4 ms·

> 99.9% of people with adblockers installed never ever touch those [advanced features] Custom matching algorithms and ability to fine tune or expand matching a

> 99.9% of people with adblockers installed never ever touch those [advanced features] Custom matching algorithms and ability to fine tune or expand matching algorithms according to new content blocking challenges are actually a kind of advanced feature that are used by all those users without them ever realizing it since their content blocker work seamlessly on their favorite sites without the need for intervention. The declarativeNetRequest (DNR) API has been quite improved since it was first announced and its great, but since it's the only one we can use now, it's no longer possible to innovate by coming up with improved matching algorithm for network requests. If the DNR had been designed 8 years ago according to the requirements of content blockers back then, it would be awfully equipped to deal with the challenges thrown at content blockers nowadays, so it's difficult to think the current one will be sufficient in the coming years.

londons_explore 4y ago
Nothing stops the API evolving... Anyone can make a build of Chromium with a better API, test it out with their own extension, and if it works better, then they can send a PR to get that API into Chrome. Obviously, extending an API is a long term commitment, so I can understand the Chrome team wanting to only do it if there is a decent benefit - "it makes my one extension with 10 users work slightly better" probably doesn't cut it.
- doomrobo 4y ago
  > Nothing stops the API evolving... Anyone can make a build of Chromium with a better API, test it out with their own extension, and if it works better, then they can send a PR to get that API into Chrome. This is not at all how API proposals are handled. There are a lot more (time, financial, logical) barriers to a change like this. See the role of the W3C in this: https://www.eff.org/deeplinks/2021/11/manifest-v3-open-web-politics-sheeps-clothing https://www.eff.org/deeplinks/2021/11/manifest-v3-open-web-p...
  - westurner 4y ago
    It is reasonable to expect BPF or a BPF-like filter. https://en.wikipedia.org/wiki/Berkeley_Packet_Filter https://en.wikipedia.org/wiki/Berkeley_Packet_Filter
- westurner 4y ago
  bromite/build/patches/Bromite-AdBlockUpdaterService.patch: https://github.com/bromite/bromite/blob/master/build/patches/Bromite-AdBlockUpdaterService.patch https://github.com/bromite/bromite/blob/master/build/patches... bromite/build/patches/disable-AdsBlockedInfoBar.patch: https://github.com/bromite/bromite/blob/master/build/patches/disable-AdsBlockedInfoBar.patch https://github.com/bromite/bromite/blob/master/build/patches... bromite/build/patches/Bromite-auto-updater.patch: () https://github.com/bromite/bromite/blob/master/build/patches/Bromite-auto-updater.patch https://github.com/bromite/bromite/blob/master/build/patches... - [ ] ENH,SEC,UPD: Bromite,Chromium: is there a url syntax like /path.tar.gz#sha256=cba312 that chromium http filter downloader could use to check e.g. sha256 and maybe even GPG ASC signatures with? (See also: TUF, Sigstore, W3C Blockcerts+DIDs) Bromite/build/patches/Re-introduce-*.patch: [...]

londons_explore 4y ago

Nothing stops the API evolving... Anyone can make a build of Chromium with a better API, test it out with their own extension, and if it works better, then they can send a PR to get that API into Chrome. Obviously, extending an API is a long term commitment, so I can understand the Chrome team wanting to only do it if there is a decent benefit - "it makes my one extension with 10 users work slightly better" probably doesn't cut it.

doomrobo 4y ago
> Nothing stops the API evolving... Anyone can make a build of Chromium with a better API, test it out with their own extension, and if it works better, then they can send a PR to get that API into Chrome. This is not at all how API proposals are handled. There are a lot more (time, financial, logical) barriers to a change like this. See the role of the W3C in this: https://www.eff.org/deeplinks/2021/11/manifest-v3-open-web-politics-sheeps-clothing https://www.eff.org/deeplinks/2021/11/manifest-v3-open-web-p...
- westurner 4y ago
  It is reasonable to expect BPF or a BPF-like filter. https://en.wikipedia.org/wiki/Berkeley_Packet_Filter https://en.wikipedia.org/wiki/Berkeley_Packet_Filter
westurner 4y ago
bromite/build/patches/Bromite-AdBlockUpdaterService.patch: https://github.com/bromite/bromite/blob/master/build/patches/Bromite-AdBlockUpdaterService.patch https://github.com/bromite/bromite/blob/master/build/patches... bromite/build/patches/disable-AdsBlockedInfoBar.patch: https://github.com/bromite/bromite/blob/master/build/patches/disable-AdsBlockedInfoBar.patch https://github.com/bromite/bromite/blob/master/build/patches... bromite/build/patches/Bromite-auto-updater.patch: () https://github.com/bromite/bromite/blob/master/build/patches/Bromite-auto-updater.patch https://github.com/bromite/bromite/blob/master/build/patches... - [ ] ENH,SEC,UPD: Bromite,Chromium: is there a url syntax like /path.tar.gz#sha256=cba312 that chromium http filter downloader could use to check e.g. sha256 and maybe even GPG ASC signatures with? (See also: TUF, Sigstore, W3C Blockcerts+DIDs) Bromite/build/patches/Re-introduce-*.patch: [...]