Changes in CUPS v1.4.7:
- Documentation changes (Issue #3710, Issue #3720, Issue #3745, Issue #3750, Issue #3757, Issue #3758, Issue #3782, Issue #3826, Issue #3829, Issue #3837)
- Web interface fixes (Issue #3412, Issue #3345, Issue #3455, Issue #3707, Issue #3755, Issue #3769, Issue #3783)
- Configure script fixes (Issue #3659, Issue #3691)
- Compilation fixes (Issue #3718, Issue #3771, Issue #3774)
- The imageto* filters could crash with bad GIF files (Issue #3867)
- The scheduler might leave old job data files in the spool directory (Issue #3795)
- CUPS did not work with locales using the ASCII character set (Issue #3832)
- httpAddrString() did not return a URI-style IPv6 numeric address (Issue #3814)
- Fixed an issue when reading compressed CUPS raster streams (Issue #3812)
- Fixed an issue with PostScript printer auto-configuration (Issue #3443)
- Fixed some compatibility issues with the libusb-based USB backend (Issue #3799)
- The network backends no longer try to collect SNMP supply and status information for raw queues (Issue #3809)
- The DBUS notifier did not report job state changes (Issue #3805)
- The scheduler did not always report that the "normal" print-quality value was supported (Issue #3803)
- The gziptoany filter did not report the correct error if it was unable to write the uncompressed document to the next filter or backend in the chain (Issue #3797)
- The Epson and Oki 9-pin drivers had a bad resolution option (Issue #3798)
- The scheduler did not always register the correct default ICC profile on Mac OS X.
- The scheduler did not use the job owner when authorizing access for the CUPS-Get-Document operation, preventing non-admins from accessing their own jobs.
- CUPS did not work with some printers that incorrectly implemented the HTTP/1.1 standard (Issue #3778, Issue #3791)
- The scheduler did not retry fax jobs properly.
- The scheduler now recognizes an empty cupsCommands PPD keyword as meaning that CUPS commands are not supported for a printer (Issue #3773)
- Fixed a crash bug in the scheduler when the application/octet-stream MIME type was not defined (Issue #3690)
- Polled printers were advertised more slowly than necessary (Issue #3574)
- cupsResolveConflicts() did not handle resolving multiple UIConstraints issues (Issue #3705)
- The SetEnv and PassEnv directives had no effect (Issue #3664)
- The libusb-based USB backend printed slowly to the LaserJet 1300 and other printers (Issue #3405)
- "lp" and "lpr" failed to print with Kerberos enabled (Issue #3768)
- The cupsctl program now displays an error if you try to directly set the Port or Listen directives (Issue #3749)
- PPD files with "*JobPatchFile: bla" no longer fail to load in relaxed conformance mode (Issue #3747)
- The scheduler generated a bad notify-text string for printer state change notifications (Issue #3739)
- The scheduler incorrectly updated printers.conf when it really needed to update classes.conf or remote.cache (Issue #3726)
- Hardwired remote printers with options did not work (Issue #3717)
- Accessing the CUPS web interface using a CNAME-based hostname would sometimes fail due to redirection to the actual hostname (Issue #3701)
- Subscription events had a misspelled attribute (Issue #3693)
- "make check" failed if LC_MESSAGES was set (Issue #3765)
- Fixed the configure script to always look for the pkg-config script (Issue #3761)
- The scheduler now only looks up interface hostnames if HostNameLookups are enabled (Issue #3737)
- Fixed a compilation problem on DragonFly BSD (Issue #3738)
- The default PageLogFormat value had the username and job ID swapped from CUPS 1.3.x (Issue #3727)
- The scheduler could crash if a browsed printer times out while a job is printing (Issue #3754)
- The scheduler incorrectly mapped custom page sizes to standard sizes (Issue #3764)
- cupsfilter and pstops did not map IPP attributes to PPD options due to a change in cupsMarkOptions (Issue #3756)
- The scheduler did not always show the most recent status message from the print filters (Issue #3731)
- The PostScript filter did not apply the mirror and number-up options properly, leading to offset and clipped output (Issue #3732)
- The network backends always reported "low toner" or "out of toner" states, even for inkjet printers (Issue #3733)