"Privacy" is defective when your exploit mitigation is deficient. You've got a lot worse privacy when your whole computer is owned, than when a few websites track which other websites you've been to.
None of this is new. https://news.ycombinator.com/item?id=27219402
Brave has a top-notch team, uses Blink, and has some of the best fingerprinting protection I've seen in any browser, much better than Safari, and better than Firefox. They have good people working on it. The fact that an attempt to improve privacy [0] is being used to bring Brave bad PR is ironic.
[0]: https://github.com/brave/brave-browser/wiki/Deviations-from-...
The link you're referencing talks about NT kernel syscalls - if the computer is already infected it doesn't matter what browser is being used.
If one is a high-value target for a nation state or something of that caliber I would agree with you - but that doesn't apply to regular users. Atleast to my knowledge.
Another counter-argument here is that people looking for vulnerabilities would much rather spend their time looking for chrome exploits than firefox exploits (because of market share), no?
Coinbase was, and that wasn't long ago. If I were a sysadmin I'd blacklist Firefox from all my machines, no matter how I feel about the open web, about Mozilla or their mission statement, it's not a reasonable risk to take.
> Another counter-argument here is that people looking for vulnerabilities would much rather spend their time looking for chrome exploits than firefox exploits (because of market share), no?
People look for both. Firefox is a much higher value target to governments than Chrome, because that's what Tor Browser is.
Tor is Firefox ESR, which only has big security fixes backported; other security flaws remain in Firefox ESR for much longer. But beyond that, the security fundamentals were never a part of the original Firefox project like they were for Chrome, and it still isn't. 13 years later, Chrome's still making faster progress with exploit mitigation than Firefox, and Firefox is the one that needed catching up. There's a reason why GrapheneOS, which people here trust, don't use Firefox. Check their website for a great readup[0].
On Zerodium, Chrome RCE+LPE goes for $500k, Firefox and Safari. People will argue that's marketshare, but get any exploit mitigation researcher in here and they'll tell you Firefox (and Safari) are fundamentally less secure than Chrome. Price isn't just determined by demand (by governments), it's supply too.
I understand the "cyberpunk" love for Firefox, they're the good guys fighting big G. But that doesn't change what their codebase is.
Sampson (Brave Team)
I really like the odd de-mozilla'd Firefox versions though.
One of the reasons why I also quit this was because some problems with their rewards crypto programs. Basically, this content creator Tom Scott, makes great videos, tweeted some stuff about Brave explaining the issue. He says that Brave users who watched him or visited his website were encouraged to donate to him through the Brave rewards program, when in fact, Tom Scott had not signed up or agreed upon it. Several people donated misunderstandignly and Brave didn’t even say that Tom Scott was not on their crypto platform, refused to either refund the people who donated and thanks to this issue now any content creator who isn’t on their crypto platform has a small text below the crypto payment.
The issues mentioned by Tom Scott (from December of 2018) were wholly unrelated to security or privacy, and instead serve as a case-study in UI/UX and messaging. Allow me to explain a bit in greater detail.
Brave held a token sale in 2017 which resulted in 300 million BAT being added to a User Growth Pool. In 2018 we invited Brave users to direct BAT grants (that is, BAT belonging to Brave) to their favorite content creators. Some users expressed a desire to support Tom Scott (as they should; his content is phenomenal).
Brave's UI explicitly marked verified creators as such, but made no explicit indication when a creator was unverified (we followed a similar model to Twitter, which gives checkmarks only to verified accounts). This resulted in some confusion among our users about whether or not Tom was a verified creator.
Some users chose to direct BAT (again, from Brave's own User Growth Pool) to Tom's YouTube account. At the time, this BAT would migrate from the UGP through to a Settlement Wallet, waiting to be collected by the creator. Again, we're talking about BAT from Brave's own wallet--no user funds involved.
After listening to Tom's feedback over Twitter, Brave took action and made a series of prompt changes. For starters, we began to explicitly identify unverified creators as such (not doing so was a naïve design decision on our part). Secondly, we modified the contribution model such that Brave (the browser) would not send tips to any unverified creator (that is, no BAT would go into a settlement wallet). Instead, issued tips now remain in the user's control for up to 90 days while the browser routinely checks to see if the intended recipient has verified.
We blogged about these changes at the time: https://brave.com/rewards-update/. We made these changes in fewer than 48 hours, IIRC. Tom's feedback was excellent, and resulted in a substantially better tipping model in Brave. Of these changes, Tom shared on Twitter "These are good changes, and they fix the complaints I had!".
I hope this helps!
Sampson (Brave Team)
Sampson (Brave Team).