Hacker News Reader

Comment by phendrenad2 | Hacker News Reader

I think there's a third option:

3. Firefox fades into obscurity/irrelevance, Chromium becomes the platform of the web, people realize this and forks abound, and eventually someone makes it into an ECMA standard and Google de facto loses control of it.

Oh also, another one:

4. The Linux Foundation or some other entity steps in and just builds a new GPL'd browser, which can mimic the success story of the Linux kernel (where the copyleft license will ensure that the vast number of open-source hackers prefer to contribute to it over something like the BSD-licensed Chromium).

View on HN

> The Linux Foundation or some other entity steps in and just builds a new GPL'd browser

The Linux Foundation was already given Servo (in late 2019 if memory is not failing me), the promising rendering engine Mozilla was working on. Which meant anyone who was working full time on it, is not right now.

So there's even a new base to work with from the ground up if we wanted to.

Servo was built and run more as a useful browser technique sandbox than a future standalone product.

Unfortunately, I think license matters here, and Servo is MPL.

A copyleft license like GPL attracts developers with more collectivist ideals, and developers who only want to contribute as a challenge or to make a name for themselves. BSD/MIT/MPL attract developers who want to make their own fork with proprietary additions.

Also the viral nature of the GPL causes people to put up a wall of separation around the project, and agree on where the boundary lies. Just look at the BSD projects, each is a blurry mess where kernel and userspace are intermixed. They're jumbled in one big repo. With Linux, the threat of license contamination has caused the modularity we see (my theory, anyway).

So the logical choice is LGPL then?

This is the ideal situation. Google already lacks control over all the Chromium forks, like Brave, Opera, Edge and ungoogled-chromium. They accept upstream patches voluntarily, and they can and do reject anything they disagree with. The shadier Google gets the more forks will reject upstream and instead share patches among themselves.

As long as Chromium is open source I don't see what there is to worry about. We literally have the code.

There's a lot to be gained by standardizing on Chromium as well. New features, improvements and optimizations only have to be implemented once instead of thrice, and developers only have to deal with bugs in one implementation instead of the union of bugs in all three.

This is fine; but still it hurts me to think that an open protocol of HTTP[S] and friends is effectively accessed by a single client at core. Yeah it's wrapped with different toppings by different vendors, but deep inside it's one client.

We're going from a world where we had IE, Google, Firefox and Safari as independent projects, we're going towards one with just one.

Microsoft couldn't build a viable browser that wasn't based on Chromium, so I don't have any faith that open-source can.

"If Microsoft can't do it, nobody can" is not an argument that holds a lot of water for me, especially if we're talking about web browsers. Microsoft has a track record of braindead decision making when it comes to their browsers. Couple that with their stubborn refusal to deprecate anything if it's a BC break and you've got an environment where not much can get done.

Case in point: https://docs.microsoft.com/en-us/internet-explorer/ie11-depl...

That article about IE 11 mentions about an IE 5 compatibility mode that can be activated. IE11 is essentially the sum of IE5, 6, 7, 8, 9, 10, and 11. So yeah: Microsoft couldn't maintain 7 browsers bundled into one, but who could? Moreover, who would? That's a terrible decision made completely from a place of fear ("what if that one customer that depends on IE5 compatibility mode gets really, really mad and drops their windows support contract!??!?!").

It'd be an absolutely massive undertaking, but it's not impossible. In fact, I'd go further and say that it's necessary.

I was referring more to their failure with Edge 1.0, where they did mostly jetison the legacy cruft and start a greenfield project to match Chrome and Safari. They threw a lot of effort into a mediocre product that almost no one used.

tbh, I forgot about that, but it wasn't exactly a green field project, right? They started with a fork of Trident. It'd be hard to know without an engineer from Microsoft chiming in, but it's entirely possible that the first bit of EdgeHTML was just Trident with all of the compatibility modes turned off. Maybe they removed them entirely, but just removing those piles of code doesn't fundamentally change the thing they hooked into, right? All the plumbing to support old cruft could still be there, which would make working on EdgeHTML almost as hard as working on Trident.

Mozilla incubated Servo with a team of five (as of ~2014). It's not a complete browser on its own by any means, but surely that points at some degree of possibility.

it was shown this week MS couldn't even ship an up-to-date version of curl with Windows

a browser is... more complicated

They could but I was probably a lot of duplicate effort when they could fork Chrome and brew in their own telemetry as opposed to Google's.

I like a variant of your third option:

3b. Chromium becomes the platform of the web and a fork becomes dominating through some unforeseen hype (without ECMA involvement)

> steps in and just builds a new GPL'd browser

I am not sure that can realistically be done anymore. Take a look at this for some estimates:

https://drewdevault.com/2020/03/18/Reckless-limitless-scope....

Reply on news.ycombinator.com