Live data from Hacker News

Brave Improves Its Ad-Blocker Performance with New Engine in Rust

brave.com

61–70 of 382 posts

Re: Brave Improves Its Ad-Blocker Performance with New Engine in Rust

#61

Earlier quoted context omitted.

How can people, who are not using those websites, avoid Facebook tracking on the rest of the web?

Install one of the many chromium extensions that block/obfuscate such tracking.

Sadly not an option on iOS, which is the reason for using Brave in the first place.

Re: Brave Improves Its Ad-Blocker Performance with New Engine in Rust

#62
post #18

Earlier quoted context omitted.

It's potentially even worse than that. Exploiting this grey area in the law will lead to a response from the courts and lawmakers. That response might well require that browsers display pages exactly as the publisher intends. In other words, the end of ad blockers and other tools, thanks to Brave. (Not at all unlikely given that Google seems to want that too.)

It feels to me like this could be legislated based on profiting on others' copyright. The issue here is one party making money, without consent of any form, based on another party's published materials.

> The issue here is one party making money, without consent of any form, based on another party's published materials.

So, google’s search engine business?

Re: Brave Improves Its Ad-Blocker Performance with New Engine in Rust

#63

I'm impressed with Brave. I've started using it as my primary browser (after a couple false starts a year or so ago) and it has all the best of Chrome without the worst. There is controversy about how Brave does stuff, and I agree there is some sketchy stuff they are doing (replacing ads with their own). I'm glad they are giving us a choice, though, and exploring different options for sites being able to makes some m…

I can't tell that I'm seeing "their" ads. The way I see ads is a notification type prompt where I have to click view and navigate to a sponsored page. Do you have a side by side where Brave replaces a site's ad with their own?

I don't think that's a thing they do, so I'm curious to hear from GP. And in terms of the notification prompts, that's something you have to enable (disabled by default) if you want to "earn" tokens.

Re: Brave Improves Its Ad-Blocker Performance with New Engine in Rust

#64

Earlier quoted context omitted.

Install one of the many chromium extensions that block/obfuscate such tracking.

Sadly not an option on iOS, which is the reason for using Brave in the first place.

Honestly, blocking ads/trackers on mobile has always been a tricky business. You could set up Pi-Hole on your home network. Or see if there's something similar to this on iOS: https://f-droid.org/en/packages/dnsfilter.android/

Re: Brave Improves Its Ad-Blocker Performance with New Engine in Rust

#65
post #25

Earlier quoted context omitted.

Algorithm. The title is a bit misleading, and conveys that "it's faster because we re-wrote it in rust". If any thing, the move to rust might cause it to be a little slower (compared to the same thing written in c++ and considering how heavily optimized they said it was; not dinging rust, but you can't just port cpp optimizations to rust with syntax changes due to safety requirements etc).

Conversely, the language semantics make some more aggressive optimizations possible by the compiler, without the programmer having to drop hints.

Why is this necessarily true? All Rust gives you is safety -- no data races and better memory safety. But since data races and use-after-free are undefined behavior in C++ anyway, C++ compilers can still aggressively optimize by assuming those two don't happen.

It's true that some compilers disable optimizations on certain undefined behavior to prevent unintuitive behavior (https://blog.regehr.org/archives/140) but that's unlikely to happen for data races or memory safety.

Re: Brave Improves Its Ad-Blocker Performance with New Engine in Rust

#66
post #51

Earlier quoted context omitted.

> But even if you're right, why is it unethical for me to install software on my own computer to render websites differently than the creator intended? It's not unethical for you to run that program, but what is being argued is that Brave (the company) is acting unethically, for example by taking donations on behalf of other people without their consent[0]. [0]: https://news.ycombinator.com/item?id=18734999

Except they don't. See above.

Well, that's good! Might give Brave a spin then.

Just curious, what exact sentence(s) from the above links to the Brave website do you see contradicting this claim about "shadow wallets"? (if that's the correct wording)

When I read something like this, it almost sound like they're making wallets (without necessarily asking beforehand):

> If you own a website, or even just a blog or other sub-domain on a hosting site, we’ll create a site wallet for you.[0]

Am I reading that wrong?

[0]: https://brave.com/how-brave-works-for-you/

Re: Brave Improves Its Ad-Blocker Performance with New Engine in Rust

#67
post #6

Earlier quoted context omitted.

They replace other ad network ads with their own ~ads~ crypto currency . They monetize the content of others and don't tell them. If you have a youtube channel, they will swap out ads and collect the ad revenue without telling the channel that they're doing so. I'm not particularly impressed with their business. In fact I'm pretty disgusted with it.

How is that any different from Google? They're both monetizing the content others placed on the internet.

The content producers who distributed their videos on YouTube agreed to Google's monetization terms. They did not agree to Brave's. I cannot unilaterally make copies of Harry Potter books, sell them for a dollar, and give a cut (even a 50% cut to make the Brave analogy complete — they like to say they give a larger cut to publishers without mentioning that it is a larger cut of a much smaller pie) to JK Rowling. Even Brave eventually understood why this was wrong.

Re: Brave Improves Its Ad-Blocker Performance with New Engine in Rust

#68
> was already implemented in heavily optimized C++ handling requests with sub-millisecond overhead, we found that we can further optimise it for a 69x average improvement

Seems like an unnecessary optimization, though I'm sure it was fun. I wrote one a little while back in C++ also[0]. It's not documented the best, but uses a kind of custom tree to match.

0 - https://github.com/cretz/doogie/blob/master/src/blocker_rule...

Re: Brave Improves Its Ad-Blocker Performance with New Engine in Rust

#69

I'm impressed with Brave. I've started using it as my primary browser (after a couple false starts a year or so ago) and it has all the best of Chrome without the worst. There is controversy about how Brave does stuff, and I agree there is some sketchy stuff they are doing (replacing ads with their own). I'm glad they are giving us a choice, though, and exploring different options for sites being able to makes some m…

>There is controversy about how Brave does stuff, and I agree there is some sketchy stuff they are doing

I would be more open to trying/using brave if someone respectable packaged it. As of writing, not a single distro packages it. It would lower my barrier to entry a lot if it were packaged by debian/fedora/f-droid. It doesn't imply anything serious, just that it can be built from source and that it passed the distro's basic diligence and guidelines. This seems like low-hanging fruit from brave's standpoint, and yet multiple issues on their github tracker about packaging downstream have remained open for a while.

Re: Brave Improves Its Ad-Blocker Performance with New Engine in Rust

#70
post #50

Earlier quoted context omitted.

It's potentially even worse than that. Exploiting this grey area in the law will lead to a response from the courts and lawmakers. That response might well require that browsers display pages exactly as the publisher intends. In other words, the end of ad blockers and other tools, thanks to Brave. (Not at all unlikely given that Google seems to want that too.)

> That response might well require that browsers display pages exactly as the publisher intends. IANAL, but I highly doubt that the government will mandate what browsers people run on their computers. Outside of North Korea, of course.

Wait for it; there will be mandated software on 1st-world machines in our lifetimes. Root on your machine is already a bit of a fiction because of government action (the cpu management engine); look for that to become more explicit after the first really big blackout/other catastrophe caused by a computer security failure.

Unrestricted mathematics are too dangerous to be allowed to normal people.

Post reply on HN