Hacker News Reader

Comment by roenxi | Hacker News Reader

I'm not sold that avoiding Chromium is any sort of win. It seems to be BSD licensed. The code comes from a long and storied OSS tradition.

It is unclear why that having multiple different browser engines is strategically important. As far as I know we've long left behind the world where a hobbyist project could reasonably expect to implement modern JavaScript. Brave's existence and success suggests that competitors building on Chromium are viable.

Not to say it is a bad thing to have alternatives floating around; but at this point it is probably easier to have multiple forks of Chromium than strive to have code with a completely independent genealogy.

View on HN

IMO the critical factor will be whether any Chromium forks keep support for Manifest V2 alive. If yes, it will be a viable alternative to Firefox (or FF forks), but I haven't seen indication that a project will pick up the slack.

A browser without proper support for uBlock Origin or similar extensions isn't a browser I'm willing to use, even if ad-blocking is integrated into the browser itself. The freedom to actually control the browsing experience is essential for long-term user freedom.

It'll be gone the moment upstream drops it. Maintaining this in a fork will be a lot of work. Fortunately it's still available even in Chrome for a while, but that's just an ultimatum

That's my expectation too, but it makes avoiding Chrome & Chromium unavoidable.

Even if maintaining Manifest V2 is too much work after upstream drops it that doesn't necessarily mean that they just have to accept Manifest V3 as shipped in upstream.

From what I've read there are two main problems with Manifest V3 when it comes to ad blocking.

Problem #1. The webRequestBlocking permission is gone:

> Note: As of Manifest V3, the "webRequestBlocking" permission is no longer available for most extensions. Consider "declarativeNetRequest", which enables use the declarativeNetRequest API. Aside from "webRequestBlocking", the webRequest API is unchanged and available for normal use. Policy installed extensions can continue to use "webRequestBlocking".

That does not necessarily mean good ad blocking must break. Most people find the ad blockers on Safari to be fine, and Safari does not have that kind of blocking. It has long required the declarative rule kind of blocking.

Some brief searching suggests that the rules for Chrome's declarative blocking are actually more powerful than Safari's.

That brings us to the second Manifest V3 problem.

Problem #2. The declarative rules only allows something like 30k rules. Apple found that such a low limit is not sufficient and raised it to 150k.

I'd guess that maintaining a fork that just raises the limit in Chrome would not be a problem for the Chromium-based third party browsers.

The big question would be whether the current 30k limit is there for technical reasons or for marketing reasons like not wanting ad blocking to be too effective. If it is a technical limit the forkers might also have to rewrite the rule processing engine which would likely be more than they would want to deal with.

[1] https://developer.chrome.com/docs/extensions/reference/api/w...

> The big question would be whether the current 30k limit is there for technical reasons or for marketing reasons like not wanting ad blocking to be too effective. If it is a technical limit the forkers might also have to rewrite the rule processing engine which would likely be more than they would want to deal with.

It’s Google. The whole Manifest is there for marketing reasons. The whole browser even. Google doesn’t want us to have a nice browser because it benefits us. It benefits them. And this will continue. With V3 they already did the biggest cut down of af blocking in their browser. They will continue this path.

Their main income is ads. They don’t care for anything else.

> That does not necessarily mean good ad blocking must break. Most people find the ad blockers on Safari to be fine, and Safari does not have that kind of blocking. It has long required the declarative rule kind of blocking.

This presupposes that advertising platforms will not react to the changes to adblocking in any way, but I don't think that's realistic.

The declarative rules can only be changed with updates to the extension (AFAIK, please correct me if I'm wrong). This means users can't block ads themselves (at least uBlock Origin Lite can't do it), and it will take between days to weeks until new block lists have propagated.

Now, remember last year when YouTube started blocking adblockers? They changed their implementations sometimes multiple times per day, and after minutes to hours new blocklists were distributed that got around them. This won't be possible anymore with Manifest V3, so Google just has to update their adblock implementation often enough to finally get rid of them.

And the same goes for other platforms. Once the majority of users is using browsers that only allow declarative blocking, more platforms will start updating more frequently.

The problem with Chromium is that it's controlled by Google. Given they control the engine in the majority of browsers, they get to call the shots on the direction of the web as a whole.

They're not in any position to control the web as a whole; Safari is too strong a competitor.

But even if Chromium was at 100% market share they couldn't control the web. If they start throwing their weight around then the other web companies will start organising around a Chromium fork. The important thing here is OSS/not-OSS rather than market share. Market share can change quickly enough in response to problems. It is like GCC - there was that spat back in the day, someone forks the code and there is a little war until people figure out which is better. Then the better version became "GCC".

It is fine to have one custodian as long as the actual power is retained by the users. Efficient, even.

What "other web companies"? Everyone has an app or an app store.

Whoever. Whatever. OSS is a system for finding the most capable entity in the world at developing a software capability and letting them get on with it. It'll find someone to pick up the slack if Google starts doing a terrible job. In this world of 8 billion humans and however many thousands of companies, there'll be one who knows how to put a web browser together.

If I were going to guess today it'd be brave.com, but the world is large and competitive.

I think that's a tad idealistic, though. You'd need the resources to keep fighting against the constant changes and ever diverging codebase the more Google goes it's own way. And already has been doing it slowly. Also, Mozilla and Apple have been important for keeping privacy relevant every time Google have tried to push for some privacy eroding web standards.

Edit: I agree with the sentiment though but I just trust the current situation more.

> You'd need the resources to keep fighting against the constant changes and ever diverging codebase

Assuming that Google is making Chromium better, sure. If they are making it worse then not keeping up with the changes is an advantage of the fork.

If Google starts making changes that reduce the value of the codebase, then a hard fork immediately becomes viable. If they are good custodians even a totally separate codebase isn't enough, it'll end up in the same position as Firefox - marginalised.

This is very handwave-y and idealistic. Sure, 8 billion humans, and thousands of companies on a cosmic timeline will lift the "most capable entity" to the top, but the road there is going to be extremely long if you just "nah they'll get it" and not do anything else.

I named a specific company, pointed out a specific example of the dynamic featuring GCC and Debian is just one iceberg in the OSS ocean demonstrating that there are more than 30,000 packages built on the principle that free software is better than closed source.

I'm not sure how much more specific and concrete you want the argument to get, but that seems fine to me. The market is big, people care about this part of it. There will be a sponsor for the work if Google isn't up to the challenge.

The only reason everyone uses Chrome is it is really good at what it does and to date Google has been a high quality steward.

Isn’t the push to deprecate and remove Manifest V2 in favor of V3 and therefore removing or at least significantly cutting down possibilities for users to block ads a major push from the Google side? A field in which they, conveniently, make their major cut of their profit? Brave is a big company already, still they’ve already announced that they won’t be able to support V2 for too long as the codebase will break over time.

Only because something is OSS doesn’t mean it’s not controlled by a single entity or consortium. We need competition, especially in something so damn critical like web browsers (which are like OSes for the web).

Mozilla is and has been a savior. They’re doing weird shit, but compared to Google and Apple they’re still as transparent as possible, even with their shit. And their shit is not even THAT bad, while I still dislike it and also thought about replacing FF for a short time. I just hope they come back to the right path, at least in their communications.

Does Google effectively own Webkit, though? This isn't a rhetorical question, I'm really asking. Whether or not Webkit can be thought of as disjoint from Chromium is really the question here.

WebKit is substantially disjointed from Blink at this point, a natural result of differing priorities from their developers.

WebKit remains highly embeddable for example, so if you drop a WebKit webview into your Cocoa/GTK/whatever toolkit app you’ve got a mostly-complete multiprocess browser and just need to build your UI around it. Blink on the other hand is hard-coupled with Chromium for things like multiprocess support, which leaves no practical alternative to forking the whole browser (which as an aside, is why Electron drags around the full heft of Chromium instead of just Blink).

That’s moot though because Google has no qualms with pushing through Chromium-only “standards” that make non-Chromium browsers less viable since sites that use these “standards” simply don’t work in anything but Chromium. It’s like the situation with Internet Explorer and its collection of IE-isms all over again.

One could argue that Mozilla and Apple should just follow Google and implement these “standards” themselves, but that lands us right back where we started in that this effectively gives Google full unbridled power to steer the direction of the web as it pleases, which is a massive conflict of interest.

> Does Google effectively own Webkit, though?

No, Google owns Blink: https://en.wikipedia.org/wiki/Blink_(browser_engine)

which outright forked in 2013, though at that point it sounds like they were already carrying enough patches to constitute a soft fork (which then became a hard fork).

> As far as I know we've long left behind the world where a hobbyist project could reasonably expect to implement modern JavaScript.

You should look into Ladybird[0] then. Their JavaScript engine ("LibJS") scores very well in conformance test suite test262[1].

[0]: https://ladybird.org/

[1]: https://test262.fyi/

>modern JavaScript

This is really one of the major problems. Much of the web is now fairly terrible, and so for a browser to get any meaningful traction it must work on nearly all websites; which means that any browser is going to be compromised unless it aims to remain incredibly niche. Some days I can't decide who I dislike more; MBAs or web developers.

People like to point at ladybird, but really the way to go before this can become a daily driver is probably in the hundreds of person years of development

I can't say that any of the changes of the last fifteen years have really brought about a better or more usable internet. Ads are worse than ever, the browser still doesn't know whether it's for browsing documents or hosting interactive apps, we're still stuck with javascript and no real standard library, and there don't appear to be real efforts towards reigning in complexity. From my perspective the current mode of development failed a long time ago.

Manifest v1/v2 is already an illustration of the problem. Google will use its power to impose (de-facto) standards that hurt users and give itself more money and power.

There's a value in keeping using a non-WebKit browser.

This is something I'm confused about. When people "build browsers" it's largely just Chromium with a couple of features tacked on top. I'm not sure why no one is trying to keep only the rendering engine and building their own, much better browser on top of it.

There are people doing that, Qutebrowser, Nyxt does that for WebKit(and they are moving to chromium engine). The main issue why others are not doing that is the issue with cutting yourself from extensions.

Reply on news.ycombinator.com