Commit ba8e3a85 authored by Tails developers's avatar Tails developers
Browse files

Merge remote-tracking branch 'origin/master' into testing

parents 6fa5119d 3d51ce9e
......@@ -44,6 +44,12 @@ enabled in the Debian kernel, and its enabling was refused in
[[!debbug 556365]] for good reasons (i.e. this feature is actually
pretty bad at testing memory) that have nothing to do with our
usecase. We are trying to have this changed in [[!debbug 646361]].
If this is refused, we'll have to build our own kernels; do we really
want to go this way? If we do, [grml's kernel building
infrastructure](http://git.grml.org/?p=grml-kernel.git) is great, and
workarounds the numerous bugs that make kernel-package currently
unusable ([[!debbug 629245]], [[!debbug 639251]], [[!debbug
638012]])... by not using it at all.
On the long run, if the Linux kernel does not wipe whatever memory
pages it allocates to use for its own data structures, the kexec'd
......
......@@ -10,12 +10,47 @@ Most of the times we are asked such a question, the answer is
So, before asking this question, please be kind enough to use our
website search functionality to look for XYZ :)
Why doesn't Tails ship software XYZ?
====================================
First of all, please ask yourself, seriously, why should Tails ship
software XYZ.
There are many, many possible reasons why Tails does not ship software
XYZ.
0. It may have licensing issues that prohibit us from shipping it (or
shipping modified versions of it, which is just as bad).
0. It may not be part of Debian stable, or even part of Debian, or
even packaged for Debian.
0. It may be in conflict with our [[specification|contribute/design]],
or it may satisfy use cases Tails is not supposed to support.
0. It may have privacy or anonymity issues. Was it ever audited in
this context?
Also, generally we try not to add too much software into Tails,
and are very careful before adding more stuff:
* More software implies more security issues.
* We do care about backward compatibility. Removing a software package
is problematic, even if it should be removed due to e.g. security
concerns, since users may have come to rely on it. In these cases we
really want to provide them with suitable alternatives.
* Tails ISO image size matters.
To end with, it might be that we simply have not thought of software
XYZ yet. Reading our [[design document|contribute/design]] may help
you understand which use cases Tails covers and hence which types of
software we may consider. If you feel that XYZ would fill up an empty
space, please [[suggest it to us|todo]] and give valid points for its
inclusion.
Why isn't Tails based on system XYZ?
====================================
0. [Why there are so many Debian derivatives](http://upsilon.cc/~zack/blog/posts/2011/09/why_there_are_so_many_debian_derivatives/), by Stefano Zacchiroli
0. We are deeply rooted and involved in Debian. The friendships,
relationships, bounds, technical expertise we have there have many
relationships, bonds, technical expertise we have there have many
benefits for Tails, and we are not ready to build the same in the
Ubuntu / OpenBSD / whatever world. See our statement about our
[[contribute/relationship_with_upstream]] for details.
......@@ -37,14 +72,14 @@ Generic answers
0. Ubuntu is based on Debian. Tails is based on Debian. Both have good
reasons to do so.
0. Ubuntu is a too fast-moving as a development platform for us. Yes,
0. Ubuntu is too fast-moving as a development platform for us. Yes,
there is LTS, but then a lot of what might make the balance weight
more on Ubuntu's side is not relevant anymore.
0. Ubuntu adds "features" in ways we find dangerous for privacy;
Ubuntu One might be an isolated example, but we don't want to play
this game of "let's see what kind of shinny'n'ugly new stuff we have to
this game of "let's see what kind of shiny'n'ugly new stuff we have to
disable in this brand new release".
0. Ubuntu is lead by a company that makes most of the important
0. Ubuntu is led by a company that makes most of the important
decisions and has the power to make them happen. Call it a
simplification if you want.
......@@ -54,7 +89,7 @@ Isn't Ubuntu more user-friendly?
At first glance Ubuntu may seem easier to install to hard-disk and to
administrate. However, this is irrelevant since Tails end-users
neither install Tails on hard-disk, nor administrate it. So from a
Tails end-user's perspective, Tails it's mostly the same as any other
Tails end-user's perspective, Tails is mostly the same as any other
GNOME desktop.
Doesn't Ubuntu support more recent hardware?
......@@ -79,7 +114,7 @@ Isn't Ubuntu more secure?
-------------------------
Yes and no. Sure, Ubuntu has some well-advertized security features
Debian has not (see below). On the other hand, the general quality of
Debian does not (see below). On the other hand, the general quality of
the maintenance work being done on packages matters from a security
perspective; Debian maintainers generally are experts in the fields
their packages deal with; Ubuntu maintainers are not that often;
......@@ -98,48 +133,13 @@ applications in Ubuntu.
right. Ubuntu makes a point here. That's why we have been helping
pushing this into Debian. This is now a release goal for Debian
Wheezy, and being actively worked on. On the other hand, such options
are not the security Graal; while they put the bar a bit higher for
are not the security Grail; while they put the bar a bit higher for
the attacker, there are known ways to workaround them, and exploit the
so-called protected binaries anyway. Given the tendency is that more
and more distribution vendors enable those options, it seems safe to
bet serious attackers take this into account, and design their
exploits accordingly.
Why doesn't Tails ship software XYZ?
====================================
First of all, please ask yourself, seriously, why should Tails ship
software XYZ.
There are many, many possible reasons why Tails does not ship software
XYZ.
0. It may have licensing issues that prohibit us from shipping it (or
ship modified versions of it, which is just as bad).
0. It may not be part of Debian stable, or even not part of Debian, or
even not packaged for Debian.
0. It may be in conflict with our [[specification|contribute/design]],
or it may satisfy usecases Tails is not supposed to support.
0. It may have privacy or anonymity issues. Was it ever audited in
this context?
Also, generally we try not to add to much software into Tails,
and are very careful before adding more stuff:
* More software implies more security issues.
* We do care about backward compatibility. Removing a software package
is problematic, even if it should be removed due to e.g. security
concerns, since users may have come to rely on it. In these cases we
really want to provide them with suitable alternatives.
* Tails ISO image size matters.
To end with, it might be that we simply have not thought of software
XYZ yet. Reading our [[design document|contribute/design]] may help
you understand which use cases Tails covers and hence which types of
software we may consider. If you feel that XYZ would fill up an empty
space, please [[suggest it to us|todo]] and give valid points for its
inclusion.
Why does Tails ship an obsolete Firefox / Iceweasel?
====================================================
......@@ -161,5 +161,5 @@ the Squeeze life-cycle. We are therefore confident this version is
safe enough for Tails users.
However, at least one feature that appeared in a newer Firefox /
Iceweasel seem interesting for Tails. Therefore, [[we've been
working|todo/Iceweasel_5.x]] to ship a more recent one.
Iceweasel seems interesting for Tails. Therefore, [[we've been
working|todo/Iceweasel_5.x]] towards shipping a more recent one.
......@@ -58,6 +58,7 @@ Now cleanup some parts of it (semi-)automatically:
perl -pi'' -e 's/\A \[ anonym \]\n//' debian/changelog
perl -pi'' -e 's/\A \[ T\(A\)ILS developers \]\n//' debian/changelog
perl -pi'' -e 's/\A \[ Tails developers \]\n//' debian/changelog
perl -pi'' -e 's/\A \[ Tails \]\n//' debian/changelog
perl -pi'' -e 's/\A \* Added a comment\n//' debian/changelog
perl -pi'' -e 's/\A \* Added a comment:.*\n//' debian/changelog
perl -pi'' -e 's/\A \* todo\+\+\n//i' debian/changelog
......
......@@ -40,8 +40,7 @@ Check the output for:
* Entering `about:plugins` in the location bar should say no plugin is
installed in every one of the following cases:
1. Torbutton add-on enabled in *Tor enabled* status
2. Torbutton add-on enabled in *Tor Disabled* status
3. Torbutton add-on disabled
2. Torbutton add-on disabled
# Pidgin
......@@ -60,8 +59,8 @@ Check the output for:
# Tor enforcement
* firewall: is the Tor-enforcement effective?
- check output of `iptables -L -n`
- check output of `iptables -t nat -L -n`
- check output of `iptables -L -n -v`
- check output of `iptables -t nat -L -n -v`
- try connecting to the Internet after unsetting `$http_proxy` and
`$HTTP_PROXY` using a piece of software that does not obey the
GNOME proxy settings, *and* is not explicitely torified in Tails:
......@@ -70,7 +69,7 @@ Check the output for:
- check output of `ip6tables -L -n`
- at a place with working IPv6: try connecting to a known-working
IPv6-enabled server on its IPv6 address over TCP and icmp6.
* is `resolv.conf` OK both before/after DHCP has been setup? it should
* is `/etc/resolv.conf` OK both before/after DHCP has been setup? it should
*always* read `nameserver 127.0.0.1`
# Use of untrusted partitions
......@@ -83,7 +82,8 @@ Check the output for:
CD/USB stick you are testing on a (possibly virtual) machine that
has a Tails system copied on a cleartext partition not managed by
LVM. The CD/USB ramdisk must use the Tails system found on the
CD/USB, and not the one found on the hard disk.
CD/USB, and not the one found on the hard disk. (Also check that
without Tails, that other Live system boots.)
# Claws
......@@ -94,8 +94,8 @@ Check the output for:
non-anonymizing SMTP relay. Then check that email's headers once
received, especially the `Received:` and `Message-ID:` ones.
* Also check that the EHLO/HELO SMTP message is not leaking anything
with a packet sniffer: Start claws with: `torify claws-mail` (which
the panel icon does in post-0.8.1 devel branch) to
with a packet sniffer: start Claws using the panel icon (which runs
`torify claws-mail`) to
avoid using the transparent proxy (which will confuse tcpdump).
Disable SSL/TLS for SMTP in Claws (so take precautions for not
leaking your password in plaintext by either changing it temporarily
......@@ -129,10 +129,7 @@ Seahorse GUI and FireGPG:
1. `sudo rm /var/lib/live/htp*`
(FIXME-FOR-0.9: update for new tordate+htpdate setup: instead
`rm /var/run/htpdate/success /var/run/tordate/done /var/lib/tor/cached-descriptors`
)
1. `sudo rm /var/run/htpdate/success /var/run/tordate/done /var/lib/tor/cached-descriptors`
2. disconnect the network cable
3. set the time to an obviously wrong one :
......@@ -175,4 +172,4 @@ steps that are worth [[a dedicated page|test/erase_memory_on_shutdown]].
language. The virtual keyboard must work and be auto-configured to
use the same keyboard layout as the X session.
* Try to start with the `truecrypt` option on boot, see if it can be found in
the Application -> Accessories menu and that it runs correctly
the *Applications* → *Accessories* menu and that it runs correctly.
......@@ -9,7 +9,7 @@ msgstr ""
"Project-Id-Version: PACKAGE VERSION\n"
"POT-Creation-Date: 2011-10-08 07:54-0600\n"
"PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n"
"Last-Translator: FULL NAME <EMAIL@ADDRESS>\n"
"Last-Translator: Rodrigo Andrade <planetaskoria@gmail.com>\n"
"Language-Team: LANGUAGE <LL@li.org>\n"
"Language: \n"
"MIME-Version: 1.0\n"
......@@ -53,11 +53,11 @@ msgid ""
msgstr ""
#. type: Content of: <div><div><h1>
msgid "Download the ISO image"
msgid "Descargar la imagen ISO"
msgstr ""
#. type: Content of: <div><div><div><h2>
msgid "Direct download"
msgid "Descarga directa"
msgstr ""
#. type: Content of: <div><div><div><h3>
......@@ -96,7 +96,7 @@ msgid ""
msgstr ""
#. type: Content of: <div><div><div><h2>
msgid "BitTorrent download"
msgid "Descarga por BitTorrent"
msgstr ""
#. type: Content of: <div><div><div><p>
......@@ -124,7 +124,7 @@ msgid ""
msgstr ""
#. type: Content of: <div><div><h1>
msgid "Verify the ISO image"
msgid "Verificar la imagen ISO"
msgstr ""
#. type: Content of: <div><div><p>
......
......@@ -29,13 +29,13 @@ msgstr ""
#. type: Content of: <div><div><p>
msgid "It helps you to:"
msgstr ""
msgstr "Te ayuda a:"
#. type: Content of: <div><div><ul><li>
msgid ""
"<strong>use the Internet anonymously</strong> almost anywhere you go and on "
"any computer:"
msgstr ""
msgstr "<strong>usar Internet anonimamente</strong> casi a donde quiera que vayas y en cualquier computadora: "
#. type: Content of: <div><div><ul><li>
#, fuzzy
......
......@@ -24,7 +24,7 @@ msgstr ""
#. type: Plain text
#, no-wrap
msgid ""
"**WARNING**: this documentation is still a work in progress. It is thus\n"
"**WARNUNG**: this documentation is still a work in progress. It is thus\n"
"incomplete, plenty of it lies on some points, and lacks screenshots. Originally\n"
"written for Incognito, it has not been fully adapted for Tails yet. Outdated\n"
"section are marked with **FIXME**. Please do **not** take them as true.\n"
......
......@@ -39,6 +39,7 @@ and !tagged(todo/translate) and !tagged(todo/documentation)
and !tagged(todo/research) and !tagged(todo/discuss)
and !tagged(todo/code) and !tagged(todo/test) and !tagged(todo/wait)
and !tagged(todo/promote) and !tagged(todo/qa)
and !tagged(todo/website)
and !link(patch) and !link(wishlist) and !todo/*/*"
actions=no archive=yes show=0 feeds=no]]
......
......@@ -24,4 +24,4 @@ stevenwelch [ at ] hot [ dot ] rr [ dot ] com
>>> Squeeze's GNOME clock applet allows configuring several locations
>>> -> could do the trick.
[[!tag todo/wait]]
>>>> No reply, closing. [[!tag todo/done]]
......@@ -6,20 +6,50 @@ Mozilla team's APT repository.
Next things to do
=================
[[!tag todo/wait todo/code]]
Extensions
----------
Some of our extensions are not marked as compatible with FF5:
- FoxyProxy: sid's 2.22.6-1 is compatible with FF4, but not marked as
compatible with FF5; tracked by [[!debbug 634071]]
- Monkeysphere 0.6 is not marked as compatible with FF5; it mostly
works with FF5, though, as we [reported to
upstream](https://labs.riseup.net/code/issues/3314) and to Debian
([[!debbug 638585]])
([[!debbug 638585]]). Let's fix this for real. [[!tag todo/code]]
Some other extensions are in a much worse state wrt. FF4+:
The FireGPG extension does not work in FF4+; work [[is being
done|todo/symmetric_OpenPGP_vs_recent_Iceweasel]] to get rid of it
anyway. Let's [[!taglink todo/wait]] for that part to be finished.
- CS Lite 1.4: unmaintained, not compatible with FF4+, needs to be
upgraded or replaced, see [[todo/iceweasel_addon_-_CS_Lite]]
- FireGPG presumably does not work in FF4+; work [[is being
done|todo/symmetric_OpenPGP_vs_recent_Iceweasel]] to find a solution.
Configuration
-------------
### Toolbars
It would be great to:
* Hide the HTTPS Everywhere and FoxyProxy icons from the
Navigation Toolbar.
* Hide the menu bar (and therefore hide the bookmarks icon from the
Navigation Toolbar).
Such settings are normally saved in `localstore.rdf`, but putting such
a customized file in `/etc/iceweasel/pref/` or
`/etc/iceweasel/profile/` does not work, => [[!taglink todo/research]]
how to seed this aspect of the configuration for new profiles.
This is probably not blocking at all, though.
A `localstore.rdf` file that would suit us would contain:
<?xml version="1.0"?>
<RDF:RDF xmlns:NC="http://home.netscape.com/NC-rdf#"
xmlns:RDF="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
<RDF:Description RDF:about="chrome://browser/content/browser.xul#toolbar-menubar"
autohide="true" />
<RDF:Description RDF:about="chrome://browser/content/browser.xul#nav-bar"
currentset="unified-back-forward-button,torbutton-button,urlbar-container,reload-button,stop-button,urlbar-search-splitter,search-container,home-button,fullscreenflex,window-controls,torbutton-context-menu,cookiemonster-status" />
<RDF:Description RDF:about="chrome://browser/content/browser.xul">
<NC:persist RDF:resource="chrome://browser/content/browser.xul#nav-bar"/>
<NC:persist RDF:resource="chrome://browser/content/browser.xul#toolbar-menubar"/>
</RDF:Description>
</RDF:RDF>
It' s impossible to register with vBulletin while using Tor. This makes participating in certain discussion forums impossible and limits funktionality of others. This is probably because the ip adress is always changing. It would be very practical to have an automated workaround for this.
> We can't just setup workarounds for every website that needs
> JavaScript and/or cookies and/or not using Tor to work. I'm in
> favour of closing this bug.
>> No input / details provided by the bug reporter for more than one
>> year, closing => [[!taglink todo/done]].
Hello. And Bye.
......@@ -64,3 +64,18 @@ found answers
* how much size does Icedove + Enigmail + l10n packages add to the
SquashFS compared to Claws Mail? -> *9MB* (as of Tails pre-0.8 devel
branch with XZ SquashFS compression)
Notes from the latest discussion
--------------------------------
Thunderbird 5.0 improves the autoconfig: at least now it is possible to stop it,
see <http://kb.mozillazine.org/Thunderbird_5.0_-_New_Features_and_Changes#Account_setup>.
We could ask Tor to block all plain text port (23, 110, 143) but that would
prevent the use of StartTLS.
For the moment, we propose instead to try (not spending too much time on it) to
code some patches on their upstream version. One option could be to disable the
steps of the autoconfig wizard corresponding to cleartext. Let us
[[!tag todo/research]] what exactly we want to patch to make the
autoconfig sane.
Tails does not detect this wireless adapter automatically, and there is no way to make it work. Help please, no internet!
> Please retry with the latest Tails: it ships a newer kernel that has
> a better chance to support your hardware.
[[!tag todo/wait]]
......@@ -89,3 +89,5 @@ Administrator password
rights. Leave this field blank to disable administrator rights.
5. responsible dev:
6. Related TODO item: [[todo/better_root_access_control]]
[[!tag todo/code]]
[[!tag todo/discuss]]
In Tails <=0.7.2 (with Torbutton <1.4) we allowed toggling Torbutton
to enable full javascript for pages it breaks etc. We made it more
difficult to toggle by mistake by setting torbutton.locked_mode=true,
......@@ -13,3 +11,12 @@ The question is, do we even want to support toggling at all at this
point, i.e. Tails 0.8 and on? That might just confuse our users as
Tor really can't be disabled thanks to our
[[firewall|contribute/design/Tor_enforcement]].
> We decided not to support toggling anymore. Main reasons:
> * interface clarity
> * Torbutton will get rid of toggling completely.
>
> The remaining issue will be dealt with otherwise:
> [[todo/iceweasel_should_support_LAN_webservers]].
[[!tag todo/done]]
......@@ -83,6 +83,13 @@ browser.
It should be noted that such a "unsafe browser" needs non-Torified
DNS resolution.
Captive portal detection
------------------------
### Ask ioerror
Seems like he is working on captive portal detection.
Beta testers
============
......
......@@ -54,6 +54,9 @@ OTOH, there might be an obscure Debian package whose installation
would reset our `iptables` rules. A malware could install such
a package, retrieve the real user IP, and send it to an attacker.
* A targeted piece of malware could do this with out the need of installing said software
Our plans
=========
......
......@@ -10,6 +10,8 @@ images, which basically means:
# Implementation
We must decide what path we choose. [[!tag todo/discuss]]
## live-build --source
live-build's `--source` option allows to build a source image
......
......@@ -2,3 +2,16 @@
Ways to recover video memory from a shutdown system may exist.
Hence, Tails should erase that memory on shutdown.
Talking with a friend of mine who used to work on video card drivers for X.org (slightly paraphrased, didn't need to take you through the entire conversation):
> < brutal_chaos> MostAwesomeDude: is it possible to erase video ram for all types of video cards during shutdown?
> < MostAwesomeDude> You can't clear the VRAM until the computer's off, because the BIOS will keep writing to it up until the last second.
> < brutal_chaos> Is there a way to flush the vram then?
> < MostAwesomeDude> Basically, become root, find the PCI BAR where VRAM's mapped, and walk over it. There's a couple things called mmapw and mmapr that can help you; Google for them.
Another idea, maybe related to the previous (this isn't really my strong point), is mapping all ram, including video memory as virtual ram and with the use of kfree in the kernel, writing all 0's to it.
......@@ -21,13 +21,14 @@ flash carries.
# Implementation
To support the youtube usecase, we additionally need to opt-in for
the html5 experimentation, which requires installing a cookie
(researched at [[!tor_bug 3347]] for TBB).
Automatically adding something like `&webm=1` (or similar) to URLs,
e.g. using a HTTPS-Everywhere custom rule or a [greasemonkey
script](https://www.userscripts.org/scripts/review/105433), could be a
better alternative.
This is now implemented in our `feature/ff4`, that uses
a [greasemonkey
script](https://www.userscripts.org/scripts/review/105433) to opt-in
for the YouTube HTML5 trial program. For the record, this problem is
being researched at [[!tor_bug 3347]] for TBB.
[[!tag todo/research]]
Will be shipped once [[other bits|todo/Iceweasel_5.x]] are ready,
hopefully in Tails 0.10.
[[!tag pending]]
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment