Some of the print modes for DeskJet 845C are incorrectly defined

Problem 1: when trying to print a test page in "Normal Grayscale" or "Photo" mode, the printer remains idle and its status in CUPS is set to "/usr/libexec/cups/filter/hpcups failed". The following is the error message in /var/log/cups/error_log which is the same for both of these modes:

D [14/May/2011:17:06:13 +0600] [Job 414] prnt/hpcups/Pcl3Gui.cpp 75: Requested resolution not supported with requested printmodeprnt/hpcups/HPCupsFilter.cpp 420: m_Job initialization failed with error = 25

The problem is the BaseResX and BaseResY values found in dj8x5PrintModes.h file not being equal to horizontal and vertical resolution specified in the PPD. After changing the former values to 600 printing in "Normal Grayscale" mode starts successfully.

Problem 2: after doing the abovementioned fix the printer still doesn't print in "Photo" mode. When trying to print a test page the paper is fed, the cartridge carriage moves back and forth for only one time and then the printer remains idle for an infinite time with a green light flashing.

The output of hpcups differs from the output of other working drivers (gutenprint, windows proprietary) using the same print mode in that the escape code specifying the quality reads as Esc*o2M for hpcups while other drivers send it as Esc*o1M. Changing the OutputType option by which the number in the above escape code is specified in PPD from OutputType(2) to OutputType(1) makes the printer to print the job successfully.

Problem 3: colors in "Normal Color" mode are output incorrectly, that shows in light rings on the color wheel. The problem here is the color depth for color cartridge in dj8x5PrintModes is set to 1 at the same time when a 4-level halftoning table (HT300x3004level_open) is used. Since the windows driver uses 4-level color in normal quality mode, the solution is to set the color depth for all color planes except black to 2. After doing that printing works as should.

The attached patch contains all the above fixes.

P. S. The question at https://answers.launchpad.net/hplip/+question/92772 describes an issue identical to the first mentioned here. Maybe the print mode definitions for that printer also need reviewing.