Slashdot Mirror
Do Slashdotters Encrypt Their Email?
An anonymous reader writes "Many years ago when I first heard of PGP, I found an add-on that made it fairly simple to use PGP to encrypt my email. Despite the fact that these days most people know that email is a highly insecure means of communication, very few people that I know ever use any form of email encryption despite the fact that it is pretty easy to use. This isn't quite what I would have expected when I first set it up. So, my question to fellow Slashdotters is 'Do you encrypt your email? If not, 'Why not?' and 'Why has email encryption using PGP or something similar not become more commonplace?' The use of cryptography used to be a hot topic once upon a time."
...but I might attach encrypted file(s) if I really wanna keep something super-secret!.
Yes, this. If I'm sending anything semi sensitive, I just encrypt a file, usually a PDF, and send the password via another method. I wouldn't use this for anything extremely sensitive such as my recurring fantasy to nuke Washington DC from orbit - but for routine stuff it's fine.
And other people can deal with it. PGP encrypted emails - no way.
Faster! Faster! Faster would be better!
I sign all my email with a PGP signature. No one has ever used it to send me an encrypted email.
You can never know everything, and part of what you do know will always be wrong. Perhaps even the most important part.
Both PGP and S/MIME are end-to-end encrypted. Not very useful for webmail users.
A.
...bringing you cynical quips since 1998
Your computer will be software or hardware bugged.
Carrieriq showed the plain text deep state joy of https efforts on your average open or closed US mobile device.
Sending encrypted mail will just make the NSA more curious.
Sit down with your family, friends, faith group, business associates and work out a few simple comments that can flow into any text.
Domestic spying is now "Benign Information Gathering"
My sig (since 2002/2001) on /. has been "Why arn't you encrypting your email?".
The answer is simple -- there was never a critical mass of people exchanging keys nor was there an easy-to-explain web of trust, nor was there a simple, free reliable certificate authority.
In 2002, Outlook Express offered integrated s/mime encryption + digital signatures. Once you installed your certificate (which, was simply double clicking a .p12 file, and entering your import password), you could encrypt or sign email going out, with a single click. It verified signatures in inbound email too, all in an integrated UI.
No one I knew used it.
Even today; Windows Live mail + Thunderbird offer integrated s/mime encryption. Maybe 1 or 2 of my technically literate friends use it. And of those 2, i think only one persists using it to this day.
Back then, when all I had was my Palm Pilot IIIxe, I thought "Whoa. I hold in my hand a portable computer that I can use to exchange digital signatures with". I even kept my pgp key in a note I could beam to someone, given the chance. Never happened.
Nowadays, even AGP on Android doesn't let me exchange keys with someone meet on the street, on the off change they happen to use it. Secure key exchange would be a trivial problem for today's smart phones (provided the carrier isn't using carrieriq to swipe your data....), but there still is no critical mass to make this worthwhile.
And, with most folks using webmail, You'd have to come up with a hackish way to encrypt mail client side (pgp copy/paste to the clipboard? w/ Rich text? attachments?), or just hand your keys to your provider. Doing the encryption server side would make the service provider an easy target for legal and hacking threats.
It's a tough nugget to crack, and it's not going to be solved until mail encryption is as easy to use as Facebook.
Why aren't you encrypting your e-mail?
Encrypted PDF is tricky. Only the string and stream data of the document is actually encrypted -- all the structural information of the document remains in plain text. The number of pages, the presence of images, size of those images, amount of text on each page can all be easily determined.
If you want to encrypt a PDF, use a file encryption tool, not PDF encryption. It doesn't work quite how you assume it does.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Why, yes. Yes I do. At least for the few recipients that do too. And
all my messages are signed.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iD8DBQFO8AWNUy30ODPkzl0RAr75AJ9qYq94sfL00DZxCb3e1tL/HX4uIACeLlbJ
RYRY0ZwfXoKwpyEJn0JzJ2Q=
=fy5a
-----END PGP SIGNATURE-----
Hell, finance in general is retarded.
I have a "regular password" that I tend to use for everything (about a year ago, I started adding a use-specific suffix so someone who stole one password wouldn't automatically have the rest).
It's a fairly secure one, but it includes a _, a $ and a * (as well as a number and letters of both cases). Linux was fine with it. Windows was fine with it. GMail was fine with it. Slashdot was fine with it. Various mailing lists were fine with it. The only things not fine with it?
Debit card PIN (only four numbers!)
Voicemail passcode (also only four numbers!)
Wachovia's online banking system (wait, what?)
Yep. Wachovia did not allow passwords with symbols. No !@#$%^&* allowed. Just letters and numbers.
Not only did that significantly decrease password strength (it went from 77^x to 62^x for a given length x), it also made it impossible for me to remember my password. I had to write down instructions on how to regenerate it (change the $ to 4, * to 8, and so on by not using shift, and drop the _ entirely). Most people would've just wrote down their password, making it even more secure.
Needless to say, I rarely used their website anyways, as it was unimaginably slow as well as pointlessly undersecure. Just waited for the monthly snail-mail summary to check my balance.
Exactly. Several years ago, I used to sign e-mails with PGP (not encrypt, just sign). At the time, some Outlook Express clients would red flag this, and display a large, glaring warning to readers about PGP-signed e-mails. Despite the fact that the bug was due to Outlook Express, I stopped using PGP... it's not like I could force all my recipients to a better mail client.
"Anyone who [rips a CD] is probably engaging in copyright infringement." - David O. Carson
I use GPG/OpenPGP for some mail and "secure" web mail for other applications. I do not use third party web mail (such as gmail) because I can't control the dissemination or privacy (or longevity) of my mail and while my life is generally boring enough to fit within Eric Schmidt's idea of privacy ("If you have something that you don't want anyone [someone] to know, maybe you shouldn't be doing it in the first place [at least not though a google property]."), I occasionally write a personal opinion of someone I wouldn't want them to be able to Google later or share a business detail that could be economically damaging or embarrassing (or is subject to NDA) and gMail and all other web mail services are effectively public.
I've used PGP (and eventually GPG) since about '94 and my keyring has about 20 people on it: more than 1 new key a year! Alas, 25% of those keys expired in the late 90s. My address book has about 1500 entries. Why so few keys? As the OP pointed out, it isn't all that difficult.
The answer for me is that the model for encouraging encryption has to be more like S-WAN than GPG-like. I'd love to turn on "encrypt everything" and forget it, but I'd get an error message for 99% of my correspondents, so obviously that isn't going to happen. So I set my prefs to reply to encrypted messages with encryption, which is fine, but it means I rarely (almost never) initiate an encrypted thread.
What I'd like is an opportunistic encryption mode where any message to an address in my keyring is encrypted by default. Any message to anyone I don't have a key for gets a nice little .sig file with a brief notice that their mail is insecure and effectively public and a link to further instructions for getting GPG set up.
One annoying problem is that encrypted mail is not searchable. To solve that, I want my client to extract a keyword list on decryption then upload that keyword list to (my own) server as an unencrypted header to enable searching (implemented, of course, with a stop list for words you wouldn't want to appear in the clear even out of context or perhaps particularly out of context).
For the truly paranoid, this list could be a hash list, though you could still fairly effectively dictionary hash fish, but it would provide some security and reduce the easy availability of information. In fact, all headers could be hashed and still generally be searchable (except maybe date ranges).
I also want my server to store my public key and encrypt all incoming mail with it. Of course it is already transported in the clear, but it makes my server less vulnerable. Once the mail has had an index extracted and the body encrypted, someone cracking into my IMAP server would, at least, not find a historical trove of clear-text data. And my friends without keys would get annoying sig files evangelizing encryption.
Not entirely - as you pointed out SSL would secure the connection between the your computer and your server, however the connection between your server and the remote server, as well as the connection between the recipients computer and their mail server would remain unencrypted, so effectively you only have encryption on 1 of 3 links.
Message encryption makes transport encryption unnecessary. I.E. you don't care if someone grabs the body of the e-mail because it's useless if you can't decrypt it. Although I do recognize that I, along with most of the rest of the world I think, consider e-mail an inherently insecure communication tool and treat it as such. If you need to send something secure through e-mail, throw it in a password protected rar file and send it as an attachment.
I believe that would be a "no" unless you consider parading your message past Google, who probably keeps a bigger file on you than any other entity, private. And it might be a worse than that--saying it's only Google that sees the message assumes that Google doesn't decrypt the message in one facility, send it from that data center to another in the clear, then re-encrypt and send to your recipient. Whose to say your mail server is in the same facility as his just because both accounts are with Google?
That's why there are S/MIME browser plugins like Penango for GMail.
I've had PGP for over 10 years, but I'm putting it aside and getting behind S/MIME.
S/MIME has great enterprise support, is built into mail clients like Outlook, OS X Mail, Mozilla Thunderbird, iPhones, iPads, and even has browser plugins for GMail. PGP has none of this, sadly.
If Mac OS X's Mail client automatically supports PGP, it is not necessary to obtain any certificate from an outside source. With an OpenPGP application installed on your own computer -- Mac, PC, UNIX, Linux, etc -- you generate your own certificate. See my http://www.rossde.com/PGP/index.html.
This is a troll - Australian ATMs use 4 digit PINs. (Don't know if longer ones are supported.)
Most human behaviour can be explained in terms of identity.
Pretty sure that's S/MIME, not PGP. Which in my opinion is the most correct of the email encryption options, and has the least support.
The main problem with OpenPGP on mail for me is that due to the unique key per recipient, if you add more than one recipient or cc, you have to encrypt the mail for each and every one of them. If you add some attachments it's pretty sure that you will hit the maximum allowed mail size of some mail server along the way.
Uh, no. It's called "session keys". The content is encrypted with a random number (the session key), and this random number is in turn encrypted with the recipients' private keys. As the content is usually compressed too before encryption, the result may even be a smaller e-mail than without...