CHANGES-1.3.txt
---------------
+CHANGES IN CUPS V1.3.11
+
+ - The scheduler did not prevent nested classes (STR #3211)
+ - The scheduler did not reprint processing jobs that were moved to
+ another destination (STR #3222)
+ - The scheduler did not reset the current job file when stopping a
+ printer (STR #3226)
+ - The scheduler did not handle POSTs to custom CGIs properly (STR #3221)
+ - The pdftops filter did not print landscape PDF pages properly
+ (STR #2881)
+ - The scheduler did not handle partial header lines properly from CGI
+ programs (STR #3194)
+ - The web interface could hang on OpenBSD (STR #3176, STR #3196)
+ - The scheduler and cupsfilter utility did not handle rules starting
+ with a negation operator properly (STR #3160)
+ - The scheduler and cupsfilter utility would crash with certain MIME
+ .types rules (STR #3159)
+ - httpSetField wasn't bracketing IPv6 numeric addresses for the Host:
+ field (STR #3164)
+ - The ServerName, if specified, was not treated as a valid alias for the
+ local system (STR #3167)
+ - "make epm" did not work (STR #3166)
+ - "lpstat -h server" showed non-shared printers (STR #3147)
+ - "make check" did not work on Linux (STR #3161)
+
+
CHANGES IN CUPS V1.3.10
- - Documentation fixes (STR #2994, STR #2995, STR #3008, STR #3056)
+ - Documentation fixes (STR #2994, STR #2995, STR #3008, STR #3056,
+ STR #3057)
+ - SECURITY: The scheduler now protects against DNS rebinding attacks
+ (STR #3118)
+ - SECURITY: Fixed TIFF integer overflow in image filters (STR #3031)
+ - The scheduler did not support the job-hold-until attribute with the
+ Restart-Job operation (STR #3130)
+ - SECURITY: The PNG image reading code did not validate the
+ image size properly, leading to a potential buffer overflow
+ (STR #2974)
+ - The rastertohp driver did not set the 1-sided printing mode when
+ needed (STR #3131)
+ - Now use a wrapper program instead of our fork of the Xpdf code to
+ support printing of PDF files. The new wrapper supports using Xpdf,
+ poppler, or Ghostscript to convert PDF files to PostScript (STR #3129)
+ - Long job names caused problems with some PJL printers (STR #3125)
+ - The lpq command did not work when showing all destinations (STR #3117)
+ - The scheduler used a codeset name of UTF8 which is not supported on
+ Solaris (STR #3113)
+ - cupsGetJobs() did not work with a NULL destination (STR #3107)
+ - Fixed a localization problem for option choices (incorrectly) named
+ "Custom" (STR #3106)
+ - The fallback OpenSSL random number seeding would not work (STR #3079)
+ - The scheduler might miss a child signal, causing high CPU usage.
+ - The scheduler did not enforce quotas after the job history was
+ unloaded (STR #3078)
+ - The job-k-limit, job-page-limit, and job-quota-period attributes
+ could not be set using the lpadmin command (STR #3077)
+ - httpSeparateURI() did not error out on URIs with a missing port
+ number after a colon.
+ - Fixed a Valgrind-detected initialization error when creating a
+ missing directory on startup.
+ - The scheduler did not always read all of the HTTP headers from a
+ CGI script/program.
+ - The scheduler did not always set the "air" property in Bonjour/DNS-SD
+ registrations.
+ - The scheduler incorrectly compared Mac OS X UUIDs for access
+ control, preventing access in certain configurations.
+ - The IPP backend incorrectly reset the required authentication
+ to Kerberos when authentication failed.
+ - The scheduler no longer looks up the local hostname by default;
+ turn on hostname lookups to restore the previous behavior.
+ - The scheduler did not always load MIME type rules correctly
+ (STR #3059)
+ - The test page did not format correctly on A4 paper (STR #3060)
- The web interface sometimes incorrectly redirected users to
127.0.0.1 (STR #3022)
- cupsPrintFile*() did not send the document filename for single
- The scheduler now rejects ATTR: messages with empty values.
- The scheduler could consume all CPU handling closed connections
(STR #2988)
- - The scheduler now protects against DNS rebinding attacks on
- localhost.
- - SECURITY: The PNG image reading code did not validate the
- image size properly, leading to a potential buffer overflow
- (STR #2974)
- Fixed some configure script bugs with rc/xinetd directories
(STR #2970)
- The Epson sample driver PPDs contained errors (STR #2979)