Why not if everybody at the end will implement the same spec? I would understand if Firefox wanted to implement its own spec, but what is the purpose of having N different implementation of the same spec with their own idiosyncrasies? At the end of the day, the engine of the major browser engines is open-source anyway.
Sorry, I don't know how I missed day zero-day.. Anyway my point still stays..
I don't think that zero-day is really an argument, given that the vast majority of users are on Chromium. If there is a zero-day on Chromium, most people have it.
> At the end of the day, the engine of the major browser engines is open-source anyway.
Open source is not enough. The question is: who controls it? AOSP is open source, Chromium is open source. But Google controls both. It means that Google can push for what is good for Google... even if it is bad for the user. E.g. preventing users from blocking ads. Not that it does not have to be with evil purposes (though Google has been shown to be evil enough already): it's enough for Google not to care about something for it to impact Chromium/AOSP.
That's the whole point of competition: you want the users to have choice, so that it pushes the companies towards building a better product. Monopolies never serve the users.
Now you say: "ok but it's open source, so if you're not happy you can fork away!" -> which precisely brings us to two browsers, like now with Chromium and Firefox.
What if the developers of the dominate engine becomes complacent and decides that it's "good enough" and we get stuck in another IE6 situation where development stops for years and years?
Yes, Chromium and Blink are open source, but they are effectively Googles open source project. If you're unhappy with their direction you'll need to fork it.
Exactly, the situation with Chrome is the exact same sort of benevolent dictator problem we had with Internet Explorer, except this time dressed up with an open source license.
Mozilla often disagrees with Google on what should get into web standards and the design of the spec, especially apis that give hardware access or seem to make privacy harder for the user. Having their own implementation is kinda crucial for that.
Look at how llvm forced gcc to improve their error messages (among other things).
Running a different compiler is also useful to find bugs in your project, and in the compiler itself. I would imagine this applies to the web just as well — a web browser implements an open spec (just like a C++ compiler), at the same time being much more complicated than a C++ compiler.
If I remember correctly, RFCs need at least two independent implementations to become standards. I think that would be a good idea for web stuff too. It's a way to make sure the spec isn't just blindly following the implementation.
