v1.2.5

Changes in CUPS v1.2.5:

• Documentation updates (Issue #2038)
• The SNMP backend no longer uses IPP for Epson printers (Issue #2028)
• Updated the configure script for Tru64 UNIX 5.1 (Issue #2033)
• Tru64 5.1B's getaddrinfo() and getnameinfo() functions leak file descriptors (Issue #2034)
• cupsAddDest() didn't add the parent destination's options and attributes.
• ppdConflicts() did not handle custom option constraints.
• Raw printing of gzip'd files did not work (Issue #2009)
• The scheduler no longer preserves default option choices when the new PPD no longer provides the old default choice (Issue #1929)
• The Linux SCSI backend is now only built if the SCSI development headers are installed.
• USB printing to Minolta printers did not work (Issue #2019)
• Windows clients could not monitor the queue status (Issue #2006)
• The scheduler didn't log the operation name in the access_log file for Create-Job and Print-Job requests.
• The PostScript filter now separates collated copies with any required JCL commands so that JCL-based finishing options act on the individual copies and not all of the copies as a single document.
• The PostScript filter now disables duplex printing when printing a 1-page document.
• cups-lpd didn't pass the correct job-originating-host-name value (Issue #2023)
• Fixed some speling errors in the German message catalog (Issue #2012)
• cupstestppd did not catch PPD files with bad UIConstraints values (Issue #2016)
• The USB backend did not work with the current udev- created printers if the first printer was disconnected (Issue #2017)
• Mirrored and rotated printing did not work with some documents (Issue #2004)
• 2-sided printing with banners did not work properly on some printers (Issue #2018)
• Updated the raw type rule to handle PJL within the first 4k of a print job (Issue #1969)
• Added an Estonian translation (Issue #1957)
• Clarified the documentation for the cupsd.conf @Local and @if(name) allow/deny functionality (Issue #1992)
• The PostScript filters did not escape the Title and For comments in the print job header (Issue #1988)
• The scheduler would use 100% CPU if browsing was disabled and the cupsd.conf file contained BrowsePoll lines (Issue #1994)
• The cupsDirRead() function did not work properly on non-POSIX-compliant systems (Issue #2001)
• The cupsFile functions didn't handle read/write errors properly (Issue #1996)
• The DBUS support now works with older versions of the DBUS library.

v1.2.4

Changes in CUPS v1.2.4:

• The --with-printcap configure option did not work (Issue #1984)
• The character set reported by cupsLangGet() did not always reflect the default character set of a given locale (Issue #1983)
• Older Lexmark and Tektronix printers did not work with IPP (Issue #1980)
• Failsafe printing did not work (PR #6328)
• Some web interface redirects did not work (Issue #1978)
• The web interface change settings button could introduce a "Port 0" line in cupsd.conf if there was no loopback connection available (Issue #1979)
• The web interface change settings and edit configuration file buttons would truncate the cupsd.conf file (Issue #1976)
• The German web interface used the wrong printer icon images (Issue #1973)
• The "All Documents" link in the on-line help was missing a trailing slash (Issue #1971)
• The Polish web interface translation used the wrong URLs for the job history (Issue #1963)
• The "reprint job" button did not work (Issue #1956)
• The scheduler did not always report printer or job events properly (Issue #1955)
• The scheduler always stopped the queue on error, regardless of the exit code, if the error policy was set to "stop-printer" (Issue #1959)
• ppdEmitJCL() included UTF-8 characters in the JCL job name, which caused problems on some printers (Issue #1959)
• Fixed a buffering problem that cause high CPU usage (Issue #1968)
• The command-line applications did not convert command-line strings to UTF-8 as needed (Issue #1958)
• cupsDirRead() incorrectly aborted when reading a symbolic link that pointed to a file/directory that did not exist (Issue #1953)
• The cupsInterpretRasterPPD() function did not handle custom page sizes properly.

v1.2.3

Changes in CUPS v1.2.3:

• The scheduler did not send job-state or job-config-changed events when a job was held, released, or changed (Issue #1947)
• The scheduler now aborts if the configuration file and directory checks fail (Issue #1941)
• Fixed a problem with ippPort() not using the port number that was set via the client.conf file or CUPS_SERVER environment variable (Issue #1945)
• HTTP headers were not buffered (Issue #1899)
• Some IPP printers (HP) did not like UTF-8 job names (Issue #1837)
• The CUPS desktop icon is now localized for Polish (Issue #1920)
• Printer options were not always honored when printing from Windows clients (Issue #1839)
• The openssl command would lock up the scheduler when generating an encryption certificate on some platforms due to a lack of entropy for the random number generator (Issue #1876)
• The web admin page did not recognize that "Listen 631" enabled remote access (Issue #1908)
• The web admin page did not check whether changes were made to the Basic Server Settings check boxes (Issue #1908)
• The IPP backend could generate N*N copies in certain edge cases.
• The scheduler did not restore remote printers properly when BrowseShortNames was enabled (Issue #1893)
• Polling did not handle changes to the network environment on Mac OS X (Issue #1896)
• The "make test" subscription tests used invalid notify-recipient-uri values (Issue #1910)
• Printers could be left in an undefined state on system sleep (Issue #1905)
• The Berkeley and System V commands did not always use the expected character set (Issue #1915)
• Remote printing fixes (Issue #1881)
• The cupstestppd utility did not validate translation strings for custom options properly.
• Multi-language PPD files were not properly localized in the web interface (Issue #1913)
• The admin page's simple settings options did not check for local domain socket or IPv6 addresses and did not use "localhost" as the listen address.
• An empty BrowseProtocols, BrowseLocalProtocols, or BrowseRemoteProtocols line would crash the scheduler instead of disabling the corresponding browsing options.
• The scheduler now logs IPP operation status as debug messages instead of info or error.
• cupsFileRewind() didn't clear the end-of-file state.
• cupstestppd didn't report the actual misspelling of the 1284DeviceID attribute (Issue #1849)
• BrowseRelay didn't work on Debian (Issue #1887)
• configure --without-languages didn't work (Issue #1879)
• Manually added remote printers did not work (Issue #1881)
• The <cups/backend.h> header was not installed.
• Updated the build files for Autoconf 2.60 (Issue #1853)
• The scheduler incorrectly terminated the polling processes after receiving a partial log line.
• The cups-lpd mini-daemon reported "No printer-state attribute found" errors when reporting the queue status (PR #6250, Issue #1821)
• SNMP backend improvements (Issue #1737, Issue #1742, Issue #1790, Issue #1835, Issue #1880)
• The scheduler erroneously reported an error with the CGI pipe (Issue #1860)
• Fixed HP-UX compile problems (Issue #1858, Issue #1859)
• cupstestppd crashed with some PPD files (Issue #1864)
• The <cups/dir.h> and <cups/file.h> header files did not work with C++.

v1.2.2

Changes in CUPS v1.2.2:

• Documentation updates (Issue #1765, Issue #1780)
• CUPS didn't know about alternate character set names for Asian text (Issue #1819)
• The lpoptions -o and -r options did not work unless you specified a printer.
• The lpoptions command incorrectly allowed users to set printer attributes like printer-type (Issue #1791)
• httpWait() did not flush the write buffer, causing "bad request" errors when communicating with CUPS 1.1.x servers (Issue #1717)
• Polling did not sanitize the printer description, location, or make and model strings like broadcasts did.
• Polled printers did not show the server's default job-sheets option value.
• The Samba password prompt was not properly localized (Issue #1814)
• Added a German translation (Issue #1842)
• The scheduler now creates self-signed SSL certficates automatically when using OpenSSL and CDSA for encryption, just as for GNU TLS.
• The SNMP backend sporatically reported some printers as "unknown" (Issue #1774)
• The scheduler now forces BrowseTimeout to be at least twice the BrowseInterval value and non-zero to avoid common configuration errors.
• The scheduler incorrectly returned printer URIs of the form "ipp://server/printers/classname" for classes (Issue #1813)
• Updated Japanese localization (Issue #1805)
• The scheduler's SSL certificate/key directory was not created on installation (Issue #1788)
• Added a mailto.conf man page and help page (Issue #1754)
• The parallel and USB backends no longer wait for the printer to go on-line - this caused problems with certain printers that don't follow with the IEEE-1284 standard (Issue #1738)
• The scheduler could crash on a reload when implicit classes were present (Issue #1828)
• The IPP backend incorrectly used the CUPS_ENCRYPTION environment variable to determine the default encryption mode when printing (Issue #1820)
• USB printing did not work on Solaris (Issue #1756)
• The scheduler sorted job priorities in the wrong order (Issue #1811)
• The scheduler did not automatically restart notifiers that exited or crashed (Issue #1793)
• IPv6 support did not work on NetBSD (Issue #1834)
• The EPM packaging file did not work (Issue #1804)
• The scheduler used up the CPU if BrowseRemoteProtocols was empty (Issue #1792)
• Custom page sizes did not work (Issue #1787)
• The SNMP backend could crash on some systems when SNMP logging was enabled (Issue #1789)
• Browsing could produce some funny printer names when ServerName was set to an IP address (Issue #1799)
• Fixed the log message for BrowseRelay (Issue #1798)
• Fixes to allow CUPS to compile on MirBSD (Issue #1796)
• The scheduler incorrectly set the FINAL_CONTENT_TYPE environment variable (Issue #1795)
• The pdftops filter incorrectly embedded a "produced by" comment, causing PDF printing not to work on some operating systems (Issue #1801)
• Sending raw jobs from a client system could cause the client's scheduler to eventually crash (Issue #1786)
• The scheduler now checks that the notifier exists prior to accepting a new subscription request.
• The scheduler now reports the supported notify-recipient schemes based on the contents of the ServerBin/notifier directory.
• Event notifications did not include the notify-sequence-number or other required attributes (Issue #1747)
• Allow/Deny addresses of the form "11.22.33.*" did not work on Linux (Issue #1769)
• cupsGetPPD() did not work if the scheduler was only listening on a domain socket (Issue #1766)
• The scheduler could crash advertising a class (Issue #1768)
• The scheduler could crash if the default printer was deleted (Issue #1776)
• Added a new default CUPS raster format (v3) which does not compress the raster stream in order to provide the same cupsRasterReadPixels() and cupsRasterWritePixels() performance as CUPS 1.1.x.
• The cupsaddsmb man page listed the wrong files for the CUPS driver.
• Some configure --with options did not work (Issue #1746)
• "Allow @if(name)" didn't work if "name" wasn't the first network interface (Issue #1758)
• The lpstat command did not use the correct character set when reporting the date and time (Issue #1751)
• The cupsaddsmb command and web interface did not update the Windows PPD files properly, resulting in corrupt PPD files for the Windows client to use (Issue #1750)
• The cupsd.conf man page didn't describe the Listen domain socket syntax (Issue #1753)
• The scheduler no longer tries to support more than FD_SETSIZE file descriptors.
• CDSA (encryption) support fixes for MacOS X.
• The lppasswd program needs to be setuid to root to create and update the /etc/cups/passwd.md5 file (Issue #1735)
• 32/64-bit library installation was broken (Issue #1741)
• The USB backend now reports a "no such device" error when using the old filename-based USB URIs instead of the "success" error.
• Increased the HTTP and IPP read timeouts to 10 seconds, as 1 second was too short on congested networks (Issue #1719)
• The SNMP backend now uses the device description over the printer-make-and-model attribute when the attribute contains a generic name (Issue #1728)
• Fixed another file descriptor leak when printing raw files (Issue #1736)
• Raw queues were not shared via LDAP (Issue #1739)
• The pstops filter didn't always embed PageSetup commands from the PPD file (Issue #1740)
• "make install" didn't work if you disabled all of the localizations.
• The scheduler didn't always choose the least costly filter.
• Fixed parsing of IPv6 addresses in Allow, Deny, BrowseAllow, BrowseDeny, and BrowseRelay directives (Issue #1713)
• Printers that were shared via LDAP did not get added to the LDAP server properly (Issue #1733)
• LDAP browsing would crash the scheduler if a required value was missing (Issue #1731)
• Special cases for the "localhost" hostname did not work, causing printing to not work when the /etc/hosts file did not contain a localhost entry (Issue #1723)
• Updated the Spanish translation (Issue #1720, Issue #1770)
• Reverse-order page output was broken when N-up or landscape orientations were used (Issue #1725)
• The parallel, serial, socket, and USB backends needed print data before they would report back-channel data, causing problems with several new drivers (Issue #1724)

v1.2.12

Changes in CUPS v1.2.12:

• The PHP cups_print_file() function crashed if the options array contained non-string option values (Issue #2430)
• The image/tiff file matching rule incorrectly identified some text files as TIFF files (Issue #2431)
• The filter(7) man page incorrectly documented the "PAGE: total #-pages" message (Issue #2427)
• PCL text files were mis-identified as HP-GL/2 and caused the HP-GL/2 filter to hang (Issue #2423)
• When printing to a queue with user ACLs, the scheduler incorrectly returned a quota error instead of a "not allowed to print" error (Issue #2409)
• cupsaddsmb could get in a loop if no printer drivers were installed (Issue #2407)
• cupsRasterReadHeader() did not byte-swap the header properly when compiled with certain versions of GCC.
• The IPP backend did not send the document-format attribute for filtered jobs (Issue #2411)
• Some PPD files could cause a crash in ppdOpen2 (Issue #2408)
• The web admin interface incorrectly handled the "share printers" and "show remote printers" settings (Issue #2393)
• The scheduler's log messages about AuthClass and AuthGroupName advised using a replacement directive but had the wrong syntax (Issue #2400)
• Updated the PostScript/PJL and HP-GL/2 MIME rules to look in the first 4k of the file, not just the first 1k (Issue #2386)
• Updated the Italian localization (Issue #2382)

v1.2.11

Changes in CUPS v1.2.11:

• Fixed the "relaying from" log message (Issue #2376)
• Updated the launchd support on Mac OS X to better support reconfiguration.
• "make distclean" didn't remove all generated files (Issue #2366)
• Fixed a bug in the advertisement of classes (Issue #2373)
• The IPP backend now stays running until the job is actually printed by the remote server; previously it would stop monitoring the job if it was held or temporarily stopped (Issue #2352)
• PDF files were not always printed using the correct orientation (Issue #2348)
• The scheduler could crash if you specified a bad file: URI for a printer (Issue #2351)
• The Renew-Subscription operation now returns the notify-lease-duration value that was used (Issue #2346)
• The IPP backend sent job options to IPP printers, however some printers tried to override the options embedded in the PS/PCL stream with those job options (Issue #2349)
• ppdLocalize() now also tries a country-specific localization for when localizing to a generic locale name.
• The cupstestppd program now allows for partial localizations to reduce the size of universal PPD files.
• Chinese PPD files were incorrectly tagged with the "cn" locale (should have been "zh")
• The backends now manage the printer-state-reasons attribute more accurately (Issue #2345)
• Java, PHP, Perl, and Python scripts did not work properly (Issue #2342)
• The scheduler would take forever to start if the maximum number of file descriptors was set to "unlimited" (Issue #2329)
• The page-ranges option was incorrectly applied to the banner pages (Issue #2336)
• Fixed some GCC compile warnings (Issue #2340)
• The DBUS notification code was broken for older versions of DBUS (Issue #2327)
• The IPv6 code did not compile on HP-UX 11.23 (Issue #2331)
• PPD constraints did not work properly with custom options.
• Regular PPD options with the name "CustomFoo" did not work.
• The USB backend did not work on NetBSD (Issue #2324)
• The printer-state-reasons attribute was incorrectly cleared after a job completed (Issue #2323)
• The scheduler did not set the printer operation policy on startup, only on soft reload (Issue #2319)
• The AP_FIRSTPAGE_InputSlot option did not clear any ManualFeed setting that was made, which caused problems with some PPD files (Issue #2318)
• cupsDoFileRequest() and cupsDoRequest() did not abort when getting an error in the response (Issue #2315)
• The scheduler did not schedule jobs properly to remote or nested classes (Issue #2317)
• Updated the mime.types and mime.convs headers to warn that the files are overwritten when CUPS is installed. Local changes should go in local.types or local.convs, respectively (Issue #2310)
• The scheduler could get in an infinite loop if a printer in an implicit class disappeared (Issue #2311)
• The pstops filter did not handle %%EndFeature comments properly (Issue #2306)
• Fixed a problem with the Polish web page printer icons (Issue #2305)
• ppdLocalize() now also localizes the cupsICCProfile attributes.
• The scheduler still had a reference to the incorrect "notify-recipient" attribute (Issue #2307)
• The "make check" and "make test" subscription tests did not set the locale (Issue #2307)
• The "make check" and "make test" subscription tests incorrectly used the notify-recipient attribute instead of notify-recipient-uri (Issue #2307)
• cupsRasterInterpretPPD() incorrectly limited the cupsBorderlessScalingFactor when specified in the job options.

v1.2.10

Changes in CUPS v1.2.10:

• ppdLocalize() now supports localizing for Japanese using the "jp" locale name used by the ppdmerge program from the CUPS DDK 1.1.0 (Issue #2301)
• _cupsAdminSetServerSettings() did not support changing of top-level directives as designed.
• The init script path check was broken.
• CUPS incorrectly used the attribute "notify-recipient" instead of "notify-recicpient-uri" in several places (Issue #2297)
• Fixed a configure script bug on MirBSD (Issue #2294)
• The pdftops filter did not limit the amount of recursion of page sets (Issue #2293)
• Custom page sizes with fractional point sizes did not work (Issue #2296)
• The lpoptions command would crash when adding or removing options on a system with no printers (Issue #2295)

v1.2.1

Changes in CUPS v1.2.1:

• "lprm -h hostname" did not work (Issue #1800)
• The web interface did not handle reloads properly for MSIE (Issue #1716)
• The configure script no longer adds linker rpath options when they are unnecessary.
• The scheduler could crash printing a debug message on Solaris (Issue #1714)
• The --enable-32bit and --enable-64bit configure options did not always work.
• The password prompt showed the domain socket address instead of "localhost" for local authentication (Issue #1706)
• The web interface filtered the list of printers even if the user wasn't logged in (Issue #1700)
• The IPP backend did not work reliably with some Xerox printers (Issue #1704)
• Trailing banners were not added when printing a single file (Issue #1698)
• The web interface support programs crashed on Solaris (Issue #1699)
• cupstestppd incorrectly reported problems with *1284DeviceID attributes (Issue #1710)
• Browsing could get disabled after a restart (Issue #1670)
• Custom page sizes were not parsed properly (Issue #1709)
• The -U option wasn't supported by lpadmin (Issue #1702)
• The -u option didn't work with lpadmin (Issue #1703)
• The scheduler did not create non-blocking back-channel pipes, which caused problems when the printer driver did not read the back-channel data (Issue #1705)
• The scheduler no longer uses chunking in responses to clients - this caused problems with older versions of CUPS like 1.1.17 (PR #6143)
• Automatic raw printing was broken (Issue #1667)
• 6-up printing was broken (Issue #1697)
• The pstops filter did not disable CTRL-D processing on the printer/RIP.
• ppdOpen*() did not load custom options properly (Issue #1680)
• "Set Printer Options" in the web interface did not update the DefaultImageableArea or DefaultPaperDimension attributes in the PPD file (Issue #1689)
• Fixed compile errors (Issue #1682, Issue #1684, Issue #1685, Issue #1690)
• The lpstat command displayed the wrong error message for a missing destination (Issue #1683)
• Revised and completed the Polish translation (Issue #1669)
• Stopped jobs did not show up in the list of active jobs (Issue #1676)
• The configure script did not use the GNU TLS "libgnutls-config" script to find the proper compiler and linker options.
• The imagetoraster filter did not correctly generate several 1, 2, and 4-bit color modes.
• cupsRasterWritePixels() could lose track of the current output row.
• cupsRasterReadPixels() did not automatically swap 12/16-bit chunked pixel data.
• Moved the private _cups_raster_s structure out of the public header.
• Updated the CUPS raster format specification to include encoding rules and colorspace definitions.
• The Zebra PPD files had the wrong PostScript code for the "default" option choices.
• The imagetoraster filter did not generate correct CIE XYZ or Lab color data.
• The cups-config script did not work when invoked from a source directory (Issue #1673)
• The SNMP backend did not compile on systems that used the getifaddrs emulation functions (Issue #1668)

v1.2.0

Changes in CUPS v1.2.0:

• Documentation updates (Issue #1618, Issue #1620, Issue #1622, Issue #1637)
• Static file copy buffers reduced from 64k to 32k to work around bogus MallocDebug library assumptions (Issue #1660)
• The scheduler did not decode the backend exit code properly (Issue #1648)
• The MacOS X USB backend did not report the 1284 device ID, nor did it fix device IDs returned by HP printers.
• The scheduler started more slowly than 1.1.x with large numbers of printers (Issue #1653)
• cupsRasterInterpretPPD() didn't support the cupsPreferredBitsPerColor attribute, and imagetoraster didn't use the new API.
• The "make test" script did not create all of the necessary subdirectories for testing (Issue #1638)
• The scheduler did not prevent rotation of logs redirected to /dev/null (Issue #1651)
• "make test" did not include the SNMP backend in the test environment (Issue #1625)
• The EPM packaging files did not work (Issue #1621)
• "Use Default Configuration" inserted a broken configuration file (Issue #1624)
• Redirects in the web interface did not always preserve the encrypted status of a connection (Issue #1603)
• Added the Apple "pap" backend.
• Added CUPS library to CUPS Image shared library linkage to support Linux --as-needed linker option (Issue #1606)
• Fixed support for --enable-pie (Issue #1609)
• The pdftops filter did not validate the length of the encryption key (Issue #1608)
• Updated the Polish localization.
• "Encryption Required" in the cupsd.conf file now only requires encryption when the connection is not over the loopback interface or domain socket.
• Printer names containing "+" were not quoted properly in the web interface (Issue #1600)
• The SNMP backend now reports the make and model in the information string so that the auto-generated printer name is more useful than just an IP address.

v1.1b5

Changes in CUPS v1.1b5:

• Documentation updates.
• The pstoraster filter didn't compile without the JPEG library.
• The cupsd server didn't support the HTTP OPTIONS request method.
• Dropped the "CLOSE" method supported by the cupsd server. (not defined in HTTP specification)
• Makefile/configure script fixes.
• Missing the job-restart template.
• Added IPP test suite for testing.
• Missing IPP documentation from binary distributions.
• Fixed multiple-document handling code when last-document not specified.
• Added more checks to IPP requests to prevent bad requests from getting through.
• Not all of the Ghostscript error output was being sent to stderr.
• The PostScript filter now added PJL commands to set the job name and display string, if supported.
• The scheduler would crash if the browse socket could not be bound. Now disables browsing if port 631 (reserved for IPP) is being used by a misbehaving daemon.
• The USB backend now looks for the older Linux 2.2.x USB printer device filenames as well as the newer ones.
• The IPP backend now uses the UTF-8 charset exclusively, since apparently only CUPS handles more than US-ASCII and UTF-8...
• Wasn't quoting ( in PostScript banners...
• Send-document requests with no document-format attribute could cause cupsd to crash.
• Old jobs in the spool directory might cause cupsd to crash.
• CUPS now supports all of the recommended job-hold-until keywords as well as name values of the form "HH:MM" and "HH:MM:SS".
• Added placeholder pointer for TLS encryption to the HTTP connection structure.
• Fixed the "fast poll" bug reported by DISA - the status pipe wasn't being closed for multi-file jobs.
• Revamped put_params code in pstoraster to fix bitmap allocation bug with FrameMaker output.
• Ripped out filename, etc. code from pstoraster as it is a potential security hole.
• Added support for RIP_CACHE environment variable in the new pstoraster.
• Fixed USB device filenames for Linux; now support new pre-2.4 devices (/dev/usb/lp#) and 2.2 devices (/dev/usblp#)
• Fixed accept-jobs crash with classes.
• Didn't include dot-matrix EPSON drivers in previous release.