable in 6-12 weeks from scratch) that allows for both front-end and back-end development? If someone can do that, and build an educational commodity around it (i.e., a scalable model for bootcamps), then I could see wasm enabling a whole new approach toward web development in which "not top-tier but definitely good enough" fullstack developers are at least as easy to find and for the same price as today's "not top-tier but definitely good enough" frontend developers.
> Who can build a single framework providing a small set of abstractions (learnable in 6-12 weeks from scratch) that allows for both front-end and back-end development?
Remeber the desktop RAD platforms, like Delphi and Visual Basic? There is no reason we couldn't have very similar things on the web.
https://anvil.works - it's explicitly modeled on VB6/Delphi. All Python (front and back end, with a simple RPC interface from client to server), and learnable in a lot less than 6 weeks. Check it out!
0%. I work as a upwork freelancer. More precisely by hourly rate and upwork desktop app takes screenshots at random times, so i can't really browse HN and other sites. But I will tell you honestly, i don't want to go back to normal work. I get shit done( there is also that "gambler" feeling, when you see how much money you earn each week ) and in good quality( I get reviews from my clients, so i can't code bad either ).
Uber - Yandex(russian alternative) etc.