Here's what I'm aiming to do: learn systems programming on the side, accumulate enough savings to live off of for nearly a year, then go independent and continue learning systems programming "as I go". Build software products on my own and start selling them.
And because of this generic term, I see people making stupid mistakes all the time.
React is great for web apps, but then you see the static pages being built by that, and you think WTF. Next thing some of these morons move back to server side rendering, discover how much better it is for SEO etc. But the problem is that the web-app folks read this, think they now all of a sudden have to move to SSR because big company X did this for their landing page.
It's truely stupid. So I think we should split the term "web development" into proper subsections, so it becomes clear what camp you are in.
If you're a website builder who does a bit of JS for some menu popups, good luck having a team of these guys writing the next google docs (no offense here, each their own specialization) But on the same ground, having a bunch of pro front-end coders build a simple website is also stupid, because they will choose technologies way too complex for that.
You will get paid more if you are a webpack ninja than if you know how to setup a simple esbuild script.
You will get paid more if you have wizard level knowledge of k8s and docker than if you can program something that solves your problems as a simple single program running on a single server.
And it's all portable. No wonder Electron has eaten the native app scene.
But the JS ecosystem with its myriad frameworks, packers, metalanguages, each halfway between immature and deprecated, and constant churn and breakage is an absolute nightmare.
Also, the browser does not have an out-of-the box set of standard UI widgets, which makes every browser-based (Electron) app redefine them in its unique way.
Because it's really difficult.
I find I can keep a big-picture, this is what the customer wants mindset, or I can have a detail-oriented this is the code that I want to improve mindset, but having both and/or switching between them is amazingly difficult. As I write the code I lose sight of what 'normal' people (non-programmers) want. I don't even know how to pitch to their level any more, or what isn't obvious to people who haven't spent 20+ years using and thinking about computers every day.
tl;dr: programming changes the way I think in a way that's unhelpful to build a product.
Next there's promotion and marketing... which is another skillset that depending how hard you want to push the product may/may not sit well with your ethics.
Thirdly and this is a personal one, I have an ADD brain and staying focused on one idea long enough to take it to market is very difficult. I have a list of ideas I dreamed up and started work on and never finished that someone else later launched successfully. As does everyone I suspect (execution >> ideas) but it's a neverending pattern (as I look at the 30+ OS projects on the go)
What I have trouble with is marketing and sales. I've never sold anything directly to consumers, so that will be the most challenging thing when going independent: selling to people.
Let's say I sell the product for $50, I would have to sell to 2000 people in one year to make $100k, which means I have to sell to at least 6 people per day. Now it would be great if people just found the product and bought it on their own, but at least the first 1000 costumers must be acquired "manually" so to speak for the business to even take off.