commit f344776616227f3e490969038aaaad991a464e1d
Author: Mike Perry <[email protected]>
Date: Wed May 6 15:32:08 2015 -0700
One more TBB design doc update..
---
projects/torbrowser/design/index.html.en | 50 ++++++++++++++++++------------
1 file changed, 30 insertions(+), 20 deletions(-)
diff --git a/projects/torbrowser/design/index.html.en
b/projects/torbrowser/design/index.html.en
index c017f4e..9ba583d 100644
--- a/projects/torbrowser/design/index.html.en
+++ b/projects/torbrowser/design/index.html.en
@@ -1,5 +1,5 @@
<?xml version="1.0" encoding="UTF-8"?>
-<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd";><html
xmlns="http://www.w3.org/1999/xhtml";><head><meta http-equiv="Content-Type"
content="text/html; charset=UTF-8" /><title>The Design and Implementation of
the Tor Browser [DRAFT]</title><meta name="generator" content="DocBook XSL
Stylesheets V1.78.1" /></head><body><div class="article"><div
class="titlepage"><div><div><h2 class="title"><a id="design"></a>The Design and
Implementation of the Tor Browser [DRAFT]</h2></div><div><div
class="author"><h3 class="author"><span class="firstname">Mike</span> <span
class="surname">Perry</span></h3><div class="affiliation"><div
class="address"><p><code class="email"><<a class="email"
href="mailto:mikeperry#torproject org">mikeperry#torprojectÂ
org</a>></code></p></div></div></div></div><div><div class="author"><h3
class="author"><span class="firstname">Erinn</span> <span
class="surname">Clark</span></h3><div class="a
ffiliation"><div class="address"><p><code class="email"><<a class="email"
href="mailto:erinn#torproject org">erinn#torprojectÂ
org</a>></code></p></div></div></div></div><div><div class="author"><h3
class="author"><span class="firstname">Steven</span> <span
class="surname">Murdoch</span></h3><div class="affiliation"><div
class="address"><p><code class="email"><<a class="email"
href="mailto:sjmurdoch#torproject org">sjmurdoch#torprojectÂ
org</a>></code></p></div></div></div></div><div><p class="pubdate">May 6th,
2015</p></div></div><hr /></div><div class="toc"><p><strong>Table of
Contents</strong></p><dl class="toc"><dt><span class="sect1"><a
href="#idp69131840">1. Introduction</a></span></dt><dd><dl><dt><span
class="sect2"><a href="#components">1.1. Browser Component
Overview</a></span></dt></dl></dd><dt><span class="sect1"><a
href="#DesignRequirements">2. Design Requirements and
Philosophy</a></span></dt><dd><dl><dt><span class="sect2"><a
href="#security">2.1. Security
Requirements</a></span></dt><dt><span class="sect2"><a href="#privacy">2.2.
Privacy Requirements</a></span></dt><dt><span class="sect2"><a
href="#philosophy">2.3. Philosophy</a></span></dt></dl></dd><dt><span
class="sect1"><a href="#adversary">3. Adversary
Model</a></span></dt><dd><dl><dt><span class="sect2"><a
href="#adversary-goals">3.1. Adversary Goals</a></span></dt><dt><span
class="sect2"><a href="#adversary-positioning">3.2. Adversary Capabilities -
Positioning</a></span></dt><dt><span class="sect2"><a href="#attacks">3.3.
Adversary Capabilities - Attacks</a></span></dt></dl></dd><dt><span
class="sect1"><a href="#Implementation">4.
Implementation</a></span></dt><dd><dl><dt><span class="sect2"><a
href="#proxy-obedience">4.1. Proxy Obedience</a></span></dt><dt><span
class="sect2"><a href="#state-separation">4.2. State
Separation</a></span></dt><dt><span class="sect2"><a
href="#disk-avoidance">4.3. Disk Avoidance</a></span></dt><dt><span
class="sect2"><a href="#app-data-isolation
">4.4. Application Data Isolation</a></span></dt><dt><span class="sect2"><a
href="#identifier-linkability">4.5. Cross-Origin Identifier
Unlinkability</a></span></dt><dt><span class="sect2"><a
href="#fingerprinting-linkability">4.6. Cross-Origin Fingerprinting
Unlinkability</a></span></dt><dt><span class="sect2"><a
href="#new-identity">4.7. Long-Term Unlinkability via "New Identity"
button</a></span></dt><dt><span class="sect2"><a href="#other-security">4.8.
Other Security Measures</a></span></dt></dl></dd><dt><span class="sect1"><a
href="#BuildSecurity">5. Build Security and Package
Integrity</a></span></dt><dd><dl><dt><span class="sect2"><a
href="#idp70162016">5.1. Achieving Binary
Reproducibility</a></span></dt><dt><span class="sect2"><a
href="#idp70184144">5.2. Package Signatures and
Verification</a></span></dt><dt><span class="sect2"><a href="#idp70188672">5.3.
Anonymous Verification</a></span></dt><dt><span class="sect2"><a
href="#update-safety">5.4. Update Safety</a></span></d
t></dl></dd><dt><span class="appendix"><a href="#Transparency">A. Towards
Transparency in Navigation Tracking</a></span></dt><dd><dl><dt><span
class="sect1"><a href="#deprecate">A.1. Deprecation
Wishlist</a></span></dt><dt><span class="sect1"><a href="#idp70225312">A.2.
Promising Standards</a></span></dt></dl></dd></dl></div><div class="sect1"><div
class="titlepage"><div><div><h2 class="title" style="clear: both"><a
id="idp69131840"></a>1. Introduction</h2></div></div></div><p>
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd";><html
xmlns="http://www.w3.org/1999/xhtml";><head><meta http-equiv="Content-Type"
content="text/html; charset=UTF-8" /><title>The Design and Implementation of
the Tor Browser [DRAFT]</title><meta name="generator" content="DocBook XSL
Stylesheets V1.78.1" /></head><body><div class="article"><div
class="titlepage"><div><div><h2 class="title"><a id="design"></a>The Design and
Implementation of the Tor Browser [DRAFT]</h2></div><div><div
class="author"><h3 class="author"><span class="firstname">Mike</span> <span
class="surname">Perry</span></h3><div class="affiliation"><div
class="address"><p><code class="email"><<a class="email"
href="mailto:mikeperry#torproject org">mikeperry#torprojectÂ
org</a>></code></p></div></div></div></div><div><div class="author"><h3
class="author"><span class="firstname">Erinn</span> <span
class="surname">Clark</span></h3><div class="a
ffiliation"><div class="address"><p><code class="email"><<a class="email"
href="mailto:erinn#torproject org">erinn#torprojectÂ
org</a>></code></p></div></div></div></div><div><div class="author"><h3
class="author"><span class="firstname">Steven</span> <span
class="surname">Murdoch</span></h3><div class="affiliation"><div
class="address"><p><code class="email"><<a class="email"
href="mailto:sjmurdoch#torproject org">sjmurdoch#torprojectÂ
org</a>></code></p></div></div></div></div><div><p class="pubdate">May 6th,
2015</p></div></div><hr /></div><div class="toc"><p><strong>Table of
Contents</strong></p><dl class="toc"><dt><span class="sect1"><a
href="#idp53435264">1. Introduction</a></span></dt><dd><dl><dt><span
class="sect2"><a href="#components">1.1. Browser Component
Overview</a></span></dt></dl></dd><dt><span class="sect1"><a
href="#DesignRequirements">2. Design Requirements and
Philosophy</a></span></dt><dd><dl><dt><span class="sect2"><a
href="#security">2.1. Security
Requirements</a></span></dt><dt><span class="sect2"><a href="#privacy">2.2.
Privacy Requirements</a></span></dt><dt><span class="sect2"><a
href="#philosophy">2.3. Philosophy</a></span></dt></dl></dd><dt><span
class="sect1"><a href="#adversary">3. Adversary
Model</a></span></dt><dd><dl><dt><span class="sect2"><a
href="#adversary-goals">3.1. Adversary Goals</a></span></dt><dt><span
class="sect2"><a href="#adversary-positioning">3.2. Adversary Capabilities -
Positioning</a></span></dt><dt><span class="sect2"><a href="#attacks">3.3.
Adversary Capabilities - Attacks</a></span></dt></dl></dd><dt><span
class="sect1"><a href="#Implementation">4.
Implementation</a></span></dt><dd><dl><dt><span class="sect2"><a
href="#proxy-obedience">4.1. Proxy Obedience</a></span></dt><dt><span
class="sect2"><a href="#state-separation">4.2. State
Separation</a></span></dt><dt><span class="sect2"><a
href="#disk-avoidance">4.3. Disk Avoidance</a></span></dt><dt><span
class="sect2"><a href="#app-data-isolation
">4.4. Application Data Isolation</a></span></dt><dt><span class="sect2"><a
href="#identifier-linkability">4.5. Cross-Origin Identifier
Unlinkability</a></span></dt><dt><span class="sect2"><a
href="#fingerprinting-linkability">4.6. Cross-Origin Fingerprinting
Unlinkability</a></span></dt><dt><span class="sect2"><a
href="#new-identity">4.7. Long-Term Unlinkability via "New Identity"
button</a></span></dt><dt><span class="sect2"><a href="#other-security">4.8.
Other Security Measures</a></span></dt></dl></dd><dt><span class="sect1"><a
href="#BuildSecurity">5. Build Security and Package
Integrity</a></span></dt><dd><dl><dt><span class="sect2"><a
href="#idp55327360">5.1. Achieving Binary
Reproducibility</a></span></dt><dt><span class="sect2"><a
href="#idp55349120">5.2. Package Signatures and
Verification</a></span></dt><dt><span class="sect2"><a href="#idp55353648">5.3.
Anonymous Verification</a></span></dt><dt><span class="sect2"><a
href="#update-safety">5.4. Update Safety</a></span></d
t></dl></dd><dt><span class="appendix"><a href="#Transparency">A. Towards
Transparency in Navigation Tracking</a></span></dt><dd><dl><dt><span
class="sect1"><a href="#deprecate">A.1. Deprecation
Wishlist</a></span></dt><dt><span class="sect1"><a href="#idp55389664">A.2.
Promising Standards</a></span></dt></dl></dd></dl></div><div class="sect1"><div
class="titlepage"><div><div><h2 class="title" style="clear: both"><a
id="idp53435264"></a>1. Introduction</h2></div></div></div><p>
This document describes the <a class="link" href="#adversary" title="3.Â
Adversary Model">adversary model</a>,
<a class="link" href="#DesignRequirements" title="2. Design Requirements and
Philosophy">design requirements</a>, and <a class="link" href="#Implementation"
title="4. Implementation">implementation</a> of the Tor Browser. It is
current as of Tor Browser
@@ -655,13 +655,13 @@ system-wide extensions (through the use of
disabled, which prevents Flash cookies from leaking from a pre-existing Flash
directory.
- </p></div><div class="sect2"><div class="titlepage"><div><div><h3
class="title"><a id="disk-avoidance"></a>4.3. Disk
Avoidance</h3></div></div></div><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a id="idp66184288"></a>Design
Goal:</h4></div></div></div><div class="blockquote"><blockquote
class="blockquote">
+ </p></div><div class="sect2"><div class="titlepage"><div><div><h3
class="title"><a id="disk-avoidance"></a>4.3. Disk
Avoidance</h3></div></div></div><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a id="idp55029872"></a>Design
Goal:</h4></div></div></div><div class="blockquote"><blockquote
class="blockquote">
The User Agent MUST (at user option) prevent all disk records of browser
activity.
The user should be able to optionally enable URL history and other history
features if they so desire.
- </blockquote></div></div><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a
id="idp66185680"></a>Implementation Status:</h4></div></div></div><div
class="blockquote"><blockquote class="blockquote">
+ </blockquote></div></div><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a
id="idp55031232"></a>Implementation Status:</h4></div></div></div><div
class="blockquote"><blockquote class="blockquote">
We achieve this goal through several mechanisms. First, we set the Firefox
Private Browsing preference
@@ -733,7 +733,7 @@ the URL bar origin for which browser state exists, possibly
with a
context-menu option to drill down into specific types of state or permissions.
An example of this simplification can be seen in Figure 1.
- </p><div class="figure"><a id="idp66208640"></a><p
class="title"><strong>Figure 1. Improving the Privacy UI</strong></p><div
class="figure-contents"><div class="mediaobject" align="center"><img
src="NewCookieManager.png" align="middle" alt="Improving the Privacy UI"
/></div><div class="caption"><p></p>
+ </p><div class="figure"><a id="idp55052928"></a><p
class="title"><strong>Figure 1. Improving the Privacy UI</strong></p><div
class="figure-contents"><div class="mediaobject" align="center"><img
src="NewCookieManager.png" align="middle" alt="Improving the Privacy UI"
/></div><div class="caption"><p></p>
This example UI is a mock-up of how isolating identifiers to the URL bar
origin can simplify the privacy UI for all data - not just cookies. Once
@@ -741,7 +741,7 @@ browser identifiers and site permissions operate on a URL
bar basis, the same
privacy window can represent browsing history, DOM Storage, HTTP Auth, search
form history, login values, and so on within a context menu for each site.
-</div></div></div><br class="figure-break" /><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a
id="idp69892352"></a>Identifier Unlinkability Defenses in the Tor
Browser</h4></div></div></div><p>
+</div></div></div><br class="figure-break" /><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a
id="idp55056352"></a>Identifier Unlinkability Defenses in the Tor
Browser</h4></div></div></div><p>
Unfortunately, many aspects of browser state can serve as identifier storage,
and no other browser vendor or standards body has invested the effort to
@@ -1124,7 +1124,7 @@ narrow domain or use case, or when there are alternate
ways of accomplishing
the same task, these features and/or certain aspects of their functionality
may be simply removed.
- </p></li></ol></div></div><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a
id="idp69985904"></a>Strategies for Defense: Randomization versus
Uniformity</h4></div></div></div><p>
+ </p></li></ol></div></div><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a
id="idp55149888"></a>Strategies for Defense: Randomization versus
Uniformity</h4></div></div></div><p>
When applying a form of defense to a specific fingerprinting vector or source,
there are two general strategies available: either the implementation for all
@@ -1298,7 +1298,9 @@ these requests are still sent by Firefox to our SOCKS
proxy (ie we set
<span class="command"><strong>network.proxy.no_proxies_on</strong></span> to
the empty string). The local
Tor client then rejects them, since it is configured to proxy for internal IP
addresses by default. Access to the local network is forbidden via the same
-mechanism.
+mechanism. We also disable the WebRTC API as mentioned previously, since even
+if it were usable over Tor, it still currently provides the local IP address
+and associated network information to websites.
</p></li><li class="listitem"><span class="command"><strong>Invasive
Authentication Mechanisms (NTLM and SPNEGO)</strong></span><p>
@@ -1311,15 +1313,23 @@ them to reveal machine information and still fail
silently prior to the
password prompt, these authentication mechanisms should either be disabled, or
placed behind a site permission before their use. We simply disable them.
- </p></li><li class="listitem"><span class="command"><strong>USB Device ID
Enumeration</strong></span><p>
+ </p></li><li class="listitem"><span class="command"><strong>USB Device ID
Enumeration via the GamePad API</strong></span><p>
The <a class="ulink"
href="https://developer.mozilla.org/en-US/docs/Web/Guide/API/Gamepad";
target="_top">GamePad
API</a> provides web pages with the <a class="ulink"
href="https://dvcs.w3.org/hg/gamepad/raw-file/default/gamepad.html#widl-Gamepad-id";
target="_top">USB
device id, product id, and driver name</a> of all connected game
-controllers, as well as detailed information about their capabilities. This API
-should be behind a site permission in Private Browsing Modes, or should
present a generic
-controller type (perhaps a two button controller that can be mapped to the
keyboard) in all cases.
-We simply disable it via the pref <span
class="command"><strong>dom.gamepad.enabled</strong></span>.
+controllers, as well as detailed information about their capabilities.
+ </p><p>
+
+It's our opinion that this API needs to be completely redesigned to provide an
+abstract notion of a game controller rather than offloading all of the
+complexity associated with handling specific game controller models to web
+content authors. For systems without a game controller, a standard controller
+can be virtualized through the keyboard, which will serve to both improve
+usability by normalizing user interaction with different games, as well as
+eliminate fingerprinting vectors. Barring that, this API should be behind a
+site permission in Private Browsing Modes. For now though, we simply disable
+it via the pref <span
class="command"><strong>dom.gamepad.enabled</strong></span>.
</p></li><li class="listitem"><span
class="command"><strong>Fonts</strong></span><p>
@@ -1599,11 +1609,11 @@ In order to avoid long-term linkability, we provide a
"New Identity" context
menu option in Torbutton. This context menu option is active if Torbutton can
read the environment variables $TOR_CONTROL_PASSWD and $TOR_CONTROL_PORT.
- </p><div class="sect3"><div class="titlepage"><div><div><h4
class="title"><a id="idp70103376"></a>Design Goal:</h4></div></div></div><div
class="blockquote"><blockquote class="blockquote">
+ </p><div class="sect3"><div class="titlepage"><div><div><h4
class="title"><a id="idp55268352"></a>Design Goal:</h4></div></div></div><div
class="blockquote"><blockquote class="blockquote">
All linkable identifiers and browser state MUST be cleared by this feature.
- </blockquote></div></div><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a
id="idp70104624"></a>Implementation Status:</h4></div></div></div><div
class="blockquote"><blockquote class="blockquote"><p>
+ </blockquote></div></div><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a
id="idp55269600"></a>Implementation Status:</h4></div></div></div><div
class="blockquote"><blockquote class="blockquote"><p>
First, Torbutton disables Javascript in all open tabs and windows by using
both the <a class="ulink"
href="https://developer.mozilla.org/en-US/docs/XPCOM_Interface_Reference/nsIDocShell#Attributes";
target="_top">browser.docShell.allowJavascript</a>
@@ -1702,7 +1712,7 @@ images (<span
class="command"><strong>svg.in-content.enabled</strong></span>).
Fingerprinting</a> is a statistical attack to attempt to recognize specific
encrypted website activity.
- </p><div class="sect3"><div class="titlepage"><div><div><h4
class="title"><a id="idp70138960"></a>Design Goal:</h4></div></div></div><div
class="blockquote"><blockquote class="blockquote"><p>
+ </p><div class="sect3"><div class="titlepage"><div><div><h4
class="title"><a id="idp55303936"></a>Design Goal:</h4></div></div></div><div
class="blockquote"><blockquote class="blockquote"><p>
We want to deploy a mechanism that reduces the accuracy of <a class="ulink"
href="https://en.wikipedia.org/wiki/Feature_selection"; target="_top">useful
features</a> available
for classification. This mechanism would either impact the true and false
@@ -1724,7 +1734,7 @@ Congestion-Sensitive BUFLO</a>. It may be also possible
to <a class="ulink" href
defenses</a> such that they only use existing spare Guard bandwidth capacity
in the Tor
network, making them also effectively no-overhead.
- </p></blockquote></div></div><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a
id="idp70145856"></a>Implementation Status:</h4></div></div></div><div
class="blockquote"><blockquote class="blockquote"><p>
+ </p></blockquote></div></div><div class="sect3"><div
class="titlepage"><div><div><h4 class="title"><a
id="idp55310832"></a>Implementation Status:</h4></div></div></div><div
class="blockquote"><blockquote class="blockquote"><p>
Currently, we patch Firefox to <a class="ulink"
href="https://gitweb.torproject.org/tor-browser.git/commit/?h=tor-browser-31.6.0esr-4.5-1&id=20a59cec9886cf2575b1fd8e92b43e31ba053fbd";
target="_top">randomize
pipeline order and depth</a>. Unfortunately, pipelining is very fragile.
Many sites do not support it, and even sites that advertise support for
@@ -1789,7 +1799,7 @@ contend with. For this reason, we have deployed a build
system
that allows anyone to use our source code to reproduce byte-for-byte identical
binary packages to the ones that we distribute.
- </p><div class="sect2"><div class="titlepage"><div><div><h3 class="title"><a
id="idp70162016"></a>5.1. Achieving Binary
Reproducibility</h3></div></div></div><p>
+ </p><div class="sect2"><div class="titlepage"><div><div><h3 class="title"><a
id="idp55327360"></a>5.1. Achieving Binary
Reproducibility</h3></div></div></div><p>
The GNU toolchain has been working on providing reproducible builds for some
time, however a large software project such as Firefox typically ends up
@@ -1900,7 +1910,7 @@ but differs under LXC. We are also investigating currently
<a class="ulink" href="https://trac.torproject.org/projects/tor/ticket/12240";
target="_top">oddities related to
time-based dependency tracking</a> that only appear in LXC containers.
- </p></li></ol></div></div><div class="sect2"><div
class="titlepage"><div><div><h3 class="title"><a id="idp70184144"></a>5.2.Â
Package Signatures and Verification</h3></div></div></div><p>
+ </p></li></ol></div></div><div class="sect2"><div
class="titlepage"><div><div><h3 class="title"><a id="idp55349120"></a>5.2.Â
Package Signatures and Verification</h3></div></div></div><p>
The build process generates a single sha256sums.txt file that contains a sorted
list of the SHA-256 hashes of every package produced for that build version.
Each
@@ -1933,7 +1943,7 @@ In order to verify package integrity, the signature must
be stripped off using
the osslsigncode tool, as described on the <a class="ulink"
href="https://www.torproject.org/docs/verifying-signatures.html.en#BuildVerification";
target="_top">Signature
Verification</a> page.
- </p></div><div class="sect2"><div class="titlepage"><div><div><h3
class="title"><a id="idp70188672"></a>5.3. Anonymous
Verification</h3></div></div></div><p>
+ </p></div><div class="sect2"><div class="titlepage"><div><div><h3
class="title"><a id="idp55353648"></a>5.3. Anonymous
Verification</h3></div></div></div><p>
Due to the fact that bit-identical packages can be produced by anyone, the
security of this build system extends beyond the security of the official
@@ -2062,7 +2072,7 @@ possible for us to <a class="ulink"
href="https://trac.torproject.org/projects/t
ourselves</a>, as they are comparatively rare and can be handled with site
permissions.
- </p></li></ol></div></div><div class="sect1"><div
class="titlepage"><div><div><h2 class="title" style="clear: both"><a
id="idp70225312"></a>A.2. Promising Standards</h2></div></div></div><div
class="orderedlist"><ol class="orderedlist" type="1"><li class="listitem"><a
class="ulink" href="http://web-send.org"; target="_top">Web-Send
Introducer</a><p>
+ </p></li></ol></div></div><div class="sect1"><div
class="titlepage"><div><div><h2 class="title" style="clear: both"><a
id="idp55389664"></a>A.2. Promising Standards</h2></div></div></div><div
class="orderedlist"><ol class="orderedlist" type="1"><li class="listitem"><a
class="ulink" href="http://web-send.org"; target="_top">Web-Send
Introducer</a><p>
Web-Send is a browser-based link sharing and federated login widget that is
designed to operate without relying on third-party tracking or abusing other
_______________________________________________
tor-commits mailing list
[email protected]
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-commits
