I don’t think he even read the link he post… The Proton definition is the correct one and used by the whole industry including mullvad. It is just some extreme people here trying to define differently, but this should be discussed in a different thread.
Thanks for all the info. I really hate when some provider makes false ads like they offer something but they actually don’t. Mac things does not concern me directly because I am using just Windows and Android but the point isn’t really about just Mac is not working but it seems that it won’t even get fixed and Proton blames something else than themselves. Mullvad does the job right so that is really lazy claim from Proton.
It is really “fun” to read about this kind of untrustly posts when I have spent last week or two to transfer every information that I have to Proton’s servers. Most time I spent to create different email aliase to every site that I have registered that allows email change. I use Proton Pass and I use Proton Mail now. I have used just VPN but I subscribed to Unlimited and started to take full advantage of everything they provide.
After reading these things, I think I should have stayed with Bitwarden in passwords and just change VPN to Mullvad which I have used before Proton. I changed Google Calender to Protons version and started to use Proton Drive alongside with OneDrive since I don’t want to out of Microsoft just yet.
But yeah, I think I shouldn’t trust Proton, but I just don’t have the energy to switch and change everything again. And obviously I made 2-year deal. 
This is false. Lockdown mode, according to the Mullvad docs, is intended to block internet access when the “disconnect” button is clicked or the app is not open. Changing servers is covered by the kill switch alone. Also there is nothing “extreme” about not wanting your IP to leak when changing servers. I and I think most would consider this basic functionality for a privacy VPN.
I don’t intend to weigh in on the conclusion of whether Proton VPN merits staying on PG or not, or if it’s a good VPN. But I’d also like to say that ProtonVPN’s iOS kill switch does not work as advertised either.
Are you a network engineer, a Linux developer, or a network software engineer? These are quite distinct titles and someone with any of them would know that. You seem to claim whichever one best supports the argument that’s about to follow.
If you were as knowledgeable as you purported yourself to be you’d know that making an extremely robust killswitch on Linux is actually trivial using network namespaces. It might be hard to make a universal solution in this way for use in a VPN app, so it’s understandable why none use this method, but to claim “there’s no such thing as a kill switch on Linux” indicates if you’re not outright lying you’re at the very least not at the place on the Duning Kruger scale you think you are.
I would expect a kill switch to disconnect my device from the internet if the VPN isn’t active for whatever reason, including my intentionally disconnecting it. I don’t think that’s an extreme view at all I think it’s a basic user expectation.
Same here. Having it work any other way from this makes it essentially worthless from an OPSEC point of view.
Maybe not.
If we take another Briar comparison:
Imagine if Briar had a button that said “disable encryption”, would you be okay with messages being sent unencrypted after clicking that? Yeah, probably. Seems strange that would be a possibility, but still.
Imagine if Briar sent messages unencrypted for a few seconds when you switch from one contact to another. Would you be okay with that? Hell no, PG would be pulling it from their recommendations instantly over that flaw.
So why is it so different for Proton? And why are you focusing on what is ultimately a design choice that disconnect truly disconnects and sends you into cleartext, instead of the unacceptable leak when you switch from one server to another?
I find the comparisons with encrypted messages or files somewhat absurd, since this depends entirely on the communication between client and server, without the OS being involved at all. There will always be leaks under certain circumstances: on macOS, Windows (see Your VPN Kill Switch Won’t Stop All Leaks) or anywhere else (almost).
Some argue the bar should be higher, as in the case of Mullvad and iVPN. Others argue the bar is fine and therefore ProtonVPN should be kept. But no one is claiming that leaks don’t exist, and you keep debating this with examples that don’t help focus on what really matters.
Not to mention the tone, which in my opinion is excessively paternalistic.
The points being made for this is: Would we accept similar dealbreaker flaws in other forms of clients? No. The answer is no.
The fact that leaks aren’t preventable in every single scenario doesn’t mean that we should ignore when a VPN client does not do all it can feasibly do to prevent them. At a minimum there should be a warning for Proton specifically that it can leak in scenarios where the other two options do not. Hence, Proton is the inferior option among the three. And if Proton has something that can reasonably be fixed by them, why shouldn’t we expect them to do so before making it back onto the list? This is fixable. Hence they should do so to keep the standard.
From what I can tell, the “others” are probably in (vocal) minority in this thread.
The “some” argue in favour of 2 bars. One for the OS to meet, and the other for VPN clients to meet. This means, PG keeps its current criteria and the current wall of text in its VPN page (which today makes a number of client-side promises on behalf of public VPN providers it recommends).
The “others” who are in the favour of ‘lowering the bar’ are invariably going to prompt a larger discussion on whether VPNs should be used “almost certainly” when it can’t prevent leaks & traffic may end up in plaintext, with the adversary (the ISP) who is squarely in a public VPN’s primary threat model (as PG markets them today). And this is the place where @Ganther’s comparision to encrypted apps comes from.
This is a misconception. It is intended to be accessible for everyone, including non-technical folks. But it’s audience is everyone.
At this point, a poll might be a good idea to see where everyone stands. Results will be displayed in a week, when the poll closes.
Choice about Proton VPN
- Proton VPN should be removed until its kill switch works well (same level as Mullvad)
- A warning should be added
- The killswitch criteria should be relaxed or removed
- I don’t know
- We should wait a bit more until we decide
- Nothing should be changed (no warning)
0
voters
The issue isn’t really that ProtonVPN’s killswitch doesn’t work as well as Mullvad’s, it’s that it doesn’t meet the minimum criteria PG has established for listing a VPN. This feels like a weird reframing of the issue at hand.
Yeah I feel like people are forgetting that this is the central issue.
The craziest part is it has been 3+ months from the issue being posted and no new criteria has been seriously considered and Privacy Guides doesn’t even warn users on its VPN page that Proton doesn’t meet the criteria.
I am not sure why its so controversial to want Privacy Guides to remove a tool that does not meet the criteria.
You keep saying the same thing. Mullvad also has leaks on Windows. Should it be removed? Or should a warning be added that leak X occurs in situation Y? You keep repeating like a mantra that Proton doesn’t meet the criteria. The point is that no provider meets those requirements, specifically regarding a perfect kill switch.
I’m in favor of either removing Proton from the recommendations or adding warnings. But saying it doesn’t meet the criteria when no other provider meets them either is outright lying, and the debate goes off the rails.
What? The issue over other tools not meeting the criteria doesn’t diminish the fact that Proton fails to meet the criteria.
Any tool in any category that does not meet the minimum criteria should be removed, period. If PG is not going to enforce its own criteria, then there is no point in having them
The Proton team has a pattern of reaching out to communities during controversies like this. It may be worth bringing this to their attention. Maybe they can provide some much needed insight.
Its equally important to review criteria when needed.
Totally agree.
When a new criteria is created and approved, then we can reconsider options like Proton and, see if they meet that new criteria. This has nothing to do with the current criteria though, which Proton has not met for months already.
Considering how long it can potentially take a new criteria to be approved (there isn’t even a thread suggesting an alternative right now AFAIK) it would be negligient to leave tools that do not meet the criteria up.
Just look at the uphill climb users have faced trying to get open source to be a requirment for Password Managers as an example of how drawn out a criteria change can be. Thats after @jonah defined what open source is for everyone, which I suspect he might have to do for “kill switch” considering past replys.