<title>What's New in this Release</title>
<meta name="GENERATOR" content=
"Modular DocBook HTML Stylesheet Version 1.79">
- <link rel="HOME" title="Privoxy 3.0.22 User Manual" href="index.html">
+ <link rel="HOME" title="Privoxy 3.0.23 User Manual" href="index.html">
<link rel="PREVIOUS" title="Installation" href="installation.html">
<link rel="NEXT" title="Quickstart to Using Privoxy" href=
"quickstart.html">
<table summary="Header navigation table" width="100%" border="0"
cellpadding="0" cellspacing="0">
<tr>
- <th colspan="3" align="center">Privoxy 3.0.22 User Manual</th>
+ <th colspan="3" align="center">Privoxy 3.0.23 User Manual</th>
</tr>
<tr>
<h1 class="SECT1"><a name="WHATSNEW" id="WHATSNEW">3. What's New in this
Release</a></h1>
- <p><span class="APPLICATION">Privoxy 3.0.22</span> stable is mainly a
- bug-fix release, it also has a couple of new features, though. Note that
- the first two entries in the ChangeLog below refer to security
- issues:</p>
+ <p><span class="APPLICATION">Privoxy 3.0.23</span> stable is a bug-fix
+ release, some of the fixed bugs are security issues (CVE requests
+ pending):</p>
<ul>
<li>
<ul>
<li>
- <p>Fixed a memory leak when rejecting client connections due to
- the socket limit being reached (CID 66382). This affected Privoxy
- 3.0.21 when compiled with IPv6 support (on most platforms this is
- the default).</p>
+ <p>Fixed a DoS issue in case of client requests with incorrect
+ chunk-encoded body. When compiled with assertions enabled (the
+ default) they could previously cause Privoxy to abort(). Reported
+ by Matthew Daley.</p>
</li>
<li>
- <p>Fixed an immediate-use-after-free bug (CID 66394) and two
- additional unconfirmed use-after-free complaints made by Coverity
- scan (CID 66391, CID 66376).</p>
+ <p>Fixed multiple segmentation faults and memory leaks in the
+ pcrs code. This fix also increases the chances that an invalid
+ pcrs command is rejected as such. Previously some invalid
+ commands would be loaded without error. Note that Privoxy's pcrs
+ sources (action and filter files) are considered trustworthy
+ input and should not be writable by untrusted third-parties.</p>
</li>
<li>
- <p>Actually show the FORCE_PREFIX value on the show-status
- page.</p>
+ <p>Fixed an 'invalid read' bug which could at least theoretically
+ cause Privoxy to crash. So far, no crashes have been
+ observed.</p>
</li>
<li>
- <p>Properly deal with Keep-Alive headers with timeout= parameters
- If the timeout still can't be parsed, use the configured timeout
- instead of preventing the client from keeping the connection
- alive. Fixes #3615312/#870 reported by Bernard Guillot.</p>
+ <p>Compiles with --disable-force again. Reported by Kay
+ Raven.</p>
</li>
<li>
- <p>Not using any filter files no longer results in warning
- messages unless an action file is referencing header taggers or
- filters. Reported by Stefan Kurtz in #3614835.</p>
- </li>
-
- <li>
- <p>Fixed a bug that prevented Privoxy from reusing some reusable
- connections. Two bit masks with different purpose unintentionally
- shared the same bit.</p>
- </li>
-
- <li>
- <p>A couple of additional bugs were discovered by Coverity Scan.
- The fixes that are not expected to affect users are not
- explicitly mentioned here, for details please have a look at the
- CVS logs.</p>
+ <p>Client requests with body that can't be delivered no longer
+ cause pipelined requests behind them to be rejected as invalid.
+ Reported by Basil Hussain.</p>
</li>
</ul>
</li>
<ul>
<li>
- <p>Introduced negative tag patterns NO-REQUEST-TAG and
- NO-RESPONSE-TAG. They apply if no matching tag is found after
- parsing client or server headers.</p>
- </li>
-
- <li>
- <p>Add support for external filters which allow to process the
- response body with a script or program written in any language
- the platform supports. External filters are enabled with
- +external-filter{} after they have been defined in one of the
- filter files with a header line starting with "EXTERNAL-FILTER:".
- External filter support is experimental, not compiled by default
- and known not to work on all platforms.</p>
- </li>
-
- <li>
- <p>Add support for the 'PATCH' method as defined in RFC5789.</p>
- </li>
-
- <li>
- <p>Reject requests with unsupported Expect header values. Fixes a
- couple of Co-Advisor tests.</p>
+ <p>If a pcrs command is rejected as invalid, Privoxy now logs the
+ cause of the problem as text. Previously the pcrs error code was
+ logged.</p>
</li>
<li>
- <p>Normalize the HTTP-version in forwarded requests and
- responses. This is an explicit RFC 2616 MUST and RFC 7230
- mandates that intermediaries send their own HTTP-version in
- forwarded messages.</p>
- </li>
-
- <li>
- <p>Client 'Keep-Alive' headers are no longer forwarded. From a
- user's point of view it doesn't really matter, but RFC 2616
- (obsolete) mandates that the header is removed and this fixes a
- Co-Advisor complaint.</p>
- </li>
-
- <li>
- <p>Change declared template file encoding to UTF-8. The templates
- already used a subset of UTF-8 anyway and changing the
- declaration allows to properly display UTF-8 characters used in
- the action files. This change may require existing action files
- with ISO-8859-1 characters that aren't valid UTF-8 to be
- converted to UTF-8. Requested by Sam Chen in #582.</p>
- </li>
-
- <li>
- <p>Do not pass rejected keep-alive timeouts to the server. It
- might not have caused any problems (we know of), but doing the
- right thing shouldn't hurt either.</p>
- </li>
-
- <li>
- <p>Let log_error() use its own buffer size #define to make
- changing the log buffer size slightly less inconvenient.</p>
- </li>
-
- <li>
- <p>Turned single-threaded into a "proper" toggle directive with
- arguments.</p>
- </li>
-
- <li>
- <p>CGI templates no longer enforce new windows for some
- links.</p>
- </li>
-
- <li>
- <p>Remove an undocumented workaround ('HOST' header removal) for
- an Apple iTunes bug that according to #729900 got fixed in
- 2003.</p>
+ <p>The tests are less likely to cause false positives.</p>
</li>
</ul>
</li>
<ul>
<li>
- <p>The pattern 'promotions.' is no longer being blocked. Reported
- by rakista in #3608540.</p>
- </li>
-
- <li>
- <p>Disable fast-redirects for .microsofttranslator.com/.</p>
- </li>
-
- <li>
- <p>Disable filter{banners-by-size} for
- .dgb-tagungszentren.de/.</p>
- </li>
-
- <li>
- <p>Add adn.speedtest.net as a site-specific unblocker. Support
- request #3612908.</p>
- </li>
-
- <li>
- <p>Disable filter{banners-by-size} for creativecommons.org/.</p>
- </li>
-
- <li>
- <p>Block requests to data.gosquared.com/. Reported by cbug in
- #3613653.</p>
- </li>
-
- <li>
- <p>Unblock .conrad./newsletter/. Reported by David Bo in
- #3614238.</p>
- </li>
-
- <li>
- <p>Unblock .bundestag.de/.</p>
- </li>
-
- <li>
- <p>Unblock .rote-hilfe.de/.</p>
- </li>
-
- <li>
- <p>Disable fast-redirects for .facebook.com/plugins/like.php.</p>
- </li>
-
- <li>
- <p>Unblock Stackexchange popup URLs that aren't used to serve
- ads. Reported by David Wagner in #3615179.</p>
- </li>
-
- <li>
- <p>Disable fast-redirects for creativecommons.org/.</p>
- </li>
-
- <li>
- <p>Unblock .stopwatchingus.info/.</p>
- </li>
-
- <li>
- <p>Block requests for .adcash.com/script/. Reported by
- Tyrexionibus in #3615289.</p>
- </li>
-
- <li>
- <p>Disable HTML filters if the response was tagged as JavaScript.
- Filtering JavaScript code with filters intended to deal with HTML
- is usually a waste of time and, more importantly, may break
- stuff.</p>
- </li>
-
- <li>
- <p>Use a custom redirect{} for
- .washingtonpost.com/wp-apps/imrs\.php\?src= Previously enabling
- the 'Advanced' settings (or manually enabling +fast-redirects{})
- prevented some images from being loaded properly.</p>
- </li>
-
- <li>
- <p>Unblock "adina*." Fixes #919 reported by Morton A.
- Goldberg.</p>
- </li>
-
- <li>
- <p>Block '/.*DigiAd'.</p>
- </li>
-
- <li>
- <p>Unblock 'adele*.'. Reported by Adele Lime in #1663.</p>
- </li>
-
- <li>
- <p>Disable banners-by-size for kggp.de/.</p>
- </li>
- </ul>
- </li>
-
- <li>
- <p>Filter file improvements & bug fixes:</p>
-
- <ul>
- <li>
- <p>Decrease the chances that js-annoyances creates invalid
- JavaScript. Submitted by John McGowan on ijbswa-users@.</p>
+ <p>'.sify.com/' is no longer blocked. Apparently it is not
+ actually a pure tracking site (anymore?). Reported by Andrew on
+ ijbswa-users@.</p>
</li>
<li>
- <p>Let the msn filter hide 'related' ads again.</p>
- </li>
-
- <li>
- <p>Remove a stray '1' in the 'html-annoyances' filter.</p>
- </li>
-
- <li>
- <p>Prevent img-reorder from messing up img tags with empty src
- attributes. Fixes #880 reported by Duncan.</p>
+ <p>Unblock banners on .amnesty.de/ which aren't ads.</p>
</li>
</ul>
</li>
<ul>
<li>
- <p>Updated the 'Would you like to donate?' section.</p>
- </li>
-
- <li>
- <p>Note that invalid forward-override{} parameter syntax isn't
- detected until the parameter is used.</p>
- </li>
-
- <li>
- <p>Add another +redirect{} example: a shortcut for illumos
- bugs.</p>
- </li>
-
- <li>
- <p>Make it more obvious that many operating systems support log
- rotation out of the box.</p>
- </li>
-
- <li>
- <p>Fixed dead links. Reported by Mark Nelson in #3614557.</p>
- </li>
-
- <li>
- <p>Rephrased the 'Why is the configuration so complicated?'
- answer to be slightly less condescending. Anonymously suggested
- in #3615122.</p>
- </li>
-
- <li>
- <p>Be more explicit about accept-intercepted-requests's lack of
- MITM support.</p>
- </li>
-
- <li>
- <p>Make 'demoronizer' FAQ entries more generic.</p>
- </li>
-
- <li>
- <p>Add an example hostname to the --pre-chroot-nslookup
- description.</p>
- </li>
-
- <li>
- <p>Add an example for a host pattern that matches an IP
- address.</p>
- </li>
-
- <li>
- <p>Rename the 'domain pattern' to 'host pattern' as it may
- contain IP addresses as well.</p>
- </li>
-
- <li>
- <p>Recommend forward-socks5t when using Tor. It seems to work
- fine and modifying the Tor configuration to profit from it hasn't
- been necessary for a while now.</p>
- </li>
-
- <li>
- <p>Add another redirect{} example to stress that redirect loops
- can and should be avoided.</p>
- </li>
-
- <li>
- <p>The usual spelling and grammar fixes. Parts of them were
- reported by Reuben Thomas in #3615276.</p>
- </li>
-
- <li>
- <p>Mention the PCRS option letters T and D in the filter
- section.</p>
- </li>
-
- <li>
- <p>Clarify that handle-as-empty-doc-returns-ok is still useful
- and will not be removed without replacement.</p>
- </li>
-
- <li>
- <p>Note that security issues shouldn't be reported using the bug
- tracker.</p>
- </li>
-
- <li>
- <p>Clarify what Privoxy does if both +block{} and +redirect{}
- apply.</p>
- </li>
-
- <li>
- <p>Removed the obsolete bookmarklets section.</p>
- </li>
- </ul>
- </li>
-
- <li>
- <p>Build system improvements:</p>
-
- <ul>
- <li>
- <p>Let --with-group properly deal with secondary groups. Patch
- submitted by Anatoly Arzhnikov in #3615187.</p>
- </li>
-
- <li>
- <p>Fix web-actions target.</p>
+ <p>The 'Would you like to donate?' section now also contains a
+ "Paypal" address.</p>
</li>
<li>
- <p>Add a web-faq target that only updates the FAQ on the
- webserver.</p>
+ <p>The list of supported operating systems has been updated.</p>
</li>
<li>
- <p>Remove already-commented-out non-portable DOSFILTER
- alternatives.</p>
+ <p>The existence of the SF support and feature trackers has been
+ deemphasized because they have been broken for months. Most of
+ the time the mailing lists still work.</p>
</li>
<li>
- <p>Remove the obsolete targets dok-put and dok-get.</p>
+ <p>The claim that default.action updates are sometimes released
+ on their own has been removed. It hasn't happened in years.</p>
</li>
<li>
- <p>Add a sf-shell target.</p>
+ <p>Explicitly mention that Tor's port may deviate from the
+ default when using a bundle. Requested by Andrew on
+ ijbswa-users@.</p>
</li>
</ul>
</li>