2007-12-07  Till Kamppeter  <till.kamppeter@gmail.com>

	* systemv/hal_lpadmin: Let hal_admin not fail if "lsusb" or
	  "hp-makeuri" are not installed. Simply do not use HPLIP URIs
	  then.
	  Replaced a wrong "os.exit()" by "sys.exit()" (Ubuntu bug #160386).
	  Made check for removal of manufacturer name from model name
	  case-insensitive (Ubuntu bug #150219)
	  Removed debug helper lines forgotten in the code.

2007-12-07  Till Kamppeter  <till.kamppeter@gmail.com>

	* systemv/hal_lpadmin: Changes for more reliable auto-setup of HP
	printers, mostly based on hints from the HPLIP developers from HP:
	  - Use only "hp-makeuri" and "lsusb" for finding out which devices
	    are supported by HPLIP and which URIs they have. "hp-probe"
	    and "hp-info" produce a very high IO load
	  - Let the print queue creation being done in a child process.
	    Due to HP's UDEV-triggered firmware upload for the HP LaserJet
	    1018/1020 these printers get uncoupled from the usblp kernel
	    module still while the queue creation is not completed and the
	    uncoupling is interpreted as device removal by HAL and so HAL
	    kills the hal_admin which creates the queues. The child process
	    does not get killed and so queue creation is assured.
	  - Do only one "time.sleep(1)" in the beginning, not one before each
	    call of hp-makeuri.

	Always produce only one notification for the system-config-printer
	tray applet, even if two queues (printer and fax) are created.

2007-10-05  Tim Waugh  <twaugh@redhat.com>

	* systemv/hal_lpadmin (HalPrinter.remove): Build command line more
	carefully.

2007-09-18  Till Kamppeter  <till.kamppeter@gmail.com>

	* systemv/hal_lpadmin: Removed the workaround applied yesterday. The
	bug in system-config-printer is fixed now.

2007-09-17  Till Kamppeter  <till.kamppeter@gmail.com>

	* autogen.sh: Use "automake --add-mising --copy" to copy the files
	needed from automake. This way autogen.sh ceated a package which
	can be built on systems without automake installed.

	* systemv/hal_lpadmin: Added a workaround to make the driver chosen
	for a queue which is generated automatically being the same as if
	the queue was set up with the add-printer wizzard of
	system-config-printer. In both cases system-config-printer has used
	completely different algorithms to find out which PPDs belong to
	the detected printer model. Especially HP's PostScript printers
	always got assigned HPIJS (PCL) PPDs in queues which are
	automatically set up by hal_lpadmin.

2007-09-14  Tim Waugh  <twaugh@redhat.com>

	* configure.in: Version 0.6.13.

2007-09-14  Till Kamppeter  <till.kamppeter@gmail.com>

	* systemv/hal_lpadmin: If a print queue was disabled due to the
	corresponding printer being turned off it did not get re-enabled
	again if the queue name was not the default one generated by
	hal_lpadmin (for example a manually created queue). Fixed.

	HPLIP-based queues got auto-disabled as sson as HPLIP accessed
	them (print job sent, extra function used, or hp-toolbox
	called). due to HPLIP uncoupling the printer from the "usblp"
	kernel module letting the /dev/usblp? file disappear. Now it is
	checked whether the printer is still available to HPLIP before
	auto-disabling the queue. Side effect: If after a accessing the
	printer once the /dev/usblp? file disappeared, HAL does not call
	hal_lpadmin any more when the printer gets really turned off. This
	means auto-disabling only works with print queues not using HPLIP.

	hal_lpadmin does not make CUPS scanning for available devices any
	more ("lpinfo -v" equivalent). These calls were only done to
	discover available HPLIP URIs. Now HPLIP is directly called which
	is much less time-consuming. This makes auto-setup and
	auto-re-enabling of HP printers vastly faster.

2007-08-10  Tim Waugh  <twaugh@redhat.com>

	* systemv/hal_lpadmin: Make logging less verbose.

2007-08-06  Tim Waugh  <twaugh@redhat.com>

	* backend/hal.c (list_devices): Use make/model strings instead of
	the description string for identifying the device.  Many devices,
	including the HP Photosmart A717, provide useless device
	description fields in the IEEE 1284 ID.

2007-08-03  Tim Waugh  <twaugh@redhat.com>

	* configure.in: Version 0.6.12.

2007-08-03  Tim Waugh  <twaugh@redhat.com>

	* systemv/hal_lpadmin: [patch from Till Kamppeter]
	- Fixed generation of USB URIs: On some HP printers URIs like
	"usb:///<make>%20<model>" instead of the correct
	"usb://<make>/<model>" were generated, the manufacturer name
	"Lexmark International" did not get corrected to "Lexmark".
	- HPLIP URIs were only used if the device reports its serial number.
	Some HP devices (even of the current models) do not report their
	serial numbers. Now we fall back to identifying them via the model
	name. So all HPLIP-supported HP printers will be set up with HPLIP
	URIs now.
	- If an HP multi-function device has fax functionality, also the
	fax queue gets auto-generated when connecting the device for the
	first time.
	- If a printer gets turned off or disconnected, the queues for it get
	disabled now, so that queued jobs do not get retried every 30
	seconds. When reconnecting and turning back on the device the
	queues are enabled again so that queued jobs get printed. Ny using
	a dedicated state message manually disabled printers are not enabled
	when reconnecting them.

	* systemv/10-hal_lpadmin.fdi: Don't use full path to hal_lpadmin
	since that gets ignored anyway.

2007-08-01  Tim Waugh  <twaugh@redhat.com>

	* systemv/hal_lpadmin (HalPrinter.configure): Use CUPS directly
	instead of foomatic, just like the add method (patch from Till
	Kamppeter).

2007-06-14  Tim Waugh  <twaugh@redhat.com>

	* configure.in: Version 0.6.11.

2007-06-14  Tim Waugh  <twaugh@redhat.com>

	* systemv/hal_lpadmin (HalPrinter.add): Use the new PPDs class to
	match Device IDs.

2007-06-07  Tim Waugh  <twaugh@redhat.com>

	* backend/hal.c (list_devices): Check return value of
	libhal_find_device_by_capability, not just num_printers (bug
	#243165).

2007-05-30  Tim Waugh  <twaugh@redhat.com>

	* configure.in: Version 0.6.10.

2007-05-28  Tim Waugh  <twaugh@redhat.com>

	* systemv/hal_lpadmin (HalPrinter.add): Use new D-Bus interface.

2007-05-03  Tim Waugh  <twaugh@redhat.com>

	* configure.in: Version 0.6.9.

2007-05-03  Tim Waugh  <twaugh@redhat.com>

	* backend/hal.c: Reverted paper-out detection and state handling
	changes for the moment, to re-appear after Fedora 7.

2007-04-26  Tim Waugh  <twaugh@redhat.com>

	* configure.in: Version 0.6.8.

2007-04-26  Tim Waugh  <twaugh@redhat.com>

	* backend/hal.c (main): Extract URI from argv[0] or DEVICE_URI.
	(main): Use backend exit constants from the API.
	(main): If the HAL device is not present, keep retrying.
	(main): Implement connecting-to-device state.
	(runloop): Moved write() loop to new function.  Use a select()
	loop.  On Linux, use a non-blocking descriptor and check printer
	status to detect paper-out.

	* systemv/hal_lpadmin (HalPrinter.get_cups_uris): Recognise both
	hp: and usb: URIs for the same device, for the purpose of deciding
	whether a new queue needs to be added.
	(HalPrinter.get_cups_uris): Fixed traceback.

2007-04-25  Tim Waugh  <twaugh@redhat.com>

	* configure.in: Version 0.6.7.

2007-04-25  Tim Waugh  <twaugh@redhat.com>

	* systemv/hal_lpadmin (HalPrinter.get_cups_uris): Use usb: URI
	scheme if we have enough information (bug #236720).

2007-04-04  Tim Waugh  <twaugh@redhat.com>

	* configure.in: Version 0.6.6.

2007-03-17  Tim Waugh  <twaugh@redhat.com>

	* systemv/hal_lpadmin (HalPrinter.__init__): If we can't connect
	to CUPS, log a message saying so rather than just a traceback (bug
	#232733).

2007-01-08  Tim Waugh  <twaugh@redhat.com>

	* configure.in: Version 0.6.5.

2007-01-08  Tim Waugh  <twaugh@redhat.com>

	* systemv/hal_lpadmin (HalPrinter.add): Set PPD page size before
	adding printer (bug #221841).

2007-01-04  Tim Waugh  <twaugh@redhat.com>

	* configure.in: Version 0.6.4.

2007-01-04  Tim Waugh  <twaugh@redhat.com>

	* systemv/hal_lpadmin (HalPrinter.add): Don't enable existing
	printers.  They may have been intentionally disabled by the
	sysadmin.
	(HalPrinter.remove): Don't disable removed printers.  Some
	devices, such as the HP PSC 1310, seem to disconnect as soon as a
	print job is sent and so this causes problems.
	(HalPrinter.get_cups_uris): Pause for a second to give HPLIP a
	chance to reconnect.

2007-01-02  Tim Waugh  <twaugh@redhat.com>

	* configure.in: Version 0.6.3.

2007-01-02  Tim Waugh  <twaugh@redhat.com>

	* systemv/Makefile.am, backend/Makefile.am: Tidying.

	* systemv/hal_lpadmin: Added syslogging.
	(HalPrinter.get_cups_uris): Use 'hp:' URIs when appropriate (bug
	#175555).
	(HalPrinter.add): Put PPD NickName in info line instead of just
	"Added by HAL".
	(HalPrinter.match): Removed unused method.
	(HalPrinter.add): Don't add new printer if one already exists with
	a matching URI.
	(HalPrinter.remove): Likewise with removal.

	* backend/hal.c (list_devices): Include IEEE 1284 Device ID in
	CUPS backend output when listing devices (bug #214755).

	* systemv/hal_lpadmin (HalPrinter.read): Fetch commandset from
	HAL.
	(HalPrinter.add): Use commandset when matching printers (bug
	#214180).

	* systemv/10-hal_lpadmin.fdi: Added method_argnames.

	* systemv/hal_lpadmin (HalPrinter.get_cups_uris): Use correct URI
	form for hal backend (bug #200682).
	(HalPrinter.add): Use correct D-Bus API name for communicating
	with eggcups.

	* configure.in: Applied remove-cups-backenddir-check patch from
	RPM package.

2006-06-02 Florian Festi <ffesti@redhat.com>

	* reimplemented hal_lpadmin 
	* new hal_lpadmin replaces cups-config-daemon
	
2006-01-12  Christopher Aillon  <caillon@redhat.com>

	* configure.in: Post-release bump to 0.5.6

2006-01-12  Christopher Aillon  <caillon@redhat.com>

        * cups-config-daemon/Makefile.am: Build with PIE enabled

