[Tails-dev] Re-Introduction of I2P to Tails

Delete this message

Reply to this message
Author: Konrad Bächler
To: tails-dev
Subject: [Tails-dev] Re-Introduction of I2P to Tails
Hi all

we would like to help and work on the re-introduction of I2P into Tails.
We understood that there is a need for the re-introduction in the "Tails
way" (easy to maintain, as close as possible to Debian). Is this true -
or is I2P unwanted? :)

We have a very long term interest in I2P (disclosure: working full time
for the non-profit association diva.exchange; I am one of the founding
members and I am used to long term missions). I2P and Tor experience is
given. I2P is one of the core elements of our system. Obviously we're
also fine to contribute to Debian or other OS's. As long as it's for the
benefit of I2P. All we do is done in public. All our code must be GPLv3

We have a strong bias towards the C++ version of I2P: I2Pd. So we would
tackle the packaging from there. Later we may also look at the java version.

There are multiple issues related to the re-introduction of I2P, like:

* if the I2P router is not properly sandboxed - it might compromise the
overall security of Tails.
* bootstrapping into the network (communicating with "reseed" servers)
must be done via Tor.
* I2P is a slow starter... the router needs some time to get integrated.
And the UX should still be great...
* Routers should be stopped gracefully otherwise the I2P network will
suffer - but honestly: that's not an OS problem :). This has to be fixed
by I2P.
* ... more ...

We're aware of all that. And we are not aware of lots of other problems
to come ;).

Now, this is our plan ("design spec"):

Prepare some deb package for Tails, including:
- i2pd
- firejail to isolate i2pd
- iptable rules in the jail
- making sure tor is used for all non-i2p traffic (like bootstrapping)

We're looking very much forward to your feedback. Thanks a lot for your
time and efforts - much appreciated!