That's two opinions, and they have nothing to do with each other.
As it happens, I am strongly biased to agree with your first point: I find working in the FE JS ecosystem to be intolerably complex with very little upside. I was so powerfully motivated by this belief that I've spent the last 18ish months working with a group of people to build StimulusReflex. SR is a library for Rails developers that makes it easy to build the same sorts of snappy, reactive UIs people think can only be done with SPAs. We keep all of the application state on the server, and send updates to the DOM via websockets. It's freaking awesome, and everyone should be getting on the train.
Fortunately or unfortunately, I am also strongly biased to thinking that building a web framework is (probably) a terrible idea, unless you're doing a PhD or someone is paying you $$$$$. The reason is that unless you're bringing something truly radical to the table, the time you sink into building infrastructure that already exists many times over is nothing more than procrastination and letting your self-doubt drive you towards small outcomes.
Look, I'm not a therapist and I say the above with only sincere compassion. Instead of building a platform, build a solution to a well-researched problem with customers who have a budget. You wouldn't be the first person to get sucked into the "well, I'd build a site but I need a framework so I'll design a computer that can run the OS I make with the language I'll have to design"... but you would be just like TempleOS.
@leastbad, thanks for chiming in and reading those pieces!
Yeah, the server-side state approach from your work on StimulusReflex, and also Laravel Livewire and Phoenix LiveView in other communities have been bringing some new things to the non-JS community. DerbyJS sort of introduced in the JS ecosystem but didn't get much traction (in part I think b/c the execution is wonky and JS tends towards SPAs rather than server-side state).
Look, I'm not a therapist and I say the above with only sincere compassion. Instead of building a platform, build a solution to a well-researched problem with customers who have a budget.
You're certainly right if my goal was to start a business from a web-framework. I don't really think that an open source web framework is the way to accomplish that. And, it's a huge time waste to build a framework in order to build a SaaS or whatever.
Fortunately or unfortunately, I am also strongly biased to thinking that building a web framework is (probably) a terrible idea, unless you're doing a PhD or someone is paying you $$$$$.
Right now, I'm thinking of it more as a way to learn more of TypeScript, address what I see as a gap in the Node/TS/JS ecosystem, and work towards building an audience of developers. It's still quite possible that it's a terrible idea but that's how I'm thinking about it now.
This reflects a lot of my thinking that I've shared with IH in a couple ways:
Unpopular Opinion: Node.js + [insert SPA framework] is the wrong choice for my next Indie Hacker project
I'm writing a web framework
I honestly think React, Angular, Vue, Svelte, etc. is most likely the wrong choice for most IH projects ... but I'm happy if you disagree with me :)
That's two opinions, and they have nothing to do with each other.
As it happens, I am strongly biased to agree with your first point: I find working in the FE JS ecosystem to be intolerably complex with very little upside. I was so powerfully motivated by this belief that I've spent the last 18ish months working with a group of people to build StimulusReflex. SR is a library for Rails developers that makes it easy to build the same sorts of snappy, reactive UIs people think can only be done with SPAs. We keep all of the application state on the server, and send updates to the DOM via websockets. It's freaking awesome, and everyone should be getting on the train.
Fortunately or unfortunately, I am also strongly biased to thinking that building a web framework is (probably) a terrible idea, unless you're doing a PhD or someone is paying you $$$$$. The reason is that unless you're bringing something truly radical to the table, the time you sink into building infrastructure that already exists many times over is nothing more than procrastination and letting your self-doubt drive you towards small outcomes.
Look, I'm not a therapist and I say the above with only sincere compassion. Instead of building a platform, build a solution to a well-researched problem with customers who have a budget. You wouldn't be the first person to get sucked into the "well, I'd build a site but I need a framework so I'll design a computer that can run the OS I make with the language I'll have to design"... but you would be just like TempleOS.
Don't be like TempleOS.
@leastbad, thanks for chiming in and reading those pieces!
Yeah, the server-side state approach from your work on StimulusReflex, and also Laravel Livewire and Phoenix LiveView in other communities have been bringing some new things to the non-JS community. DerbyJS sort of introduced in the JS ecosystem but didn't get much traction (in part I think b/c the execution is wonky and JS tends towards SPAs rather than server-side state).
You're certainly right if my goal was to start a business from a web-framework. I don't really think that an open source web framework is the way to accomplish that. And, it's a huge time waste to build a framework in order to build a SaaS or whatever.
Right now, I'm thinking of it more as a way to learn more of TypeScript, address what I see as a gap in the Node/TS/JS ecosystem, and work towards building an audience of developers. It's still quite possible that it's a terrible idea but that's how I'm thinking about it now.
This comment was deleted a year ago.