[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: [tor-talk] [tor-dev] Porting Tor Browser to the BSDs

On 2015-04-17 1:34 pm, Apple Apple wrote:
On 14 Apr 2015 21:27, "Dave Warren" <davew@xxxxxxxxxxxx> wrote:

On 2015-04-14 06:05, Apple Apple wrote:

On Tails, you have to assume that the software you're running isn't
actively trying to thwart you, which may not be the case since browsers
often have vulnerabilities.

It's not perfect, but it would seem to dramatically raise the bar since a
browser based exploit alone is no longer sufficient to unmask a user like
with TBB, and potentially with Tails.

I think Tails and indeed, operating systems in general are being a bit
underrated here.

BTW, while I primarily use Whonix, I am a Tails user too. :)

An exploited Tor browser running under Tails can only access the internet
through Tor because of the IPtables rules. It cannot interfere with
processes running under a different Unix account because of process
separation. It can only read and write files that the Tor browser user has
access to.

Despite that being the good faith intention, I don't think this is true.

I believe the general principle is called "privilege escalation".

When malware escalates its privilege in bare metal Linux/BSD/Tails...game over.

Tails, or indeed stock Linux is not massively insecure on its own.

With the qualifier of "massively", it depends upon the sophistication of attack one is comparing this statement too. Or one's own personal needs/standards.

Personally, I'm not comfortable with bare metal Linux or BSD distros anymore.

Far too much TCB attack surface for my tastes.

On 17 Apr 2015 06:55, "Yuri" <yuri@xxxxxxxxx> wrote:

On 04/14/2015 15:38, WhonixQubes wrote:

-- Harder:  Whonix with VirtualBox, KVM, etc isolation for Tor

--- Hardest:  Whonix with Qubes isolation for Tor

I only don't understand why you are you so sure that the system with the
hypervisor involved is more secure.

I think the problem is that virtual machines are such a simple concept, you
flick a switch and magic security things happen. Few people stop and
realize that they don't know the first thing about how hypervisors actually work on a hardware and software level. In reality something as trivial and obscure as an absent iommu can completely negate any "isolation" you think
you are getting.

I would strongly advise people to educate themselves before pinning all
their privacy, anonymity and security on technologies they don't understand
or worse still, actively encouraging other people to do the same.

If you went to Google Scholar and grabbed all the papers you could find on Tor and other anonymity solutions before using them then why don't you do
the same thing with virtualization?


But, for security purposes, why not run Qubes + Whonix or Qubes + Tails, instead of just Whonix or Tails without Qubes' added system isolation?

On 14 Apr 2015 22:41, "WhonixQubes" <whonixqubes@xxxxxxxxxx> wrote:
However, with disk encryption, deleting VMs after usage, and overwriting
disk space, this same anti-forensics effect can be accomplished with Whonix.

Have you thought about what happens when the host PC runs out of memory and
begins to swap? Hell, what about closing the lid of your laptop?

The contents of virtual machines, details of Tor circuits, opened webpages
could all be permanently saved to disk in plaintext right?

This is something Tails explicitly designs against.

I understand people are passionate about Whonix and I'm not trying to bash it, I just want a balanced discussion. I don't think it benefits anyone to
convince ourselves that a particular solution is the perfect one true
answer and everything else is crap.

Right. And I personally use both, but more so Whonix.

Tails suffers from limitations due to its nature as a live distro.

Whonix suffers from limitations due to its nature as an install distro.

tor-talk mailing list - tor-talk@xxxxxxxxxxxxxxxxxxxx
To unsubscribe or change other settings go to