X-Git-Url: http://www.privoxy.org/gitweb/?p=privoxy.git;a=blobdiff_plain;f=doc%2Fwebserver%2Fuser-manual%2Fwhatsnew.html;h=eeb32d20740bbf29ff7b6d9f01ef9a309367d47c;hp=d1bf8633b0d541b1102cfe40a4fe33923503132a;hb=3db7a58b2bbed7b6356b2a0600e93ec4f2846499;hpb=70b9ed04dec24ef32a00a14dbe2a1556275e9bd3 diff --git a/doc/webserver/user-manual/whatsnew.html b/doc/webserver/user-manual/whatsnew.html index d1bf8633..eeb32d20 100644 --- a/doc/webserver/user-manual/whatsnew.html +++ b/doc/webserver/user-manual/whatsnew.html @@ -7,7 +7,7 @@ NAME="GENERATOR" CONTENT="Modular DocBook HTML Stylesheet Version 1.79">
Added the config file option handle-as-empty-doc-returns-ok to - work around Firefox bug #492459, which causes Firefox to hang - if JavaScripts are blocked in certain situations. The option is - enabled in the default config file. +> Fixed last-chunk-detection for responses where the content was small + enough to be read with the body, causing Privoxy to wait for the + end of the content until the server closed the connection. + Reported by "Karsten" in #3028326.
Added the config file option default-server-timeout to control the - assumed default server timeout. Since Privoxy no longer returns - an error message for connection resets on reused client connections, - assuming larger server timeout values appears to actually work - pretty well as long as connections aren't shared. +> Responses with status code 204 weren't properly detected as body-less + like RFC2616 mandates. Like the previous bug, this caused Privoxy + to wait for the end of the content until the server closed the connection. + Fixes #3022042 and #3025553, reported by a user with no visible name. + Most likely also fixes a bunch of other AJAX-related problem reports + that got closed in the past due to insufficient information and lack + of feedback.
Added optional support for FreeBSD's accf_http(9). Use the - configure option --enable-accept-filter to enable it. +> Fixed an ACL bug that made it impossible to build a blacklist. + Usually the ACL directives are used in a whitelist, which worked + as expected, but blacklisting is still useful for public proxies + where one only needs to deny known abusers access.
Added fancier Privoxy icons for win32. Contributed by Jeff H. +> Added LOG_LEVEL_RECEIVED to log the not-yet-parsed data read from the + network. This should make debugging various parsing issues a lot easier.
In daemon mode, fd 0, 1 and 2 are bound to /dev/null. +> The IPv6 code is enabled by default on Windows versions that support it. + Patch submitted by oCameLo in #2942729.
Resolve localhost using whatever address family the operating - system feels like. Previous betas would try to use IPv4 as this - is what most users expect, but this didn't work reliable on - GNU/Linux systems. +> In mingw32 versions, the user.filter file is reachable through the + GUI, just like default.filter is. Feature request 3040263.
In the action lists on CGI pages, actions and their parameters are - no longer separated with a space. The action file parser doesn't - actually allow this and will throw an invalid syntax error if actions - and parameters in the action files are separated. Not adding the - spaces means copy and pasting CGI output into the action files works. +> Added the configure option --enable-large-file-support to set a few + defines that are required by platforms like GNU/Linux to support files + larger then 2GB. Mainly interesting for users without proper logfile + management.
The default keep-alive timeout has been reduced to 5 seconds to work - around hangs in clients that treat the proxy like any other host and - stop allowing any new connections if the "maximum number of - connections per host" is reached. +> Logging with "debug 16" no longer stops at the first nul byte which is + pretty useless. Non-printable characters are replaced with their hex value + so the result can't span multiple lines making parsing them harder then + necessary.
Several webbug URLs that look like they are leading to images are now - blocked as image instead of empty documents. Doing the latter causes - WebKit-based clients to show a "missing image" icon which may mess up - the layout. +> Privoxy logs when reading an action, filter or trust file.
The no-such-domain template if used if for DNS resolution - problems with FEATURE_IPV6_SUPPORT enabled. Previously the - connect-failed template was used. Reported by 'zebul666'. +> Fixed incorrect regression test markup which caused a test in + 3.0.16 to fail while Privoxy itself was working correctly. + While Privoxy accepts hide-referer, too, the action name is actually + hide-referrer which is also the name used one the final results page, + where the test expected the alias.
Accepts quoted expiration dates even though RFC 2109 10.1.2 - doesn't seem to allow them. Reported anonymously. -
CGI interface improvements: +In finish_http_response(), continue to add the 'Connection: close' + header if the client connection will not be kept alive. + Anonymously pointed out in #2987454. +
Don't try to forget connections if connection sharing is disabled. - This wasn't a real problem but caused an unnecessary log message. -
Apostrophes in block messages no longer cause parse errors + when the blocked page is viewed with JavaScript enabled. + Reported by dg1727 in #3062296. +The still undocumented --enable-extended-host-patterns configure - option has a better description. -
Fix a bunch of anchors that used underscores instead of dashes. +Fixed an error message that would claim a write to the server - failed when actually writing to the client failed. -
Allow to keep the client connection alive after crunching the previous request. + Already opened server connections can be kept alive, too. +Log the crunch reason before trying to write to the client. - The log is easier to read that way. -
In cgi_show_url_info(), don't forget to prefix URLs that only contain http:// or https:// in the path. + Fixes #2975765 reported by Adam Piggott. +Several log messages about client connections also mention - the socket number. +> Show the 404 CGI page if cgi_send_user_manual() is called while + local user manual delivery is disabled. +
handle-as-empty-document no longer depends on the image blocking - code being enabled. -
Action file improvements: +Enable user.filter by default. Suggested by David White in #3001830. +
Privoxy-Log-Parser is roughly 40% faster in highlighting mode. -
Block .sitestat.com/. Reported by johnd16 in #3002725. +uagen, a Firefox User-Agent generator for Privoxy and Mozilla - browsers has been imported and is available in the tarballs - tools directory. -
Block .atemda.com/. Reported by johnd16 in #3002723. +The scripts in the tools directory treat unknown parameters - as fatal errors. -
Block js.adlink.net/. Reported by johnd16 in #3002720. +If you missed the previous two beta versions, you may also be - interested in the additional changes since 3.0.12, the - last stable release:
Block .analytics.yahoo.com/. Reported by johnd16 in #3002713. +
Added IPv6 support. Thanks to Petr Pisar who not only provided - the initial patch but also helped a lot with the integration. -
Fix problems noticed on Yahoo mail and news pages. +Added client-side keep-alive support. -
Remove the too broad yahoo section, only keeping the + fast-redirects exception as discussed on ijbswa-devel@. +The connection sharing code is only used if the connection-sharing - option is enabled. -
Don't block adesklets.sourceforge.net. Reported in #2974204. +The latency is taken into account when evaluating whether or not to - reuse a connection. This should significantly reduce the number of - connections problems several users reported. -
Block chartbeat ping tracking. Reported in #2975895. +The max-client-connections option has been added to restrict - the number of client connections below a value enforced by - the operating system. -
Tag CSS and image requests with cautious and medium settings, too. +If the server doesn't specify how long the connection stays alive, - Privoxy errs on the safe side of caution and assumes it's only a second. -
Don't handle view.atdmt.com as image. It's used for click-throughs + so users should be able to "go there anyway". + Reported by Adam Piggott in #2975927. +Setting keep-alive-timeout to 0 disables keep-alive support. Previously - Privoxy would claim to allow persistence but not reuse the connection. -
Also let the refresh-tags filter remove invalid refresh tags where + the 'url=' part is missing. Anonymously reported in #2986382. + While at it, update the description to mention the fact that only + refresh tags with refresh times above 9 seconds are covered. +Pipelined requests are less likely to be mistaken for the request - body of the previous request. Note that Privoxy still has no real - pipeline support and will either serialize pipelined requests or - drop them in which case the client has to resent them. -
javascript needs to be blocked with +handle-as-empty-document to + work around Firefox bug 492459. So move .js blockers from + +block{Might be a web-bug.} -handle-as-empty-document + to + +block{Might be a web-bug.} +handle-as-empty-document +Fixed a crash on some Windows versions when header randomization - is enabled and the date couldn't be parsed. -
ijbswa-Feature Requests-3006719 - Block 160x578 Banners. +Privoxy's keep-alive timeout for the current connection is reduced - to the one specified in the client's Keep-Alive header. -
Block another omniture tracking domain +For HTTP/1.1 requests, Privoxy implies keep-alive support by not - setting any Connection header instead of using 'Connection: keep-alive'. -
Added a range-requests tagger. +If the socket isn't reusable, Privoxy doesn't temporarily waste - a socket slot to remember the connection. +> Added two sections to get Flickr's Ajax interface working with + default pre-settings. If you change the configuration to block + cookies by default, you'll need additional exceptions. + Reported by Mathias Homann in #3101419 and by Patrick on ijbswa-users@ +
If keep-alive support is disabled but compiled in, the client's - Keep-Alive header is removed. -
Documentation improvements: +Explicitly mention how to match all URLs. +
Fixed a bug on mingw32 where downloading large files failed if - keep-alive support was enabled. -
Consistently recommend socks5 in the Tor FAQ entry and mention + its advantage compared to socks4a. Reported by David in #2960129. +Fixed a bug that (at least theoretically) could cause log - timestamps to be occasionally off by about a second. -
Slightly improve the explanation of why filtering may appear + slower than it is. +The configure script respects the $PATH variable when searching - for groups and id. +> Grammar fixes for the ACL section. +
Compressed content with extra fields couldn't be decompressed - and would get passed to the client unfiltered. This problem - has only be detected through statical analysis with clang as - nobody seems to be using extra fields anyway. -
Privoxy-Log-Parser improvements: +Also gather statistics for blocked and redirected requests. +
If the server resets the Connection after sending only the headers - Privoxy forwards what it got to the client. Previously Privoxy - would deliver an error message instead. -
Provide the percentage of keep-alive offers the client accepted. +Error messages in case of connection timeouts use the right - HTTP status code. -
Add a --url-statistics-threshold option. +If spawning a child to handle a request fails, the client - gets an error message and Privoxy continues to listen for - new requests right away. -
Also gather statistics for ressources, methods, and HTTP versions + used by the client. +The error messages in case of server-connection timeouts or - prematurely closed server connections are now template-based. -
Add a --host-statistics-threshold option to also gather + statistics about how many request where made per host. +If zlib support isn't compiled in, Privoxy no longer tries to - filter compressed content unless explicitly asked to do so. -
Fix a bug in handle_loglevel_header() where a 'scan: ' got lost. +In case of connections that are denied based on ACL directives, - the memory used for the client IP is no longer leaked. -
Add a --shorten-thread-ids option to replace the thread id with + a decimal number. +Fixed another small memory leak if the client request times out - while waiting for client headers other than the request line. -
Accept and ignore: Looks like we got the last chunk together + with the server headers. We better stop reading. +The client socket is kept open until the server socket has - been marked as unused. This should increase the chances that - the still-open connection will be reused for the client's next - request to the same destination. Note that this only matters - if connection-sharing is enabled. -
Accept and ignore: Continue hack in da house. +A TODO list has been added to the source tarballs to give potential - volunteers a better idea of what the current goals are. Donations - are still welcome too: http://www.privoxy.org/faq/general.html#DONATE -
Accept and higlight: Rejecting connection from 10.0.0.2. + Maximum number of connections reached. +In case of missing server data, no error message is send to the - client if the request arrived on a reused connection. The client - is then supposed to silently retry the request without bothering - the user. This should significantly reduce the frequency of the - "No server or forwarder data received" error message many users - reported. -
Accept and highlight: Loading actions file: /usr/local/etc/privoxy/default.action +More reliable detection of prematurely closed client sockets - with keep-alive enabled. -
Accept and highlight: Loading filter file: /usr/local/etc/privoxy/default.filter +FEATURE_CONNECTION_KEEP_ALIVE is decoupled from - FEATURE_CONNECTION_SHARING and now available on - all platforms. -
Accept and highlight: Killed all-caps Host header line: HOST: bestproxydb.com +Improved handling of POST requests on reused connections. - Should fix problems with stalled connections after submitting - form data with some browser configurations. -
Accept and highlight: Reducing expected bytes to 0. Marking + the server socket tainted after throwing 4 bytes away. +Fixed various latency calculation issues. +> Accept: Merged multiple header lines to: 'X-FORWARDED-PROTO: http X-HOST: 127.0.0.1' +
Allows the client to pass NTLM authentication requests to a - forwarding proxy. This was already assumed and hinted to work - in 3.0.13 beta but actually didn't. Now it's confirmed to work - with IE, Firefox and Chrome. - Thanks to Francois Botha and Wan-Teh Chang -
Code cleanups: +Remove the next member from the client_state struct. Only the main + thread needs access to all client states so give it its own struct. +
Fixed a calculation problem if receiving the server headers - takes more than two reads, that could cause Privoxy to terminate - the connection prematurely. Reported by Oliver. -
Garbage-collect request_contains_null_bytes(). +Compiles again on platforms such as OpenBSD and systems - using earlier glibc version that don't support AI_ADDRCONFIG. - Anonymously submitted in #2872591. -
Ditch redundant code in unload_configfile(). +A bunch of MS VC project files and Suse and Redhat RPM spec - files have been removed as they were no longer maintained for - quite some time. -
Ditch LogGetURLUnderCursor() which doesn't seem to be used anywhere. +Overly long action lines are properly rejected with a proper - error message. Previously they would be either rejected as - invalid or cause a core dump through abort(). -
In write_socket(), remove the write-only variable write_len in + an ifdef __OS2__ block. Spotted by cppcheck. +Already timed-out connections are no longer temporarily remembered. - They weren't reused anyway, but wasted a socket slot. -
In connect_to(), don't declare the variable 'flags' on OS/2 where + it isn't used. Spotted by cppcheck. +len refers to the number of bytes actually read which might - differ from the ones received. Adjust log messages accordingly. -
Limit the scope of various variables. Spotted by cppcheck. +The optional JavaScript on the CGI page uses encodeURIComponent() - instead of escape() which doesn't encode all characters that matter. - Anonymously reported in #2832722. -
In add_to_iob(), turn an interestingly looking for loop into a + boring while loop. +Fix gcc45 warnings in decompress_iob(). -
Code cleanup in preparation for external filters. +Various log message improvements. -
In listen_loop(), mention the socket on which we accepted the + connection, not just the source IP address. +In write_socket(), also log the socket we're writing to. +
In log_error(), assert that escaped characters get logged + completely or not at all. +
Privoxy-Regression-Test supports redirect tests. +> In log_error(), assert that ival and sval have reasonable values. + There's no reason not to abort() if they don't. +
Remove an incorrect cgi_error_unknown() call in a + cannnot-happen-situation in send_crunch_response(). +
Clean up white-space in http_response definition and + move the crunch_reason to the beginning. +
Turn http_response.reason into an enum and rename it + to http_response.crunch_reason. +
Silence a 'gcc (Debian 4.3.2-1.1) 4.3.2' warning on i686 GNU/Linux. +
Privoxy-Log-Parser can gather some connection statistics. +> GNUmakefile improvements: +
Use $(SSH) instead of ssh, so one only needs to specify a username once. +
Removed references to the action feedback thingy that hasn't been + working for years. +
Consistently use shell.sourceforge.net instead of shell.sf.net so + one doesn't need to check server fingerprints twice. +
Removed GNUisms in the webserver and webactions targets so they + work with standard tar. +