Easy to use email encryption

So tobacco, isn't that exactly the problem? If he has five emails, why? I mean, I know he has his reasons, but he only loves the Hotmail; the rest are overflow for particular matters that probably should be secured anyway. And overall, if the user wants to secure at all (and you said he did), then that's really evidence that he's not happy with his current setup. So ideally, he can secure his chosen webmail with InstantCrypt or some friendly software for use with you (because you're the only person he knows that's its been an issue with...I'm assuming you're part of the equation) and then he can use a web-based encryption for those correspondents who aren't ready to deal with desktop encryption.

I think screenshots are vital learning tools. It guilds the user into the machine. Tor needs a meta-level tutorial, I think. I haven't found one.

JT, I was only kidding about pushing for Mackeral. You've supplied links for two great papers (yours and Whitten) which are keeping me plenty busy with some interesting thoughts. But still, I'm very interested in Mackeral.

Warren Weaver said Shannon's information theory was "sufficently imaginatively motivated" to deal with the "inner core" of the communication problem, and the evidence of this was that the theory contributes to the basic theory of cryptography.

I think we're starting to see some of the implications of that statement.

 

What interests me about InstantCrypt is the goal towards simplifying the user interface of e-mail encryption. (This is a really tough goal, and I hope my referencing of the HCIsec work will be of some assistance.) Not only that, but it [InstantCrypt] strives to:

• retain convenience, by working with existing e-mail clients (webmail included)
• minimize the number of cryptographic decisions the user needs to make
• maintains interoperability with other applications that use the OpenPGP standard

Convenience trumps security every time, so it's vital that any real-world cryptographic application takes this into account by being as transparent and seamless as possible. Users expect security; they don't want to work for it. They shouldn't have to.

No worries. I was in a bit of a hurry and completely forgot to include the rest of what I was going to say. (Due to my hatred for spontaneous browser crashing, I typically type posts in a text editor, then copy and paste.)

Green Cryptography focuses on assembling the implementation the right way, such that the functions of encryption and authenticate will be appropriately handled, and more easily realized, by developers. Mackerel focuses on applying the implementation to the right stuff, by aligning current cryptographic models with real-world communication models, and adds a layer of abstraction on top of a green cryptographic model, such that the concepts of confidentiality and integrity are more tangible and palatable for users.

To tie this into the thread's topic: Just as InstantCrypt aims to better the interface between the implementation and the user, so does Mackerel aim to better the interface between cryptography and its stewards -- cryptographers, developers, and users.

 

Well, JT, we definitely need a Mackerel attack to simplify this process, because the longer the week went on, the more I had to agree with tobacco. My experiments with InstantCrypt this week, while heartening, still leave me with a bridge to gap.

My recipients just hate key exchange. It's counter-intuitive and they don't know what to do with the key downloads. I think InstantCrypt has done a great job with setting up the sender with a key and enabling the sender to enable encryption. The flaw in the system is that the InstantCrypt sender (who is probably a novice) needs to then guide a recipient through InstantCrypt. And so on, and so on, and so on. The recipient always comes to encryption cold. This is not InstantCrypt's fault in any way. But the software ideally has to be really friendly, enough that a recipient can both learn and teach in short order. InstantCrypt is more geared toward a second-tier type user (like the sender.)

Also, practically speaking, if the software is not super simple, the numerous emails and phone calls needed to talk the recipient just through key exchange (all unencrypted communication) leave a vast trail. That's where it gets silly. That's what I meant above about "you're already communicating with the guy." Yes, about how to have an encrypted communication If the proto-communication (about encryption, and how, when and why it should be enabled) goes on long enough, you start to feel like you might as well just meet for drinks and hand him the data. I'm joking, but only sort of.

To return to the Hushmail problem, here's what Wired said:

"The first time a Hushmail user logs on, his browser downloads a Java applet that takes care of the decryption and encryption of messages on his computer, after the user types in the right passphrase. So messages reach Hushmail’s server already encrypted. The Java code also decrypts the message on the recipient’s computer, so an unencrypted copy never crosses the internet or hits Hushmails servers.

In this scenario, if a law enforcement agency demands all the e-mails sent to or from an account, Hushmail can only turn over the scrambled messages since it has no way of reversing the encryption.

However, installing Java and loading and running the Java applet can be annoying. So in 2006, Hushmail began offering a service more akin to traditional web mail. Users connect to the service via a SSL (https://) connection and Hushmail runs the Encryption Engine on their side. Users then tell the server-side engine what the right passphrase is and all the messages in the account can then be read as they would in any other web-based email account.

The rub of that option is that Hushmail has — even if only for a brief moment — a copy of your passphrase. As they disclose in the technical comparison of the two options, this means that an attacker with access to Hushmail’s servers can get at the passphrase and thus all of the messages."

Sorry for the long quote, (and I have always maintained that Hush is better than nothing, even given this flaw) but I wondered if anyone can clarify the function of the java applet here and in the discussion below, re: the Cryptoheaven client non-encrypted contact access, per discussion below. Am I to understand that this also comes into play with the Cryptoheaven web edition?

Cryptoheaven might be perfect if it were free However, payment presents issues. No complaints about value. That's obvious. But it requires commitment that might not be practicable for short-term correspondence, nor appropriate to ask for short-term business contacts. And payment pierces anonymity, should that be a concern.

Nevertheless, I'll experiment with the free trial. The encryption itself is impeccable and besides the payment problem, it looks otherwise close to ideal.

 

Hello Nix

When you get right down to it, InstantCrypt and for that matter - Thunderbird/Enigmail/GPG4Win is not really that complicated if the want is there for securing email and patience is shown doing alittle reading and familiarizing oneself with the programs. It's not! But therein lies the problem - "Patience". Most users want it easy - click&go, click&go!

The java applet ensures the encryption/decryption process takes place clientside which is secure and from my understanding, can only be breached if you are targeted with a specialized applet designed to capture your info . Not perfect but better than anything done serverside which is highly insecure.

I've been experimenting with Cryptoheaven for a few months now (gave me a 3 month trial) and it's a great and highly flexible service. The ability to interact with non-secure users is a plus as long as the recipient doesn't mind following a link to the secure email/having the applet load and creating a password during "first" use which will be used for any further correspondence. They can reply "securely" to the CH user with the applet. Any non secure email sent to a CH user is obviously "non-secure" but when hitting the CH server, is automatically encrypted (AES-256) with the user's public key.

The CH user is given the option at first use of private key storage clientside or serverside. CH support also showed me how to make CH portable which comes in handy for not only flash drive use but also if you wish not to have "java" installed on your computer.

If interested, do the following: Remove all instances of java from your machine. Then download and install the CH client. No java installation will be detected so it will download and install java "in the CH folder only". Now, copy the install folder to wherever you wish to run it portable. If you've created your keys already, copy it there as well. Now uninstall the CH client from add/remove programs and your set. You now have a portable version with it's own java ready to use anywhere with your private key still clientside. However, there is a config file which gets written to the documents folder that needs to have it's path changed to be written to the portable folder. Not sure how to do this yet

Another option which is free but closed source is TrulyMail. Works somewhat like CH and in version 3 which will be in beta in about a month's time, will offer expanded abilities like CH to deal with non-secure email users.

 

I put this question to CH support for you and this was their response:

 

I've misspelled Mackerel a number of times. I won't be able to correct it when it appears as a quote, so I apologize to JT.

tobacco, thanks for asking about the prepaid card. I'm interested in CH for a number of reasons and will initiate the free trial tomorrow, as it appears it is still available.

O!polis is something a little new. I like their email options, i.e. the meassge control options, some of which mimic some of the old read-notify features like message destruct, and notice when message is forwarded, etc., along with a few others like the print notice, etc.

Here's their paper on message security:

http://www.opolis.eu/securemail_TheOpolisOnionModel.html

Thoughts, anyone, on either security or ease of use or (hopefully) both?
It's free, by the way.

 

It does look secure but after watching their setup screencast here - http://www.opolis.eu/video_create_01.html i decided to give it a pass as the setup process is rather involved. And you know my contacts

Secondly, it asks during the setup for too much - way too much personal information for my taste. It uses this info to create a verification certificate. However, think this service would be great for business use though.

 

Agreed on all counts, mr. tobacco I have privacy concerns with this one. I went through sign-up last night myself.

 

What would be "easy to use email encryption" IMO is a firefox & explorer addon that would encrypt and scramble text without key pairs, only using passwords.

Why?
-browsers are widely used for email-ing by the common user and addons can be added and removed in one click (if you bother with 3rd party software for encryption, i doubt that receiver will want to download extra software to wrestle with it.)
-no keypairs means that there is no need to send any additional data, no extra work for "don't-bother-me" user, less chances to mess something up in the encryption-decryption process.

Downfall of this suggestion is that there are many addons that promise text encryption but they rarely use aes, they are rarely transparent and don't use any open source encryption protocol. If there are any programmers that could to this sort of thing, i would give it a two thumbs up.

 

As far as passwords go, you could try this:

- type your communication in a text file and then use something like Winrar to create a password protected, self decrypting archive.

- use the free email client from DreamMail - http://www.dreammail.eu/intl/en/home.html

Both parties need dreammail (to decrypt) but emails can be sent password protected. Portable version also available.

- give these options a look see:

Ecipher - http://download.cnet.com/eCipher/3000-2092_4-10842835.html Can be used as a limited free version but hasn't been updated recently so don't know if it works on 7 or 64.

Zsentry (web based) - http://www.zsentry.com/ Free basic version.

 

Thanks for hint!

Thank you, Justin, for that hint. I downloaded the paper, looked very interesting. And thank you for your interest in general.

InstantCrypt