A place to discuss privacy and freedom in the digital world.
Privacy has become a very important issue in modern society, with companies and governments constantly abusing their power, more and more people are waking up to the importance of digital privacy.
In this community everyone is welcome to post links and discuss topics related to privacy.
Some Rules
- Posting a link to a website containing tracking isn’t great, if contents of the website are behind a paywall maybe copy them into the post
- Don’t promote proprietary software
- Try to keep things on topic
- If you have a question, please try searching for previous discussions, maybe it has already been answered
- Reposts are fine, but should have at least a couple of weeks in between so that the post can reach a new audience
- Be nice :)
Related communities
Chat rooms
much thanks to @gary_host_laptop for the logo design :)
- 0 users online
- 57 users / day
- 383 users / week
- 1.5K users / month
- 5.7K users / 6 months
- 1 subscriber
- 2.97K Posts
- 74.6K Comments
- Modlog
This article makes some good points generally, but it is ultimately marketing for a commercial snakeoil service which has a gigantic backdoor in its very threat model: when a tutanota users send an “end to end encrypted email” to a non-tutanota user what actually happens is that they receive a link to a web page which they type the encryption key in to.
Even if the javascript on that page is open source and audited, it is not possible (even for sophisticated users) to verify that the server is actually sending the correct javascript each time that a user accesses it. So, the server can easily target specific users and circumvent their encryption. The same applies to tutanota users emailing eachother when one of them is using the webmail interface.
This effectively reduces the security of their e2ee to “it works as long as the server remains honest”. But, if you fully trust the server to always do what it says it will, why bother with e2ee at all? They may as well just promise not to read your email.
I am removing this from !privacy@lemmy.ml with the reason “advertising for snakeoil”. (If you’re reading this on another instance and the post isn’t deleted, ask your instance admins to upgrade… outdated versions of lemmy had a bug which prevents some moderation actions from federating.)
i’m genuinely curious about some alternatives to this sort of surveillance to solve issues like CSAM etc., which aren’t “it’s the parents responsibility”. section 230 reform? links to further reading appreciated.
But we have to think of the children.
Two details:
FBI expands rapidly to DHS and then the entire US Police State. If you cross borders, expect ICE AND CBP to be up in your body cavities. If the local county sheriff doesn’t like you, or you’re being stalked by an officer (say, an ex) expect them also to have access.
When you think Hackers think of not only data mining interests like Palantir but also industrial spies. If you have any business interests on your phone subject to an NDA (or you’re motivated not to share because reasons) these guys will sell that information to your competitors, if they weren’t hired by them in the first place.
If you run more than a mom-and-pop then the default security of your smartphone is not enough. But a lot of sizeable companies supply their officers with unprotected phones.
Slightly off topic - someone mentioned they don’t use Tutanota for social interactions because the domain is weird and I agree whole heartedly. Everytime I’m on the phone with a support dept. or tell my friends and I spell it out I feel so silly. Not to mention my wife has gotten it wrong several times.
Love the solution, their support is responsive as well, but yeah…
That’s the main reason I didn’t even consider them. “Proton(mail)” just sounds more professional when used in actually important contexts and is easier for people to get right.
In general, I’ve noticed that a lot of privacy focused software, particularly FOSS, are really bad at choosing names which make people want to use them. They tend to have names which might appeal to some crypto-nerds, but which make them sound just weird or questionable or niche to the average user. Like (the precursor to) Signal the messaging app used to be called TextSecure. There’s no way I would’ve gotten my parents and siblings to use something called TextSecure. The name just sounds so geeky and niche.
Tbf Google is also a weird name, Yahoo was also a bit weird even if not entirely, there probably are more examples but it’s not just that the name is not great but also that these things aren’t advertised as well
Custom domains are pretty cheap, and you can hook them up to Tutanota easily. Then you never have to spell it for anyone.
That’s certainly true, but to catch mainstream it’s still a decent gap. My mom struggled even signing into Gmail. Think about their patience as you describe the process to configure DNS records? All while there are free options available.
For us privacy focused folks it’s viable and I might consider it too.
Tutanota isn’t suitable for many people. You can’t use standard email clients, screen readers don’t work with it, pinch to zoom it’s kind of iffy.
I’d say for people who are either technologically inclined, or working for a company with a help desk to explain how to use it, it’s fine. Still got some rough edges. But it’s definitely not suitable for the elderly computer users
Leaving a back door in is the same logic as leaving a key under a fake rock by your house.
That you as the home owner don’t know about and anyone with a home constructed by the same people whom did your home have the exact same key under the exact same rock.
Uhm question, how is Tutanota E2EE? Other than making PGP setup easier. Afaik they just use a different protocol for client-server
it is a shitty E2EE implementation in JS incompatible with the email standard OpenPGP.
but I like that they wrote this post, even if it is for marketing purposes, because Tutanota is based on the EU and hopefully the EU Parliament will listen if enough people tells them.
They have a JavaScript version, it’s true. But they also have apps. Meaning you don’t have to rely on JavaScript security. If you want to lock it down.
Encrypted email, should never be considered end to end encrypted. This includes protonmail which does implement PGP. Email is a clear text protocol. Encrypted email providers provide encryption at rest for the email.
The issue with protonmail, and PGP in general, is the metadata is unencrypted, to from subject. Metadata gets people killed. Metadata is valuable data.
So you have to choose for your data at rest do you want everything encrypted, then you go with Tutanota, if you only want the body of the email encrypted then proton mail/ PGP.
Since most email is clear text anyway, and if you want end to end encrypted you should use signal or simplex, I think full encryption at rest is the better option here.
All of that’s to say it’s not a shitty implementation, it’s an implementation with different trade-offs than what you value
their clients use the same JS implementation, they are the web version wrapped in electron.
The major problem with these JS implementations (including Proton and any other program that uses JS for encryption) is that it would be trivial for them to grab your private key from your browser and send it to their servers. And yes, we have the code. But it’s virtually impossible to verify that the code they are sending to your browser each time is exactly the same one that they publish on github, after JS minimizers and all that.
A third party that found a vulnerability in a browser could also inject their own JS and steal your private keys.
You’re obviously right about everything else and email’s inherent insecure nature.
I still find it useful because it’s the only online communication channel that is widely adopted, that can be self-hosted without depending on third party servers or you can simply choose a provider you trust. I’d love to have that with XMPP or SimpleX or something like that, but currently we’re stuck with email.
The point of a fully embodied app, is you don’t have to pull the JavaScript from the website. It’s distributed via the app system. Fdroid in many cases
yes, the clients should be good in most cases, as long as builds are reproducible or you compile yourself from the public code (which is not most cases).
Still, I’d rather do OpenPGP encryption on my client of choice with my implementation of choice that is provider agnostic.
Fair. Glad you found an email system that works for you, PGP is great.
They are in the f Droid repo, which means fdroid does build them from the source code.
I just don’t think tutanota is shitty, they’ve just made different trade-offs.
I personally don’t like Tutanota for a lot of reasons. The other day I recommended Tutanota to someone that needed a new email account and they weren’t able to create the Tutanota account using Tor. They tried using a VPN and they weren’t able. Tutanota said their IP address was being used for abuse.
What’s the point of a private email if you block anonymizers?
Some people might find a use case for it, of course. And their post advocating against anti-encryption laws is good. But I don’t think it’s a good email provider and I won’t be recommending them again.
They don’t use PGP.
Tutanota is end and encrypted between different users of tutanota. But any external email you send or receive is unencrypted. They do have an option to send an encrypted link to the other party, but that’s cumbersome.
The big thing about this mail service, is the data is stored at rest encrypted with your key. So once it’s received clear text, it’s encrypted and stored on the disk only with your key. After that they can’t decrypt it.
Those famous good guys we all know and love.
It’s the same person.
Love the concrete examples and then “etc.”
There are no good guys. Only rotating members of the aligned interest gang.
Nailed it.
Can’t believe this is real. But on a mass scale we can use FOSS with full encryption.
not if governments are compromising the encryption algorithms themselves.
edit: source https://slashdot.org/story/420213 https://www.newscientist.com/article/2396510-mathematician-warns-us-spies-may-be-weakening-next-gen-encryption/
They did this before with the eliptic curve cryptography, and we knew it had this problem before it was implemented as a standard.
So if the NSA offers a standard, don’t trust it and include in your encryption software the option to use something different.
Good luck with that. FOSS is transparency on a source code level, there’s no obscurity they can hide their back door behind.
https://slashdot.org/story/420213
You realize nobody would know about this in the first place if it was Proprietary, right?
FOSS allows for whistleblowers, scrutiny, and audits. Proprietary ‘security via obscurity’ does not.
I’m perfectly aware of all that. but cryptography is an extremely complicated discipline that even the most experienced mathematicians have a hard time to design and scrutinize an algorithm, they heavily rely on peer review. If one major institution like the NIST is biased by the NSA, they will have a bigger chance of compromising algorithms if that are their intentions.
You’d be surprised what the world wide collective of Cryptographers are capable of when they’re able to scrutinize a project in the first place. Which would you prefer? A closed unscrutinizable encryption algorithm or one that’s entirely open from the ground up?
NIST could do damage if they’re biased, but it’s not like people aren’t keeping a close eye on them and scrutinizing as many mistakes as possible. Especially for an algorithm as globally important as PQC.
I’m totally against anything proprietary. That’s the first requisite for anything I use. And I’m not advocating for proprietary algorithms at all, that would be very much the demise of encryption.
I’m just worried that a sufficiently influent actor (let’s say a government) could theoretically bribe these institutions to promote weaker encryption standards. I’m not even saying they are trying to introduce backdoors, just that like the article suggest they might bias organizations to support weaker algorithms.
AES 128 bits is still considered secure in public institutions, when modern computers can do much stronger encryption without being noticeable slower.
It’s the fact that the intelligence agencies have proven themselves to be unable to responsibility use their powers, and instead find every sneaky way possible to infiltrate and spy on their own citizens while preventing nothing. That’s what has pushed the world to say enough is enough and we are going to encrypt everything we can. Now the global powers are crying poor about how they need access to stop terrorism, while being completely unable to point to a single instance where they stopped a terror attack and contrarily there’s plenty of terror attacks that were never stopped.
removed by mod
Government’s : We’re going to counter terrorism by backdooring into every device our citizens use.
Real Terrorists : We have our own devices that are fully encrypted and free from backdoors.
Thanks for giving us the key to spy on all of your citizens tho, very helpful.
deleted by creator
The recent advent of governments worldwide trying to force corporations to build backdoors into their services for the ‘safety of children’ or to ‘counter terrorism’ arguably does more harm than good for the common people.
It’s never about the public interest but the preservation of plutocratic power.
We’re going to ‘counter terrorism’ by mass spying on our own citizens and hope to god real terrorists don’t gain access to our backdoor.
The point of the exercise is ALWAYS to spy on innocent citizens. It’s about surveillance and control, not countering terrorism or protecting children.
if we spy on their spies, we can counterspy their spyworks before they spy on us!
… and it’s called “defensive measures”, sir.
/s
All mail should force to be postcards, so that no secrets may go unchecked.