New Update Google’s Manifest V3 Still Hurts Privacy, Security, and Innovation

A very thorough explanation of MV3's impact on Ghostery, discussion of alternative implementations and the overall implications for users, adblockers and other developers. I recently found this but have no idea when it was first published.

Interested members may follow discussions of MV3 technical issues here: Issues · w3c/webextensions
The limitations seem too much, and the workarounds seem like too much of a hurdle and not good enough.
 
The limitations seem too much, and the workarounds seem like too much of a hurdle and not good enough.
Indeed, which is why I posted it as another example. I believe that content blocking will be only as good as the browser's built-in mechanism allows - plus whatever limited (and mangled) function is available via MV3 extensions. Too many technical issues to implement at this late stage even if Google agreed to positive changes. Users will have to pick their poison with browser choice.
 
AdGuard V4 has a new engine which can also process new Manifestv3 rules with other API. There are still some issuee (with Mv3) with Google (e.g grant rights to the extension per website), but it looks like AG extension will survive Mv3.

That is why I am using Edge with tracking protection on strict and have set AGv4 to allow per website (to get used to using browser build-in with allow AG per website of AG and extra rules for that website).

My other bet for Mv3 is Brave. Also using build-in browser adblocker and copying AG's rules to my personal rules.
 
I don't understand the details ... how?
He probably meant this
1.png
 
@SeriousHoax answered your question @oldschool

In Mv3 background pages are not allowed anymore and service worker alternative offered in Mv3 seems to have limitations also or does not work properly yet. I don't whether these are just startup bugs or unrevealed limitations Google is sneakingly imposing on extensions (Google has often communicated that they would like to require the user to explicitly allow permissions to an extension for every single website).
 
The biggest problem is that most of the current solutions rely heavily on community driven filter lists.
Brave uses u-block origin compatible filter lists, Adguard's filters are based on EasyList / EasyPrivacy, etc.

I am afraid that most of the community filter lists will deprecate and fade away when MV3 gets fully enforced.
If this happens, u-block origin will just die, Adguard & Ghostery will have to invent new technologies (AI / smart ad & tracking prevention) which will not use community lists.
 
I am afraid that what @Gangelo posts is true. AdGuard has filterlists convertor (from ABP and uBO to AG and from AG to uBO). The new filterformat can be automatically converted, so I expect AG to add another 'convertor'. Also AG maintains its own filters, so I expect them not to fade. Also I promiss to convert my Kees1958 filter to new format :-)
 
Adguard & Ghostery will have to invent new technologies (AI / smart ad & tracking prevention)
... which is what I've presumed, given their business models, and MV3 will leave the small developers in the dust.

Maybe reality is starting to set in for users of content blockers.
 
... which is what I've presumed, given their business models, and MV3 will leave the small developers in the dust.

Maybe reality is starting to set in for users of content blockers.
Exactly.

Aguard & Ghostery already have a paid subscription model for their suites and applications so they can support filter list maintenance and development of new technologies.
Ad blocking in general will need to move on from filter lists eventually.
 
The biggest problem is that most of the current solutions rely heavily on community driven filter lists.
Brave uses u-block origin compatible filter lists, Adguard's filters are based on EasyList / EasyPrivacy, etc.

I am afraid that most of the community filter lists will deprecate and fade away when MV3 gets fully enforced.
If this happens, u-block origin will just die, Adguard & Ghostery will have to invent new technologies (AI / smart ad & tracking prevention) which will not use community lists.

I dont think it will be that extreme, Brave has hired "Fanboy" ( Ryan Brown), the top committer of EasyList and EasyPrivacy and Brave adblocker isnt affected at all by MV3.


So with this commitment from Brave I think AdGuard and Ghostery will be fine, considering that they have their own resources and manpower too.
 
Maybe the combination of DNS that can use filter lists and heuristic extensions will be the way to go...
Smart AdBlock was ahead of its time. It was based on heuristics and it did quite a good job at blocking while using less resources than even uBO.

It only had one flaw: It doesn't support cosmetic filters. So you are left with gaps and iframes saying that they didn't load etc. But yeah heuristics seems to be good enough to detect virtually all ads. Just not good enough to fill in the gaps they leave.
 
Adguard & Ghostery will have to invent new technologies (AI / smart ad & tracking prevention) which will not use community lists.
BTW, aren't Adguard and Ghostery both browser extension open source? Does it mean if gorhill of uBO can't manage to come up with a solution, he can simply look at Adguard & Ghostery's code to have an idea and come up with a similar solution?
 
BTW, aren't Adguard and Ghostery both browser extension open source? Does it mean if gorhill of uBO can't manage to come up with a solution, he can simply look at Adguard & Ghostery's code to have an idea and come up with a similar solution?
I do not think that the extensions will do much with MV3. When I mentioned Adguard & Ghostery in my previous posts I was referring to the paid apps /suites.

The extensions are indeed open source.
 
Last edited: