# Ad blocking test at d3ward is extremely flawed **URL:** https://discuss.privacyguides.net/t/ad-blocking-test-at-d3ward-is-extremely-flawed/20644 **Category:** General **Created:** 2024-09-06T13:39:03Z **Posts:** 33 ## Post 1 by @jerm — 2024-09-06T13:39:03Z [https://xcancel.com/GrapheneOS/status/1831877968895078464#m](https://xcancel.com/GrapheneOS/status/1831877968895078464#m) > Vanadium will not include rules for cheating at content filtering tests because we think it would be a breach of user trust, regardless of flaws in a test. > **[How “test your adblocker” websites can harm users and the adblocker ecosystem...](https://brave.com/blog/adblocker-testing-websites-harm-users/)** > > Adblocker testing sites can be misleading—or even harmful—for a number of reasons. We've outlined the four major issues. --- ## Post 2 by @anon32558482 — 2024-09-07T19:02:42Z There is also a small thread on their Mastodon page. --- ## Post 4 by @xe3 — 2024-09-07T21:48:58Z The usefulness of this type of test is always going to be kind of limited. But I don’t really understand why they chose to characterize other adblockers as “_cheating_”. Seems like hyperbole and another example of the “everyone else is wrong/acting in bad-faith” mentality. That chain of tweets unnecessarily assumes other projects are all acting in bad-faith, rather than taking the time to understand that the frustration they now feel for this sort of test is actually shared by the peers they just accused of “cheating.” Here are the actual sections they chose to characterize as ‘cheating’ (pay special attention to the comments/explanations): > **uBlock Origin** > > > [Gorhill]: Please do not use those online tools which claim to measure how good is your content blocker, they are often flawed. > > ``` > ! defuse pointless online tool causing hardship to volunteers > ! https://twitter.com/gorhill/status/1583581072197312512 > *$3p,domain=d3ward.github.io > @@*$redirect-rule,domain=d3ward.github.io > d3ward.github.io##.textads > ``` > **Adguard:** > > ``` > ! SECTION: d3ward.github.io > ! > ! We're getting tons of questions from users who do not understand why some test on the internet does not show 100%. > ! We're tired of explaining why this or that domain should not be blocked in reality so there's that > ! > ! Following domains should not be blocked globally because they are not used to collect data (admin panels, bug tracking services) > ! > ! Google Analytics admin panel. We're blocking what's required on this web site, but more granuralry (i.e. `/collect` location). Unfortunately, this test check for the whole domain blocking, hence here's a rule > ||analytics.google.com^$domain=d3ward.github.io > ! Used as a click-through tracking domain for Google Analytics when a user clicks on a ad link that has been tagged with Google Analytics tracking code > ||click.googleanalytics.com^$domain=d3ward.github.io > ! Blocking of this domain may break Yahoo apps > ||log.fc.yahoo.com^$domain=d3ward.github.io > ! Used for managing Yandex AppMetrica > ||appmetrica.yandex.ru^$domain=d3ward.github.io > ||metrika.yandex.ru^$domain=d3ward.github.io > ! Used by TikTok Business API for managing advertising campaigns > ||business-api.tiktok.com^$domain=d3ward.github.io > ! Used for creating Pinterest widgets > ||widgets.pinterest.com^$domain=d3ward.github.io > ! Pinterest Analytics dashboard > ||analytics.pinterest.com^$domain=d3ward.github.io > ! > ||udc.yahoo.com^$domain=d3ward.github.io > ||udcm.yahoo.com^$domain=d3ward.github.io > ! Bug tracker services > ||notify.bugsnag.com^$domain=d3ward.github.io > ||sessions.bugsnag.com^$domain=d3ward.github.io > ||api.bugsnag.com^$domain=d3ward.github.io > ||app.bugsnag.com^$domain=d3ward.github.io > ||browser.sentry-cdn.com^$domain=d3ward.github.io > ||app.getsentry.com^$domain=d3ward.github.io > ! > ! `afs_ads` class - popular anti-adblock bait > d3ward.github.io##.afs_ads > ! NOTE: d3ward.github.io end ⬆️ > ``` * * * > **Here is what I think is a less hostile/bitter sounding rephrasing of the partially valid point they tried to articulate:** > > Users shouldn’t make generalizations about the quality or effectiveness of an adblocker based on tests like [d3ward.github.io](https://d3ward.github.io/toolz/). While these tests can have value and do serve a purpose, they cause much misunderstanding among users who don’t understand how to interpret the results nor understand the limitations of the test and flaws in the methodology. For this reason tests like this probably do more harm than good. > > In addition to this, tests like this one create a lot of unnecessary work and frustration for volunteers and filter list maintainers, due to persistent misunderstandings among the userbase about the significance of the test results, the % value, as well as flaws in the test itself. This is a time suck. > > GrapheneOS is a new entrant to the adblocking space, but other adblockers like uBlock Origin and Adguard have long been frustrated by the situation." After years of trying to educate users and a longstanding bug in the test effecting uBO, uBO chose to just block all 3rd party resources on [d3ward.github.io](http://d3ward.github.io), and Adguard has created rules for that specific domain to block specific 3rd party resources for that site only that they don’t block globally, and don’t recommend blocking globally. > > In both cases the goal is to reduce the number of support requests, and misguided reports and complaints by users who misunderstand the results. But GrapheneOS considers this approach to be misleading in and of itself. Instead of taking a similar approach, we are trying to raise awareness about the limitations of these tests, and hope our users will understand. --- ## Post 5 by @anon32558482 — 2024-09-07T22:41:41Z > [@xe3](#): > > But I don’t really understand why Daniel chose to characterize other adblockers as “_cheating_”. Seems like hyperbole and another example of the “everyone else is wrong/acting in bad-faith” mentality. That is my sense as well. I do like the OS but the constant blaming other projects is getting old. --- ## Post 6 by @Valynor — 2024-09-07T23:00:26Z I never felt that d3ward was a scientific-precision ad block tester at all but more of a quick general test that gives you a reasonable idea of where you stand. Some people really need to relax a bit. :wink: --- ## Post 7 by @anon63378630 — 2024-09-07T23:33:07Z That test is/was actually broken. I can get a 100% on it without any blocker. --- ## Post 8 by @anon55142611 — 2024-09-07T23:40:25Z PRobably because of DNS ? --- ## Post 9 by @jonah — 2024-09-07T23:58:37Z Why exactly should I care whether some random website is a good test. Is it commonly used? Never heard of it myself lol --- ## Post 10 by @exaCORE — 2024-09-07T23:59:49Z > [@jonah](#): > > Why exactly should I care whether some random website is a good test You have to care about this issue immediately because GrapheneOS said so on Twitter! /s --- ## Post 13 by @anon48875053 — 2024-09-08T08:02:28Z I just typed “adblock test” into Brave Search and that test was the first result. --- ## Post 15 by @anon46412288 — 2024-09-08T08:05:34Z I have used it a bit in the past for testing ad-blockers. It’s very commonly pointed to to test the effectiveness of an ad-blocker. It’s not some random website tbh. --- ## Post 16 by @anon48875053 — 2024-09-08T08:17:14Z One more thing that should be noted is that these are project accounts, more than one person has access to them, but I like how everyone just assumed that Daniel was the one who made these posts. --- ## Post 17 by @eqrlzo8t — 2024-09-08T08:20:01Z > [@xe3](#): > > That chain of tweets unnecessarily assumes other projects are all acting in bad-faith, rather than taking the time to understand that the frustration he now feels for this sort of test is actually shared by the peers he has just accused of “cheating.” Exactly this. Originally uBO repeatedly declined to fix the test / POC sites. This site alone dated back to 2021: - [d3ward.github.io: ads · Issue #10487 · uBlockOrigin/uAssets · GitHub](https://github.com/uBlockOrigin/uAssets/issues/10487#issuecomment-968988473) > gorhill: Anyways, I just realize OP is just reporting a proof-of-concept page, we do not fix these. - [d3ward.github.io · Issue #11103 · uBlockOrigin/uAssets · GitHub](https://github.com/uBlockOrigin/uAssets/issues/11103#issuecomment-1002143461) > gorhill: We do not add filters for proof-of-concepts, we add filters for real-world cases after ensuring no widespread breakage occurs as a result. - [d3ward.github.io: detection · Issue #12564 · uBlockOrigin/uAssets · GitHub](https://github.com/uBlockOrigin/uAssets/issues/12564) > okiehsch: We don’t fix test cases. When the prerequisites for new issues became tighter, and people were able to search duplicated issues, github reports slowed down. But then, they turned to uBO subreddit. It was asked so many times, and uBO team had to reply again and again. It got even more ridiculous when I saw a reddit user wanted to use the old “uBlock” ( **not** uBO) since it has higher scores, despite not updated for a long time, simply because it doesn’t have as many redirection filters as uBO. Eventually uBO team had enough, since the questions just come despite all the tweets gorhill posted and all the explanations from the team, and “fixed” that test site. After that, questions stopped and volunteers can focus on the real issues of real sites. Funny that Graphene did nothing to help during those whole times with uBO and Adguard. And now when they finally implement their own blocker, and experience the same questions from their users, they start to accuse other projects. --- ## Post 20 by @anon17263739 — 2024-09-08T17:57:08Z “Cheating” is definitely the correct word. Just because a project you don’t like points something out, doesn’t make it invalid. The project account did not say Adguard bad or uBlock sucks, just that they cheat the results. Read jonah’s minutes of meeting about not wanting free accounts for reviews because developers might cheat, so is he accusing every subscription based service of being evil, or is he using the word as it is intended to be used? (Disclaimer: am a disposable account of a lurker) --- ## Post 23 by @xe3 — 2024-09-08T18:04:18Z > [@anon17263739](#): > > Just because a project you don’t like Most of us like GrapheneOS. Please don’t imply that anyone who disagrees with GOS’s misleading characterization must only be disagreeing because they dislike GOS. > doesn’t make it invalid. Nobody’s argument rested on dislike for GOS. --- ## Post 27 by @jonah — 2024-09-08T18:16:00Z Anyways! I think it is fair to say that uBlock Origin and AdGuard are cheating tests if they are indeed blocking requests on the domains of test suites. I just think it is unfortunate that someone decided to frame this situation as bad behavior on their part, in order to advertise the new ad blocking capabilities in their own web browser. The real takeaway is very simple, and doesn’t require all the drama that has been injected into the conversation: these in-browser testing suites are a poor way to validate the effectiveness of these content blockers. It is very similar to the gross overuse of “browser fingerprint test” websites that unnecessarily scare people into thinking the privacy tools they use are inadequate. TL;DR just don’t use d3ward’s site or ones like it and move on with your lives lol --- ## Post 28 by @anon48875053 — 2024-09-08T18:19:08Z > [@jonah](#): > > I just think it is unfortunate that someone decided to frame this situation as bad behavior on their part, in order to advertise the new ad blocking capabilities in their own web browser. Nobody framed it as “bad behavior.” It was just framed as cheating because it’s indeed cheating. As for advertising, it wouldn’t make sense at all. GOS couldn’t care less about ad blocking, security and privacy are their priorities, their adblocker doesn’t even block YouTube ads. --- ## Post 31 by @Milus — 2024-09-08T18:23:24Z > GOS couldn’t care less about ad blocking, security and privacy are their priorities You keep repeating this. Do you know how much malicious code can be injected through Ads? --- ## Post 32 by @anon48875053 — 2024-09-08T18:26:56Z Why should I care if that malicious code runs on the site or through ads? The browser should protect me regardless. --- ## Post 33 by @eqrlzo8t — 2024-09-08T21:03:01Z They couldn’t care less about adblocking, and they make long threads of accusation to other blocker projects to satisfy their drama need. They actually cared less about adblocking when they didn’t implement their own blocker and their users didn’t complain about that site. They are just framed as advertising because they are indeed advertising > Vanadium comes with a default enabled ad blocking implementation. It currently uses Chromium’s standard filter engine without cosmetic filtering or the advanced uBlock Origin syntax so it doesn’t quite block as much as uBlock Origin or Brave, but it’s more efficient and simpler. --- ## Post 34 by @jonah — 2024-09-08T21:48:07Z Unfortunately, every time GrapheneOS is so much as merely mentioned in a conversation, the entire thing devolves into petty nonsense about them. Locking because it seems the actual conversation about adblock test sites concluded long ago, and we’re just stuck on GrapheneOS drama _again_. --- ## Post 35 by @jonah — 2026-06-29T19:47:00Z cleaning up off-topic posts