I understand that argument but 100% disagree. The web is the cross platform that runs everywhere (as in in a browser or webview). True it has many issues and some forms of it (electron) are not ideal for some use case but I believe it’s a better platform than flutter in almost every way.
You say “true” cross platform and could argue that the web isn’t “true” cross platform. But really is anything?
Of course this is only relevant if that deeper plumbing into the OS is needed, and of course it's quickly changing. Perhaps the day comes when web apps can truly be first class citizens in all major OSes, but it ain't here yet.
So if your claim is simply that CSS/HTML/JS is capable of creating good user interfaces, I'm not arguing against it.
bruh it's not just UI, the JS/TS code is where the business logic code is. And anyway the UI is the whole point, we are debating UI libraries. But most importantly native plugins are niche, you use them very rarely and for very specific needs, they occupy generally less than 1-5% of the code base. e.g. when you need to access the camera or sensors, which you usually don't need and most apps don't actually benefit from plugins. I'm not even talking about the fact that most of those niche needs are generally covered by equivalent modern web APIs (e.g. web share) but since plugins are trivial to use, people often use them even when a web api already exist.
However, I believe it might not matter. 95% apps can be done as web apps today, especially what comes to boring business applications. 5% we can wait for some time.
What comes to Flutter, it can never integrate well within a browser as it sufferes from the same issue as Java applets. It does not use the web browser native rendering engine and widgets will always feel kludgy, have accessibility issues and so on. Not a big deal for some apps like games. And Google Docs is already using Canvas based rendering so seems like Flutter canvas based apps might have hope.
Well, isn’t it the Java applets model as you call it as well? But is a colored div that much better of a button?
Though I don’t disagree with you, indeed most apps can be web-based.
However given the app store is a cash cow for Apple I highly doubt they’d want to integrate web more deeply.
Local storage already does a lot of needed data storage (though apple has recently made this complicated with their auto delete after certain period). Local file access for uploads is also now possible on iOS using the Files app.
Nope, The web is The Platform. And while it offers numerous benefits it has plenty of flaws / inability to do some things as well.
I understand that argument but 100% disagree. Development is horrible, it's anything but standard, comes with a lot of baggage, the performance is atrocious. It doesn't even run anywhere, it runs in a browser. A huge, bloated, hungry browser.
You say Electron is not ideal, I say it's the worst thing that had happened to desktop computers ever. It's bad for the planet too. Some argue that without "js on desktop" we wouldn't have these amazing apps. I think it's not worth it to have those apps if they are built using those technologies. They make me miserable every single day. It's the only thing that forces me to upgrade my computer. It's so bad. Things that were possible 20 years ago are suddenly an unachievable goal. Stop it.
I am not saying that electron-based apps are generally good or efficient. Most of them are frustratingly slow and awkward. But this inefficiency rarely seems to be constrained by local resources: the apps often stand still, consuming barely any resources (CPU time, memory, etc.) and taking ages to load each page. At the same time you can often visit the same page in Chrome and it will load almost instantly.
It is frustrating and I have no idea what is going on, but it doesn't seem like upgrading the PC would solve the issue.
Yes, but the web is crappy system for designing UIs in.
As a community we've got good at designing UIs with web technologies (HTML+CSS), because we have no other choice if you want to build web-apps. But the level of experience we've gained with those tools has hidden the fact that they have some real issues - mostly due to the fact that they've developed piecemeal over many years. Getting the layout of your page to behave correctly under all the circumstances it needs to requires a lot of experience with CSS, a lot of simple page behaviour we'd now expect as users require adding JS sprinkled around (and I'm not talking app logic, just UI logic).
I freely admit that I'm not great with CSS - I'm not a web developer although I have done some web development. If I do any at work (e.g. web UI for embedded device) I need a pro to come along and make things look nice. But when I did app development in Flutter it took about 2 days to be able to build screens that looked good, did complicated dynamic behaviour, and behaved reliably regardless of screen size and shape.
For one, you mention Electron. That targets all desktop platforms, but not mobile. Mobile development from webapps is absolutely possible, but you need to use something else for that. Meanwhile, on the flutter side of things, if you want to target desktop platforms, you use flutter. If you want to target mobile platforms, you use flutter. There isn't a need to learn a whole new framework to target a new platform.
Aside from that, local file storage and offline use can also be a bit finnicky with Electron. Absolutely possible, but its another thing you have to figure out that Flutter just does by default.
If I'm wrong on any of this, please do let me know, I know web dev far better than I know Flutter so it'd make my life a lot easier, but the above reasons are why flutter hasn't budged from my todo list