[NSFW] komikdewasa.club Pop Up

Description of the issue:
Pop up appears on this site.ABPindo List is active nothing changed, When I tested with uBO there is no issue with ABPindo List

Exact URL of the website in question:
komikdewasa.club

Screenshot of the ad as it appears in Brave
**Did the issue present with default Shields settings? (yes/no)**YES

**Does the site function as expected when Shields are turned off?**NO

**Is there a specific Shields configuration that causes the site to break? If so, tell us that configuration. (yes/no):**NO

**Does the site work as expected when using Chrome?**NO

**Brave version (check About Brave):**1.48.164

Same issue as [NSFW] komikmanhwa.me ads it is a regex issue, not a ABPindo thing. That issue with .me was fixed btw, it is in Easylist now so you can delete the custom rule I told you to add if you still have it.

You can add in the meantime this custom rule until it is fixed:

||armlockyakker.com^

Thanks, i hope they fix adblock rust

EDIT: BTW you shouldn’t mark my post as solution, in fact, if you have to tag FanboyNZ do it, make sure this gets added to the proper list and then mark that as solution, if not other users will not get the proper fix.

Well, it’s a tricky ‘fix’ because RUST lang is the one that doesn’t want to implement the lookaround features as I linked on the other post, it is because of performance and maintenance and he, the maintainer, believes you shouldn’t have to use lookaround features anyway, so unless someone just goes and adds it and do it, he won’t.

So Brave would have to completely switch engines or crate but the truth is, regex is like procedural cosmetics, they can make things easier in a way, but shouldn’t be used because they are just not efficient enough.

Of course, In this case using regex is justified, they make it nice and easy to block many requests by regex, it’s fine because these awful tracker/ads scripts have similar structure in the URL so it seems to work fine without breaking any good script, which is great!

BUT the problem is the regex rule uBlock created is using a lookaround feature, which is not doing anything really, I don’t know what it is there, and that’s the logic RUST guy says, you shouldn’t relay on lookaround if you can make it normal. I modified the rule not to have lookaround and it works, just like I have seen people doing whatever they want it as long as ‘it works’.
I have seen terrible made rules in uBlock lists, where Procedural cosmetics are used and are not needed and are actually redundant, so people do whatever they want, instead of making things simple and more efficient if possible.

But for the way adblockers work, because I am sure in other RUST implementations lookaround features can be useful, Adblocker should be the first one not using lookaround because that’s not what you want to do in an adblocker rule.
Adblockers are always matching a URLs for requests to block so adding just a to the custom filters will result in the same as using ||brave^ because it will match anything in the URL that has the letter a on it, so why do you want to make it backwards and say for example say ‘match all the alphabet but not the letter a’? it doesn’t make sense 99.9% of the time.

That’s the logic of using lookaround in adblockers. Adblockers are not supposed to make some conditional rules where you negate some stuff and then match other, or where want to match something based on words or letters or numbers matching the whatever before or after, it’s just not efficient or good practice, especially when you can restrict rules per domain, you are not matching multiline or anything like big lines or anything, it is just: just one URL at the time.

It would be easier and more efficient just to make the rules normal and move on, no regex required for that!

I mean, the only example where I can think of needing lookaround would be to workaround a feature called denyallow which is not supported by Brave, but what it does is you tell the adblocker to allow ONLY domain1 and domain2 in example.com, and then block everything else.
Sounds nice, but it’s rarely used for its nature, but Regex in that case would be needed to match and do the same thing as denyallow, two block everything based on the whatever domains you want to allow.
It’s rare when you know you can do it differently, but also there is a feature that can do the exact same thing (and more) Content Security Policy (CSP), so choosing between regex and CSP, obviously CSP would win.

So I think ‘fixing’ Brave adblocker is not really need, it would be better if they just added a fixed version of the regex to the Brave unbreak without any lookaround which is useless in this type of rules anyway, or tell uBlock to edit their rules to stop using something that shouldn’t even be used in 99.9% of cases.
Or do what they do today, you see weird ads or something in the page, you report it, it gets added to Easylist or Brave unbreak and move on. Of course uBlock users will not report it because it will work for them and they will not notice those domains are not in Easylist, so practically it is in the shoulders of Brave users to do it.

@fanboynz The fix still hasn’t arrived

Thanks for the explanation. Hopefully Brave users like ublock’s users, rarely seeing ads

Improvements are incoming, we’ve added around 500+ revolving adservers in the last 24hrs, not every site is covered and depending on the site it may update daily. But it should be much of improvement. Give it 24/48hrs.

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.