[pembrook]

The solution is to improve web app capabilities and security, not to devolve back to platform-specific fiefdoms just because there’s kinks to work out.

Disentangling apps from OS-specific environments is how we introduce competition back into computing.

Apple and Microsoft are the most valuable companies in the history of earth due to these fiefdoms and it’s about time we got some reversion to the mean.

[drdaeman]

“Web” as an OS is overbloated, with decades of legacy junk. It was never meant to run applications, and yet it does.

The strong indicator for this is that there are only, what, 3 (two sharing common ancestor, so more like 2.5?) non-dead rendering engines, and it’s practically impossible to implement a new one? (Yes, a toy engine is possible, but anything practical is out of reach.)

My point is, you can’t introduce competition over this platform - it’s the same sort of a walled garden, just with different owners.

[szasamasa]

why is a browser more overbloated than windows?

it was meant to run applications, that is why V8 is developed and optimized like hell, that is why we have webassembly to port c++ programs (try stockfish on lichess, it is webassembly)

originally, the browser was not meant to run applications, modern browsers are... strong indicator for this is chromium V8 performance and webassembly and the evolving js language with modules and node.js on server side

jvm introduced a competition to c++ and java is still top language, a browser is a kind of a jvm... you can use kotlin and other languages to compile to jvm and you can use js and webassembly to write apps on browsers

my point is, this is the best platform to introduce competition to native apps on walled gardens

the walled garden is not the os or browser but the company behind a closed os platform or closed single browser

on iOS, you had both: closed OS and closed single webkit browser engine

on every other os platforms you have true browser choice... since chromium is open source, you can create not just your toy browser but one of the most powerful ones... you can download the source code and if it goes in a restrictive direction, everybody can fork the web standard friendly versions

it is open and powerful, that is why most companies give a shit about browsers but it is not difficult for the EU for example to hire great browser engineers and create a beautiful stron free browser from an appropriate chromium version if things turned bad

whereas there are only 2 mobile operating systems that really work, from 2 giant US companies

where it is much more difficult to introduce competition is creating alternative mobile OSs... and then what, developers should create apps for this additional one two with another codebase?

browsers are the perfect level to introduce competition, in a way blink, webkit and gecko are all open source and chromium has a stron focus on improving web capabilities, pushed by 2 giants too (Google, Microsoft)

[szasamasa]

yes, there are sooo many OS out there

linux and windows, then 3x linux derivatives (android, mac, ios)

so we have the linux kernel and windows, we still exist (1% open source)

browsers have the open source chromium kernel plus safari and firefox and some small others

it seems to me browser choice is good enough

I dont mind if google and microsoft engineers all work on chromium together...

[drdaeman]

I'm sorry, but your comment has a lot of simple facts stated plain wrong, and then it entirely misses the point (which could be my fault for not making it clear what it was about).

macOS and iOS are based on Darwin, which is a BSD-derivative, IIRC there's nothing from Linux in there at all (although IIRC macOS used to ship some GNU programs). Chromium is not the engine, Blink and WebKit are. Android has almost no relation to GNU - sure, it has Linux kernel, but, it's entirely different system on top of it... but all of this is not important.

It doesn't matter if the browser choice is good enough or not. That's not the point at all - today's browsers are going to become irrelevant, just like '90s and 00's browsers don't matter anymore (Trident? Presto? CERN and NSCA stuff?). The fact there are only few engines remaining is the indicator, not the problem.

The core issue that the entry barrier is extremely high, and the foundations are still almost as bad as they were when Web2.0 was conceived. There's a giant baggage of legacy stuff that has pretty large impedance mismatch. The colossuses are growing so fast no one can realistically build a competing one anymore, yet all of them have feet of clay, as they were never meant/designed to grow into what they became through their twisted evolution.

By the way, I can totally understand if someone would disagree with this. Evolution takes weird paths, but somehow ends up achieving impressive results, and if we would stop caring about how things are working underneath (think of a recurrent laryngeal nerve analogy) and only focus on what they can do, maybe it's fine.

[szasamasa]

thanks for the details, you are probably right, then change linux to unix and chromium to blink in my comment, however, browser engines are all open source and they are the exact pitch perfect level to introduce competition

I am not so sure that chromium "kernel" or linux "kernel" ever become irrelevant... gecko or webkit maybe

even if only chromium engine stays, it is open source, if something goes wrong, you can fork or get back to gecko, webkit... we are all carbon based lifeforms, still there is great diversity I guess

why would we need to make so much effort to compete on browser engine level if it is so great that you cannot differentiate native to browser run programs?

I really do not understand what the point of your comments actually are... it is simply not true the browsers are not good for competition and I do not see what alternatives you suggest instead?