Question Which Ubo filters 2023 do you use ?

[floalma]

Hi friends,

Spoiler

This is an update to a previous thread here about Ubo filters lists.
These are my UBO filter lists.
Any comments or recommendations are welcome.
Thanks you for your contribution.

 

[ForgottenSeer 97327]

Indeed, I saw that and won't use it for that reason. Your suggestion to use Easy List & Privacy minified instead seems like a good one.

Well I posted that uBO digests 90%of AG filters, but only 23 errors on over 64 rules (in optimized version) means that uBO digest 99,99%

Spoiler: Here is a screen shot of the rejected rules, no big deal

 

[oldschool]

Well I posted that uBO digests 90%of AG filters, but only 23 errors on over 64 rules (in optimized version) means that uBO digest 99,99%

Spoiler: Here is a screen shot of the rejected rules, no big deal

View attachment 278001

I guess you look into these things more than I do. I didn't bother doing the math either.

BTW, I wonder why Adguard keeps Fanboy's Enhanced Tracking in its lists? To assuage the "More is better" user?

 

[ForgottenSeer 97327]

I guess you look into these things more than I do. I didn't bother doing the math either.

BTW, I wonder why Adguard keeps Fanboy's Enhanced Tracking in its lists? To assuage the "More is better" user?

I guess AG put it in initially to offer their customers the old-trusted-known filters on top of their own filters. So yes, that would be a valid explanation.

Another guess is that they don't have people looking at third-party filters manually anymore. When they should have checked (like @Jan Willy did) they would have seen it is not maintained anymore. I think they have automated the maintenance and update. As @South Park posted a subset is maintained, so their (automated) update check confirms that it changes from time to time. Also in my post#94 I added two links of Brave research. In one of them they also discovered that a substantial part of "old" (5 year old) rules keep blocking stuff. Which means that in their (automated) triggering check they see that some of those block rules still block stuff. As said all guestimates, but this could be the reason it stays in the list.

 

[Jan Willy]

According to Yuki2718, AG Tracking Protection (alias Spyware Filter) and EasyPrivacy set different accents. View

[QUESTION ]Spyware Filter and/or EasyPrivacy · Issue #77015 · AdguardTeam/AdguardFilters

I use both Spyware Filter and the EasyPrivacy filter. However if EasyPrivacy uses all its rules the AdGuard Spyware Filter uses only the half of its rules. So there are a lot of double filter rules...

github.com

 

[Jan Willy]

This is beside the point that testing shows them to be pretty ineffective.

The reason could be what Lenny_Fox said:
'The URL/IP-blacklists of DNS-services and AV-companies are much larger than all the bllocklist you can add to uBO.'

View Need Advice - Which block lists do you use with UBO, AG and ABP ?

 

[nicolaasjan]

I have to say that all these different opinions are quite confusing to me.

I have now disabled the Malicious URL Blocklist and enabled Firefox own Safe Browsing protection.
Firefox Enhanced Tracking Protection is set to "Strict".

Disabling Easylist reduced my score on d3ward to from 100% to 96%, so I turned it on again (yes, I know this test is flawed).
Easylist minified didn't remedy this.

[Edit]
I don't have Peter Lowe's list enabled, because that one is already in my hosts file.
I also use Quad9 DNS, which blocks malicious sites as well.

Should I also enable EasyPrivacy again?

 

[Gandalf_The_Grey]

You could add EasyPrivacy (Optimized (less rules that EasyPrivacy).
I get now 100% on this flawed test

 

[nicolaasjan]

You could add EasyPrivacy (Optimized (less rules that EasyPrivacy).

That didn't get me beyond 96% either. haha!

I guess I'll have to keep this dubious stock EasyList enabled for now...
Or learn not to trust dubious "Adblocker Tests".

 

[Jan Willy]

That didn't get me beyond 96% either. haha!

I guess I'll have to keep this dubious stock EasyList enabled for now...
Or learn not to trust dubious "Adblocker Tests".

You can add custom rules specific for the few items that are not covered so far.
BTW I don't understand why stock EasyList should be 'dubious'.

 

[Zappathustra]

That didn't get me beyond 96% either. haha!

I guess I'll have to keep this dubious stock EasyList enabled for now...
Or learn not to trust dubious "Adblocker Tests".

The only thing dubious is the page that tests adblockers.

 

[Gandalf_The_Grey]

That didn't get me beyond 96% either. haha!

I guess I'll have to keep this dubious stock EasyList enabled for now...
Or learn not to trust dubious "Adblocker Tests".

EasyList also has an optimized version.
And yes, do not trust dubious adblocker tests

 

[ForgottenSeer 97327]

uBO cheats

Spoiler: uBO cheats with specific rules for this faulted test

Spoiler: Do you need more proof that this test is faulted?

with less than 2% of the rule count nearly 200% result, do you also believe in Santa Claus?

 

[Gandalf_The_Grey]

uBO cheats

Spoiler: uBO cheats with specific rules for this faulted test

View attachment 278019

I think Raymond Hill aka gorhill was fed up with this tool, probably because of this:

This tool has been doing more damage to content blockers because it reports flawed results.

Hot Take - Evaluate your content blocker with Ad Block Tester

they allow users to disable the telemetry, etc. This is why @upnorth uses it and tracks their releases on the forum. I do? I thought it was because of their cute logotype. :p😅

malwaretips.com

 

[nicolaasjan]

BTW I don't understand why stock EasyList should be 'dubious'.

My apologies for that rant.
I take my words back, if y'all think that list is fine.

I got confused by @Max90 earlier post, I guess.

Picture 3 - Only 201 rules of EasyList account for 90% of the blocks (source Brave - warning for stale rules)

 

[simmerskool]

I got confused by @Max90 earlier post, I guess.

you are not alone... (not necessarily @Max90 -- I find this thread interesting but somewhat confusing...)

 

[ForgottenSeer 97327]

@simmerskool and @nicolaasjan

The problem is only AdGuard knows which 201 rules accounted for 90% of the blocks and we (at least I) don't know which. Because AdGuard has a DNS service (and a feedback opt-optout in their extension) they have a pretty good idea which domains are blocked (and can assess which rules triggered those blocks).

I had no intend to bring EasyList filters are useless or full of bloat. The exponential graph of the Brave research shows that the added protection maxes out. Mind teaser was that it was probably more effective to use AdGuard onlt filters in uBO and disable all other stuff (except for the ones you added manually).

 

[simmerskool]

@simmerskool and @nicolaasjan

The problem is only AdGuard knows which 201 rules accounted for 90% of the blocks and we (at least I) don't know which. Because AdGuard has a DNS service (and a feedback opt-optout in their extension) they have a pretty good idea which domains are blocked (and can assess which rules triggered those blocks).

I had no intend to bring EasyList filters are useless or full of bloat. The exponential graph of the Brave research shows that the added protection maxes out. Mind teaser was that it was probably more effective to use AdGuard onlt filters in uBO and disable all other stuff (except for the ones you added manually).

...ok, I follow what you are saying. I'd understand this thread better if I opened my uBo settings to see exactly what filters I'm using, but doesn't seem broken here, so I'm just reading along... ... in case I read something that strikes a nerve.

 

[nicolaasjan]

The problem is only AdGuard knows which 201 rules accounted for 90% of the blocks and we (at least I) don't know which.

Why don't they share that knowledge?

probably more effective to use AdGuard only filters in uBO and disable all other stuff

I wouldn't dare to disable the uBO filters. They're very effective.

 

[JohnR]

I'm not a fan of YT Shorts, and use https://raw.githubusercontent.com/gijsdev/ublock-hide-yt-shorts/master/list.txt

 

[Zappathustra]

you are not alone... (not necessarily @Max90 -- I find this thread interesting but somewhat confusing...)

It depends on who you are. Are you a minimalistic kind of person like Marie Kondo? Choose minified Easylists.

You think less is good, but not necessarily more? Choose optimized Easylists.

You love headaches or banging your head against the wall like Roger Waters? Choose all the lists you find: less than1 milion rules and 900 000 cosmetic filters it's a failure