I had a bit of a rant about this on IRC the other day so I'll post a cleaned
up version here of me reading between the lines:
> Own-Mailbox is a personal email server you can run in your own home
You better hope it doesn't break or have downtime, and trust that your home
network is secure enough to handle anyone intruding through this internet
facing service.
> Your email address will be like [email protected]. The domain name
name.omb.one will belong to your for free for life and will automatically
point to your Own-Mailbox, even when you change IP.
Once we go out of business you're screwed. You also have to trust us not to
proxy and snoop your traffic.
> Some ISP will block only port 25, but allow to forward other ports. In this
case we will offer a free port tunneling service for 5 years for port 25.
So we'll actually proxy your traffic maybe. Just trust us not to snoop?
> you can seamlessly send and receive encrypted emails
Great, you won't know when you're not encrypted.
> through a smartphone app, or using external email software
Man it'll sure be easy making all this software work together properly.
> Own-Mailbox automatically encrypts your emails using Gnu Privacy Guard, a
strong encryption software. This is the same software that has been used by
Edward Snowden
If we use GPG as part of software that makes the rest of it secure, right?
Also Edward Snowden used some software we're going to use, so by extension it
makes our engineering practices secure.
> Own-Mailbox allows you to send and receive 100% confidential messages even
with people who don't use email encryption yet.
Impossible, you both need to be using authenticated encryption to have
confidential messages. If it's not authenticated, it's easy to attack. I
might want to add here that the project throws around that '100%' figure a
lot. 100% open hardware, but it uses an Allwinner A13 board, so you still
have a nonfree chipset. It says 100% free software too which is kinda weird
given Allwinner has a history of GPL infringement, though they might be able
to avoid those parts given how the device just needs networking.
> For this purpose we introduce PLM, a new technique that allows you to send
a filtered and temporary HTTPS link to your contacts. This link points to
your private message hosted on your Own-Mailbox.
New crypto techniques? This isn't good either. Who hosts the link database?
> The link is temporary: once clicked by your correspondent it is too late to
spy, the link does not work anymore.
Or if someone's man-in-the-middling it they can just show it before it was
destroyed.
> The link is filtered by a question. Depending on the level of surveillance
you think you are in, the question can be a simple captcha to avoid bots, a
secret question that your correspondent can answer but not the NSA, or a
request for a password previously exchanged with your correspondent, or no
question at all.
That's somewhat useful since you do get some authentication. I hope the
secret question and password filters actually encrypt the link or the people
hosting the database can see right through it, just like with CAPTCHAs or no
question at all.
> In practice a simple captcha will allow you to be safe from mass
surveillance, since only targeted surveillance can be done by human beings.
We already have problems making CAPTCHAs to filter spam bots, but obviously
the most equipped surveillance agencies on Earth won't be able to get by
them.
> On top of that any spy will be detected, and have his IP address revealed.
Detected how? As for the IP address, wouldn't governments just use Tor to
mask themselves? If that's so above them, why don't we just block government
agency IP blocks? Mass surveillance solved!
> On our test, no PLM has ever been spyed even with no question at all.
Those sound like famous last words.
