What Google says vs what's going on aren't necessarily the same thing, they have a long history of selling us the 'for your convenience' line while removing functionality that people depended on but that ultimately hurt Google's business interests: to be able to force feed you more ads.
They have long outlived their credit in the bank of the benefit of the doubt.
Docs: https://developer.chrome.com/docs/extensions/reference/decla...
(Disclosure: I work on ads at Google, speaking only for myself)
If you're going to argue this is better, don't point us to such a clearly worse result.
[0]:https://github.com/el1t/uBlock-Safari/issues/158 [1]:https://adguard.com/en/blog/safari-adblock-extensions.html
It is not a question that V3 breaks the gold standard privacy protecting extension.
Therefore, it's not just about changing the mechanism, the end result is clearly a lot worse. One can say, they crippled ad-blocking which this change. Hopefully, once the millions of people using ublock origin start noticing what's happening, they will move away from Chrome. I already did, ublock origin is worth more to me than any feature Google puts in Chrome.
I understand the average user probably shouldn't be able to easily hand over so much control to extensions, but on the other hand, dynamic ads shouldn't be able to serve malware or cryptominers.
I'd be a bit more open to the idea of a locked-down manifest if we had seen more good-faith attempts from AdTech to change the paradigm that makes content blockers almost a requirement.
Today we know AMP was also a anticompetitive plot to kill off header bidding.
Why should we to believe a word of what Google says about Manifest v3?
I completely disagree, and I think this will become clear as information continues to come out.
I don't expect anything that comes right out and says something like "we must implement AMP to as a strategy to remove competition from header bidding". Instead it will mostly be just the standard talking points about user experience and load times.
I could be wrong though: plenty of things have been revealed in things like text messages where people don't think of them as being part of an official record.
In this case, it appears that multiple large orgs are involved: G and M. One provides the environment and the other appears to have dropped their drawers and crapped in it but in the end a telephone should always be able to make emergency calls regardless of what is installed or configured on it.
In the end this sort of thing might look like lack of responsibility due to arrogance due to lack of competition. I'm sure other interpretations are available.
Normally the above should be considered an example of whataboutery but I think your response deserves little else. If you have something to contribute then please do but not that sort of thing.
(1) https://www.theregister.com/2021/12/09/android_911_teams/
[Edit: weird formatting snag, content unchanged]
So if you weren't asked to, you're just astroturfing voluntarily. That's not better, it's worse: If one gives up one's integrity, one should at least get paid for it. Otherwise, one isn't acting just scummily, but scummily and stupidly.
> I work on ads at Google, speaking only for myself
I'm doubtful about a person's ability to speak for themselves, when they have been consistently defending their employer on HN for years, at every occasion they got.
This is somewhat debunked in the article of this post.
> the change is that it's now declarative.
In the App MANIFEST. To my understanding, each update of those lists will require an app update (going through Store approval process).
If only it would have been dynamic, the end result would have been much better.
And again, there is no reason to disable dynamic updates if they are only lists of blocked URLs.
This is true, but I think overly performance-focused. It doesn't feel like that much time, so I think there's a valid complaint that it doesn't make sense to kneecap flexibility for speed.
> more private because an ad/content blocker doesn't need to be given such broad permissions
Sure, but this doesn't seem like the only possible solution for the people that own the browser. Why not only allow a restricted subset of JS that lacks any form of IO? Or if that's impossible/risky, why not something like Starlark or Lua?
I think this is based on a fundamental misreading of the problem. The privacy concern is that your data will leak, not merely that it's accessible to a third party. The cat binary can read my data, and I'm not at all concerned about that. So can my shell, and likewise on the concern.
Privacy doesn't necessitate this solution. It is one of the possible solutions, but I think is hard to sell as the best solution to the problem. It is likely the easiest.
Those serious "serious tradeoffs" made me completely stop using the web on my iPhone. Yeah, Safari content block can block roughly 80% of web ads, but those extra 20% are extremely annoying.
Web owners use all sorts of trickery to bypass adblockers and serve malware filled ads. Handicapping our current best defense tech against this is a sure as hell way to make me never open chrome again and completely purging it from any friends and family computer.
There is still "block request" functionality, the change is that it's now declarative. This is the same way it works in Safari, and is (a) more efficient because you don't need to execute JS to evaluate each request and (b) more private because an ad/content blocker doesn't need to be given such broad permissions. There are serious tradeoffs (no request time js makes it less flexible) but it's still very capable and easily can be used to block Google ads. Docs: https://developer.chrome.com/docs/extensions/reference/decla...
(Disclosure: I don't work at Google or on ads anywhere, speaking only for myself)
[1] I am well aware that was not the original context of the quote, but it's a nice rallying cry of the sentiment behind the movement.
I didn't see dissecting the security details as the point they were trying to make. Instead it was to partially undermine the reasons Google said they were doing this.
Basically "here's why it's bad for privacy, and here are why Google's stated reasons for the update are insufficient to justify that"
It seems like this is the clear interpretation of what I wrote and that you may be purposely misconstruing my comment in order to level a soft insult and condescending language at me over an opinion you don't agree with.
To give you a little benefit of the doubt though and assume you may just be passionate about the issue and don't intend to be insulting I will address your other point: Users giving away access to their web requests without realizing it is a problem. It is also one that can be addressed without making it much harder for privacy-minded users and the providers of those extensions to get what they want as well.
After manifest v3 it will be worse off.
Ublock origin provides a lot of privacy benefits, so people like me are 100% worse off.
Well I switched to Firefox, only really use chrome for testing nowdays.
When Google says "pull in code remotely" they dont mean from a remote server. Instead its 'code remote to Google' aka code you wrote yourself sitting on your hard drive. This kills greasemonkey/tampermonkey and all the other UserScript extensions. Google saw how great Apple is doing and fell in love with the concept of walled garden. Its their browser and they wont let you execute any code that wasnt approved by them.
User Agent no more, Its Google Agent now.
To say browser extensions pose a risk is true, but it hardly makes it in the top list of threats anymore. Malicious sites however still do and Google just restricted our ability to let third party tools provide essential services. Sure, these could be malicious, but that is generally not a wide spread IT problem of today. That should be also obvious to Chrome developers.
Accidentally they also restrict ad blockers? Come on, you are getting played.
Installing random plugins is a security issue. But web tracking is by far the more significant threat.
Specifically, "we have decided to implement DNR and continue maintaining support for blocking webRequest. Our initial goal for implementing DNR is to provide compatibility with Chrome so developers do not have to support multiple code bases if they do not want to. With both APIs supported in Firefox, developers can choose the approach that works best for them and their users."
https://blog.mozilla.org/addons/2021/05/27/manifest-v3-updat...
The entire argument regarding security doesn't mention any of the reasons Chrome developers cite its security improvement, instead it brings up that Firefox "does good enough already" and that malicious extensions can still get past the review process. the review process is by itself improved with V3 as extensions that pull in code remotely can no longer get past the review process[0], especially with how many current extensions implement RCE C&C intentionally. They also say extensions are "usually interested in simply observing the conversation between your browser and whatever websites you visit" - that's 'usually', though; malicious extensions intercepting and modifying requests for their own benefit isn't unheard of.
Instead of only stating 'this is bad', it would be beneficial to include both (A) what they say (B) their basis for the decision, if any (C) why that line of reason is incorrect/deceiving.
0: https://developer.chrome.com/docs/extensions/mv3/intro/mv3-o...