[Tails-dev] Debian popularity contest
Hi, did you consider participating with Tails in Debian popularity contest? (popcorn) I saw you disabled it, but couldn't find an explanation in the Design. Cheers, adrelanos ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev
Re: [Tails-dev] [PATCH] Remove the last absolute path in our SYSLINUX config
hi, intrigeri wrote (25 Oct 2012 10:50:06 GMT) : > Great! So, I think next steps are: > 0. someone else tests the patch a bit and ACKs it: I'll do it Done => ACK > 1. a ticket is created to remind us to upstream this later > 2. the release manager decides if he wants to merge it ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev
[Tails-dev] Tails 0.14 rc1 virtualization testing & howto install virtualbox and vmplayer
Tails 0.14 rc1 686-pae sees all my cpu cores and RAM Time to test virtualization. virtualbox and vmplayer will use the 3.2.0-4-686-pae kernel headers and compile and insert some kernel modules needed to run virtual machines, create virtual network cards etc Kernel headers 3.2.0-4-686-pae, vmplayer & virtualbox need gcc 4.6 There is no squeeze backport for gcc-4.6 A Solution: dpkg --install gcc-4.4_4.4.5-8_i386.deb ln -s /usr/bin/gcc-4.4 /usr/bin/gcc-4.6 vmplayer will now install, compile & insert kernel modules virtualbox 4.2 will now install, compile & insert kernel modules Side effects of gcc-4.4 = gcc-4.6, none observed TODO: 1. View release notes & changelogs between gcc 4.4 and 4.6 Everything is running well. vmplayer in particular is *very* fast when I copy the tails-0.14-rc1 iso to the ramdisk and boot it in a VM, GIMP and all other apps load very fast apt-get, it appears secure, using the debian public key(s) stored on the tails livecd to verify the Releases file (which has the hash of the packages file), then the hash of the Packages file(which has the hash of the individual .debs), then the hash of the .deb so it should be ok to install using apt-get over tor, I havent audited it yet though, there must be bugs Once you apt-get install gcc-4.4 and symlink it to gcc-4.6 you can apt-get install virtualbox4.2 and it will install fine. https://www.virtualbox.org/wiki/Linux_Downloads is verified by verisign, so you only get verisign/ssl-level security The webpage text shows 7B0F AB3A 13B9 0743 5925 D9C9 5442 2A4B 98AB 5139 Oracle Corporation (VirtualBox archive signing key) as the key fingerprint for oracle_vbox.asc which you will need to add to your apt-key repository, and edit /etc/apt/sources.list and add deb http://download.virtualbox.org/virtualbox/debian squeeze contrib non-free to the list. Full instructions are at https://www.virtualbox.org If anyone wants to run virtualbox or vmplayer from within their tails livecd you can do it TODO: 1. Calculate what size requirements there would be if virtualbox was ever shipped with tails 2. See how a git patch could be made that is easy simple and just makes everything work well Running virtual Tails from within a Tails live-cd: Advantages: * Can hide hardware serial numbers, even if an attacker gets root * Allows stronger enforcement of tor-only connections, an attacker must break out of a virtual machine, in addition to previous steps taken. A VM can be configured to only be able to send traffic through the tor process running on the host machine. * Enables the features described at https://tails.boum.org/todo/Two-layered_virtualized_system/ https://tails.boum.org/todo/Two-layered_virtualized_system/virtails.png is a great diagram, but could be made even more secure by using multiple computers to separate things even further, whonix has a multi-computer design Example setup, each item in brackets[] is a separate computer, connected to the other computers via a crossover cable: [Tails Storage Server] Runs encrypted storage from within a vm | Gives access to encrypted storage vis sshfs | Encryption keys are never in RAM of vulnerable apache server | running many end-user services | [Tails Server Edition] Runs Apache inside a virtual tails | | | [Tails Gateway] Runs only Tor process inside a virtual tails | | | [OpenBSD livecd] Transparent Bridge Firewall - Runs only pf and allows only connections to a list of Tor bridges | | Internet ^Bridges may not be recommended for tor hidden services? but this will also work for a end-user client setup ^At any ethernet crossover cable, an intrusion detection system tap can be added A hacker could root a number of machines, and would have a very difficult time ever revealing the real IP of the machine. Please critique! I'm not sure what the point of this email is, other than to get more people interested in testing out virtualization! and making it easier for anyone who saw some error messages using apt-get I was told to discuss here ideas about virtualization and tails as it is still early in the discussion process What does everyone think about virtualization and tails? References: https://tails.boum.org/todo/Two-layered_virtualized_system/ https://tails.boum.org/todo/amd64_kernel/ ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev
Re: [Tails-dev] Debian popularity contest
Hi, adrelanos wrote (26 Oct 2012 10:42:48 GMT) : > did you consider participating with Tails in Debian popularity > contest? (popcorn) I don't remember doing more than disabling it, as a way to avoid having to think about it. > I saw you disabled it, but couldn't find an explanation in > the Design. First, Tails has no outgoing SMTP client configured by default, so popcon would not work out of the box. Second, even if Tails had the needed facility, I'm unsure Debian mail servers would accept email coming from Tor exit nodes. I suppose we could setup a dedicated limited SMTP relay as we have for WhisperBack, but that is quite some effort to setup and maintain. Personally, I'm not interested in doing this work. Third, I like Tails not to call home by default, and only then, we can make exceptions when we feel it safe and needed (e.g. the security issue check). Cheers! ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev
Re: [Tails-dev] Debian popularity contest
intrigeri: > Hi, > > adrelanos wrote (26 Oct 2012 10:42:48 GMT) : >> did you consider participating with Tails in Debian popularity >> contest? (popcorn) > > I don't remember doing more than disabling it, > as a way to avoid having to think about it. > >> I saw you disabled it, but couldn't find an explanation in >> the Design. > > First, Tails has no outgoing SMTP client configured by default, > so popcon would not work out of the box. It tries http first. > Second, even if Tails had the needed facility, I'm unsure Debian mail > servers would accept email coming from Tor exit nodes. Valid point. > I suppose we > could setup a dedicated limited SMTP relay as we have for WhisperBack, > but that is quite some effort to setup and maintain. Personally, I'm > not interested in doing this work. > > Third, I like Tails not to call home by default, > and only then, we can make exceptions when we feel it safe and needed > (e.g. the security issue check). I considered it for Whonix today and it's a real bad idea to add it. Reasons are listed here: https://sourceforge.net/p/whonix/wiki/Security/#popularity-contest If you are interested I could adjust it for Tails and add it to the Tails design. Cheers, adrelanos ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev
Re: [Tails-dev] [PATCH] Remove the last absolute path in our SYSLINUX config
26/10/12 13:03, intrigeri wrote: > hi, > > intrigeri wrote (25 Oct 2012 10:50:06 GMT) : >> Great! So, I think next steps are: > >> 0. someone else tests the patch a bit and ACKs it: I'll do it > > Done => ACK > >> 1. a ticket is created to remind us to upstream this later >> 2. the release manager decides if he wants to merge it I merged this into testing (and devel) in hope that it will fix our issues with the Universal USB Installer in rc2, or at least make fixing them easier. Cheers! ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev
[Tails-dev] Tails Attack Surface Reduction - Bridge Enforcement
Hello Work is being done on adding bridge support to Tails In addition to supporting bridges, here is another idea: [Standard Tails Livecd] 1. | | | [Bridging Firewall - Allow only bridgeip:bridgeport] 2. 1. The standard tails livecd is just the normal tails os, the user enters their bridges into vidalia, or however it ends up being supported in tails 2. This is simply a bridge firewall, for example a Tails livecd where iptables is configured to be a transparent bridge firewall. The user also enters their bridge IPs in here The bridging firewall is simply iptables in bridge mode (no IP addresses used) and is told to ONLY allow traffic going to and from the user supplied bridge IP addresses The attack surface for revealing a users IP is now reduced to being able to exploit a vulnerability in iptables, these are *extremely* rare compared to vulnerabilities in the end-user applications used, local kernel exploits etc Worth the effort making a bridge firewall CD ? ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev
[Tails-dev] Please review and merge feature/korean_input
Hi, please review and merge (into devel): branch: feature/korean_input ticket: todo/korean_input_system "Tested", as in if I choose Korean language in Tails greeter, then I get a SCIM applet in the panel, in which I can choose the Hangul input method. We've got someone willing to test early ISO images once they're out (I guess that would be 0.15~rc1 or something). Cheers, -- intrigeri | GnuPG key @ https://gaffer.ptitcanardnoir.org/intrigeri/intrigeri.asc | OTR fingerprint @ https://gaffer.ptitcanardnoir.org/intrigeri/otr.asc ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev
[Tails-dev] Please review and merge feature/dpkg-origin
Hi, branch: feature/dpkg-origin ticket: todo/custom_dpkg_origin candidate for post-0.14. commit 833df8b944c80b93b94623b5a5aec229e9e1e14e Author: Tails developers Date: Fri Oct 26 16:33:45 2012 +0200 Add vendor-specific dpkg origin information. This makes dpkg-vendor return correct information. See deb-origin(5) and dpkg-vendor(1) for details. Thanks to Paul Wise for suggesting this on the debian-derivatives mailing-list. Cheers, -- intrigeri | GnuPG key @ https://gaffer.ptitcanardnoir.org/intrigeri/intrigeri.asc | OTR fingerprint @ https://gaffer.ptitcanardnoir.org/intrigeri/otr.asc ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev
Re: [Tails-dev] Tails Attack Surface Reduction - Bridge Enforcement
Hi, (meta: do you read the list of shall we Cc: you?) a...@riseup.net wrote (26 Oct 2012 18:23:29 GMT) : > The attack surface for revealing a users IP is now reduced to being > able to exploit a vulnerability in iptables, these are *extremely* > rare compared to vulnerabilities in the end-user applications used, > local kernel exploits etc FWIW, I think this is related to * https://tails.boum.org/todo/Two-layered_virtualized_system/ * Whonix design This looks all interesting and valuable, but right now, we clearly don't have time to tackle it seriously. See https://tails.boum.org/contribute/roadmap/ for our priorities. Help is welcome. Cheers! ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev
Re: [Tails-dev] Tails 0.14 rc1 virtualization testing & howto install virtualbox and vmplayer
hi, a...@riseup.net wrote (26 Oct 2012 15:43:09 GMT) : > Tails 0.14 rc1 686-pae sees all my cpu cores and RAM Nice to hear. > Time to test virtualization. Ah. FYI this is tracked on https://tails.boum.org/todo/add_virtualbox_host_software/ (I'll ignore the proprietary vmware thing in what follows.) > virtualbox 4.2 will now install, compile & insert kernel modules Nice to read! > https://www.virtualbox.org/wiki/Linux_Downloads is verified by verisign, > so you only get verisign/ssl-level security A long-term solution for Tails would have to be based on Debian, rather than on Oracle's packages. Current status in Tails is a bit kludgy: we are shipping a 4.1.10-dfsg-1~bpo60+1 custom backport of the guest tools and drivers (custom because they are built against the xorg from squeeze-backports). > TODO: > 1. Calculate what size requirements there would be if virtualbox was ever > shipped with tails > 2. See how a git patch could be made that is easy simple and just makes > everything work well + check that issue, quoted directly from the aforementioned ticket: IIRC, VirtualBox host software sets iptables/netfilter up in a way that makes the guest system bypass the existing firewall / or be blocked by it, so some care should be taken on this side. > What does everyone think about virtualization and tails? Personally, I'd be very happy to see todo/add_virtualbox_host_software solved, but I lack time to do it any time soon. You are most welcome to go on working on this! :) Cheers, -- intrigeri | GnuPG key @ https://gaffer.ptitcanardnoir.org/intrigeri/intrigeri.asc | OTR fingerprint @ https://gaffer.ptitcanardnoir.org/intrigeri/otr.asc ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev
Re: [Tails-dev] Debian popularity contest
adrelanos wrote (26 Oct 2012 15:58:44 GMT) : >> First, Tails has no outgoing SMTP client configured by default, >> so popcon would not work out of the box. > It tries http first. Thanks for correcting me. > I considered it for Whonix today and it's a real bad idea to add it. > Reasons are listed here: > https://sourceforge.net/p/whonix/wiki/Security/#popularity-contest > If you are interested I could adjust it for Tails and add it to the > Tails design. I'd be very thankful if contributed such an adapted version. Sending it here first would be preferred. BTW, it's "popcon", not "popcorn" :) Cheers, -- intrigeri | GnuPG key @ https://gaffer.ptitcanardnoir.org/intrigeri/intrigeri.asc | OTR fingerprint @ https://gaffer.ptitcanardnoir.org/intrigeri/otr.asc ___ tails-dev mailing list tails-dev@boum.org https://mailman.boum.org/listinfo/tails-dev