3 ms·

> As soon as some other client application is built which pushes user experience beyond what the browser is capable The problem is that the browser is so good

> As soon as some other client application is built which pushes user experience beyond what the browser is capable The problem is that the browser is so good at being a ubiquitous, open-source, open-standards-based, cross-platform, near-instant-loading secure remote code execution environment. As Brendan Eich says, always bet on JS and WASM. I would extend that to -- always bet on browsers. With the amount of change that has occurred in browser technology over the last 30 years, I strongly think the future looks like more improvement in browser technology (and perhaps deprecation/removal of old standards) than an entirely new paradigm. I know various early web pioneers want to build a post-Web application delivery system (notably Tim Berners-Lee and Douglas Crockford), but given the accumulated legacy of modern browsers and web apps and the ongoing investments into browser technology, I don't see how any post-Web system could gain any traction within the next 25 years. But yes - your point still stands. If anything better than browsers ever appears, the internet will indeed still be there to host it.

BrendanEich 3y ago
A couple of ways for a new "browser" to leapfrog: 1/ As guest in current browsers acting as hosts, for the zero-install ubiquity of the Web. This would require WebGPU, Wasm, fast JS<=>Wasm calls, and installable PWAs at least. Gatekeepers notably Apple can and will hinder. Once established, this new hosted app could be nativized to drop the browser dependency. 2/ As new obligate not facultative browser-like app in new device, e.g., XR eyewear. The device would have to take off at very large scale, and its maker(s) would want to do things not easily done via (1). The extant Web would be mapped onto in-world surfaces but the new device and its content stack would support things that go way beyond evolutionary improvements to the Web. I used to think (2) might happen with Oculus being acquired by Meta, or by Apple eyewear, but I'm less optimistic now. The Big Tech players, especially Apple, can build nice h/w, but it won't get to scale quickly enough to subsume the Web.
- TimTheTinker 3y ago
  > As guest in current browsers acting as hosts, for the zero-install ubiquity of the Web. That sounds promising, except I don't see how the affordance of URLs could be duplicated in that context without introducing competing janky proxies—e.g. https://new-web.com/[insert-new-url-type-here] https://new-web.com/[insert-new-url-type-here] Or were you thinking that installable PWAs would obviate that? I wonder if a wasm app framework ("a new way to build web apps without HTML/CSS/JS") would be viable, as it could work seamlessly with existing URLs and people could host apps built in it anywhere. If that gains traction, introduce a new, "cleaner" build target (sans HTML/CSS/JS linking) and a "faster, more lightweight" client for built apps. Have it use a new URL scheme to match - maybe wasm://. Web-targeted wasm app builds could even check if the URL scheme is available, and switch to it at runtime if it is. Kind of like how web apps sometimes open the ios app if it's installed. From there, expand into a platform for competing frameworks to target...
  - TimTheTinker 3y ago
    Endgame here is browsers run wasm:// applications (not a bad outcome), new client dies, and we still have web browsers. But maybe we get a more lightweight Electron.
    - BrendanEich 3y ago
      At high enough altitude, doesn't matter, but I was counting on PWAs. So Apple bad moves in Europe matter.
      - TimTheTinker 3y ago
        Some news dropped today, looks like PWAs on iOS in the EU are coming back: https://appleinsider.com/articles/24/03/01/apple-reverses-course-on-death-of-progressive-web-apps-in-eu https://appleinsider.com/articles/24/03/01/apple-reverses-co...
    - BrendanEich 3y ago
      One observation re: Electron: it was a security nightmare (no sandboxing! then lots of script injection / prototype pollution attack surface), so I think the better path to a host runtime is an extended browser. It has a lot of hard-fought security and privacy defense built in.

BrendanEich 3y ago

A couple of ways for a new "browser" to leapfrog: 1/ As guest in current browsers acting as hosts, for the zero-install ubiquity of the Web. This would require WebGPU, Wasm, fast JS<=>Wasm calls, and installable PWAs at least. Gatekeepers notably Apple can and will hinder. Once established, this new hosted app could be nativized to drop the browser dependency. 2/ As new obligate not facultative browser-like app in new device, e.g., XR eyewear. The device would have to take off at very large scale, and its maker(s) would want to do things not easily done via (1). The extant Web would be mapped onto in-world surfaces but the new device and its content stack would support things that go way beyond evolutionary improvements to the Web. I used to think (2) might happen with Oculus being acquired by Meta, or by Apple eyewear, but I'm less optimistic now. The Big Tech players, especially Apple, can build nice h/w, but it won't get to scale quickly enough to subsume the Web.

TimTheTinker 3y ago
> As guest in current browsers acting as hosts, for the zero-install ubiquity of the Web. That sounds promising, except I don't see how the affordance of URLs could be duplicated in that context without introducing competing janky proxies—e.g. https://new-web.com/[insert-new-url-type-here] https://new-web.com/[insert-new-url-type-here] Or were you thinking that installable PWAs would obviate that? I wonder if a wasm app framework ("a new way to build web apps without HTML/CSS/JS") would be viable, as it could work seamlessly with existing URLs and people could host apps built in it anywhere. If that gains traction, introduce a new, "cleaner" build target (sans HTML/CSS/JS linking) and a "faster, more lightweight" client for built apps. Have it use a new URL scheme to match - maybe wasm://. Web-targeted wasm app builds could even check if the URL scheme is available, and switch to it at runtime if it is. Kind of like how web apps sometimes open the ios app if it's installed. From there, expand into a platform for competing frameworks to target...
- TimTheTinker 3y ago
  Endgame here is browsers run wasm:// applications (not a bad outcome), new client dies, and we still have web browsers. But maybe we get a more lightweight Electron.
  - BrendanEich 3y ago
    At high enough altitude, doesn't matter, but I was counting on PWAs. So Apple bad moves in Europe matter.
    - TimTheTinker 3y ago
      Some news dropped today, looks like PWAs on iOS in the EU are coming back: https://appleinsider.com/articles/24/03/01/apple-reverses-course-on-death-of-progressive-web-apps-in-eu https://appleinsider.com/articles/24/03/01/apple-reverses-co...
  - BrendanEich 3y ago
    One observation re: Electron: it was a security nightmare (no sandboxing! then lots of script injection / prototype pollution attack surface), so I think the better path to a host runtime is an extended browser. It has a lot of hard-fought security and privacy defense built in.