It is incredible that the community have built this general purpose platform, and that it's not only versatile but also incredibly performant.
Modern JS engines are only 50% the speed of native C, no compiler or toolkit needed. How insane is that!
CSS/HTML is growing from strength to strength, it's a gui toolkit with so much power universally supported on all computing platforms, from phones to desktops. It has so many features, and is so extendable, but anyone can use it anywhere. There's no limit to its capability, or who can use it to build their inventions.
The same toolkit can be used by small children to explore the ideas around "programming" all the way up to the best developers in the world.
Honestly the argument that we have somehow gone wrong with html/css/js is absurd. What this platform has contributed to humanity is immeasurable.
The "original vision" of the web as a publishing platform for text may be dead. But it lives on as a publishing platform for human invention, ingenuity and creativity.
The two places I'm most excited about it is using it in combination with WebGPU for games/engineering/ai/science, and with the "origin privet file system" enabling client side databases such as SQLite and DuckDB. I think it's the final peace of the jigsaw to make PWAs take off like a rocket.
Honestly, PWAs are the true future of the web, the perfect extension of its vision.
I know there's some boundary issues with performance and hope that it can be worked out in a meaningful/reasonable way. I think WASM gets much more interesting in server hosted platform environments. You get largely portable code that can run wherever (x86, arm, risc-v, etc) and is in a secure environment, read-only into itself.
Js used to be pretty bad as well, but hats down, they did alleviate many of the shortcomings and made it into a decent language (though I do get grumpy when it is overused at places where it has no reason to be)
WinForms may be a great GUI toolkit, but it isn't as versatile or accessible as the web platform.
If you only ever use a tool the way it was intended how can you make it better?
> If you only ever use a tool the way it was intended how can you make it better?
I don’t think that it’s how things work.. you don’t have to misuse something to think ahead and make something better/develop it into better directions.
Sounds like: "We're proud of the craziest kludge in history. Look mom, DHTML!"
One could even write a full parody of one of the previous posts. (Sorry, SCNR):
"""
The JVM platform, Java in a browser, is one of the greatest inventions of all time. Its versatility is immeasurable. The fact we have one toolkit that scales from a simple text based platform to full application development environment, is insane. It covers all paradigms of computing, from thin client to full offline enabled local computing.
It is incredible that the community have built this general purpose platform, and that it's not only versatile but also incredibly performant.
Modern Java interpreters are only 50% the speed of native C, no compiler or toolkit needed. How insane is that!
JavaFX/CSS is growing from strength to strength, it's a gui toolkit with so much power universally supported on all computing platforms, from phones to desktops. It has so many features, and is so extendable, but anyone can use it anywhere. There's no limit to its capability, or who can use it to build their inventions.
The same toolkit can be used by small children to explore the ideas around "programming" all the way up to the best developers in the world.
Honestly the argument that we have somehow gone wrong with Java is absurd. What this platform has contributed to humanity is immeasurable.
The "original vision" of Java as a platform only for embedded devices may be dead. But it lives on as a platform for human invention, ingenuity and creativity.
"""
It’s been 30 years, it’s natural for popular sites to be complicated.
And while using the browser as an OS inside the OS has some benefits, it's even worse than Java in the sense that it's hard to build a good program when you're constantly fighting with the browser for the interface... (consider how rare it is to see a webapp with lots of keyboard shortcuts and support for multiple windows for instance)
HTTP/3 is definitely still HTTP. QUIC is replacing TCP.
Requiring CAs and not implementing support for other anchors of trust is an implementation decision, and is not mandated by the spec. The spec mandates TLS 1.3, not "the version of TLS 1.3 used by these three web browsers". QUIC was designed with non-browser use in mind too; it's for any situation where you want to maintain connection integrity in an unreliable network.
Like I said, the megacorps that designed QUIC designed it for their use cases and that does not include allowing human people to host websites by themselves. Not if it potentially impacts e-commerce security. Their implementations reflect this.
Note: and the fact that there is no search engine inside companies has led them to disregard the concept of permalinks. All SPAs there simply never change the URL in the address bar. Which is an HORRIBLE experience as a user!
That's a great point. You wouldn't believe the hemming and hawing that ensues at work when teams encounter SEO-driven constraints.