Yeah, that along with their improper communication does make Firefox seem like a bit of a lost cause. I do still hold out some hope, as they have begun to rollout more features the past couple of months, but I am cautious due to the apparent push for LLM integration, which has been opt-in so far, but may go awry in the future (the local AI stuff, like translations and now auto tab grouping seem fine).
On a separate note, Brave is better security-wise because of the stronger sandboxing, but I feel like I have to go out of my way to make it feel less bloated, while Firefox with Arkenfox is simpler.
Browser choice is mostly about choosing what’s right for your use case and threat model, so the return of “honorable mentions” would make sense to me.
Yes Firefox is a lost cause at this point and we’re only using it because Chrome is even worse (more invasive and bloated, less configurable, chrome web store is a mess). Hopefully Servo or Ladybird will become stable one day.
I’m not a fan of this. “Honorable mention” is confusing, what does it mean? Do we recommend it or not? We agreed a while ago that we would just stick with something is either recommended or it’s not.
As far as the suggested browsers go, Firefox is the upstream to all of these but Brave.
Brave is good because they actually add a lot, Chrome doesn’t have fingerprint randomization or tracker blocking or really any attempt at anti fingerprinting whatsoever. With the Firefox forks it’s mainly Firefox with a coat of paint on it.
The Tor browser and Mullvad have involvement from the Tor project and they have years of research and even upstreaming improvements through the Tor uplift project. Librewolf and Zen browser really just use already built-in Firefox features and all they do is add another party to trust and some delay for updates. I don’t really see the value in recommending these. Especially if we’re going to remove Firefox then Arkenfox is going to be extremely confusing. People are going think it’s a browser instead of a config for Firefox.
I really don’t want the browser page to just be a million Firefox forks, I don’t really see the point when we can just recommend Firefox and show how to configure it for maximum privacy/security.
The thing for forks like Librewolf, like Mullvad is that these are configured out of the box for the most part, not all users wanna go through that hassle of arkenfox and hardening, etc. and that’s a fact.
I don’t see how moving arkenfox from the “advanced” category of the browsers page to a required modification under firefox adds confusion. You can then remove most of the recommended gui configs as arkenfox does it all by default.
If you really want to remove any confusion you can just list “Firefox with Arkenfox” or a version of that as the browser name.
Not everyone wants to use vanilla Firefox, own a Google Pixel phone, or use centralized messengers like Signal.
I think his point is that these browsers wouldn’t necessarily be recommended but they could still cater to specific niches. As GorujoCY said not everyone is going to go through the hassle of arkenfox and it’d be easier to install a browser that does it by default, so I’m all for adding LibreWolf as an honorable mention but not Zen Browser since LibreWolf appears to be more privacy-focused and we don’t need to list every Firefox fork. It could also be worth mentioning projects that meet the criteria but bad usability (off-topic but I can think of two not already recommended Linux distros that could easily be honorable mentions in this case), and it’s easier and more convenient to install a browser than it is to modify a browser with policies and user.js.
Besides, I don’t think security is a black-or-white thing where everything is either super secure or wildly insecure, and we need assurance that just because something isn’t recommended doesn’t mean it’s the end of the world if you still use it anyways and that it could still be better than what Google, Apple, and Microsoft offer.
Instead of a “my way or the highway” approach, it’d be better to list a best approach than some alternatives for those who can’t or won’t (ideological reasons probably) follow that approach. Most people are guided into making the right choices while the rest of us can still freely oppose Google and other large tech companies based on ideology without huge sacrifices of security.
I don’t see where recommending LibreWolf makes any sense. If you care enough about fingerprinting, use Mullvad or Tor Browser. For logins, take the time to read the arkenfox wiki and install it yourself.
If you plan on using LibreWolf, you’d have to read their documentation anyway—and at that point, why not just use arkenfox and read the arkenfox wiki?
You’d be adding trust to another party just to maybe save a few minutes of reading, which you end up doing anyway when something doesn’t work.
The same reasoning (“open-source” criterion aside) can be used to include Windows in the OS recommendations list because it can be configured for privacy as well, by disabling all telemetry.
Unmodified Firefox is not a good option if you care about privacy. At the very least there should be a big fat warning in the FF section about this and that the configuration below is a must not just a “recommendation” if you care about privacy.
This whole paragraph… I don’t even know where to start…
This feels like an absolutist position to me. I would rather someone be on Firefox with little to no changes than have them be on Google Chrome. It feels like we are letting perfect be the enemy of “good enough” or “significantly better”
The main problem is that we no longer have any Firefox-based browsers that don’t require active maintenance.
Previously our stance was that Firefox was “good enough” out of the box, and you could optionally improve it with Arkenfox. This covered both users who want something reasonable with low hassle, and users who could keep up with more advanced maintenance from the Arkenfox wiki.
Nowadays it seems to be generally considered in this community that Firefox isn’t good enough out of the box, and Arkenfox is essentially mandatory, so our recommendation is only good for the second group of users.
Unless we don’t think that’s actually true. If we instead agree with what @gregandcin just said above, and Firefox is still “fine” for people, then our existing recommendations are fine and no changes are necessary.
Just to follow this up, Firefox is not perfect, but I don’t even run Arkenfox on my usual install. I’m not a privacy absolutist, but this community helps me understand what is and what is not worth my time and money, which is why I think Firefox is a great entrypoint for someone who wants an alternative to Chromium-based browsers.
I will say, LibreWolf and Zen Browser live in the same world as Brave for me. They are LITERALLY forks of an upstream browser that require the downstream maintainer to update as fast as they can on upstream. I don’t see why we treat them as if Brave doesn’t also have some longer-than-wanted update cycles
I don’t think that this point has really been demonstrated or even argued by many. We already provide tips for some basic hardening that together with uBlock Origin can go a long way, and privacy-wise would probably be pretty similar to Brave. On the other hand, using Arkenfox would arguably go even beyond Brave on this regard, so I don’t know why the current recommendation + the recommended settings wouldn’t be good enough anymore.
I agree with the sentiments of this, but “good enough” and “better than Google Chrome” should not be the criteria to get recommended. If that was the case we could recommend countless more browsers such as vivaldi which at least have more features than Firefox.
If someone is taking the effort to look on this website for an alternative browser, I can’t really imagine a senario where we would want them to pick non-hardened firefox. Brave is more usable, easier to setup to a privacy respecting level, and has basic browser functionality like profiles. Arkenfox is a step up, trading much of Brave’s funtionality for more privacy. Mullvad is a another step up, getting even more privacy for less funtionality. I don’t see why we would recommend a browser with both poor funtionality and poor privacy like unhardened Firefox and why we would want someone to use it.
I’d argue the contrary, Brave has more security, allowing it to enforce more privacy.
If you’re referring to RFP fingerprinting protections, that’s not enabled on Arkenfox by default, and I don’t believe any fingerprinting experts have recently said that RFP is better than Brave’s fingerprint protections.
I agree not every browser better than Chrome should be recommended, but again there are different use cases and niches that are actually relevant here. If someone wants to use Firefox, they should either harden it themselves with Arkenfox (more timely updates but requires constant maintenance and configuration) or use LibreWolf (hardened out-of-the-box, little maintenance required, but risk of slower, unreliable updates). Unless Zen browser offers something over LibreWolf I don’t see why it should be an honorable mention.
Also, Vivaldi isn’t better (from a security and privacy perspective) than Google Chrome.
At least from the privacy perspective this is just plain wrong, since Vivaldi doesn’t collect much information from their users, when with Chrome Google collects everything. I don’t know how you can seriously make this kind of statement.
What ads? Do you mean their bookmark partners perhaps? Those and everything else can be disabled, which is something one cannot say about Chrome and Google’s data collection. I would also say that Vivaldi’s business model is miles more ethical than Google, whose business model centers around your data.