Closed michaelrsweet closed 12 years ago
CUPS.org User: rickr.mn.rr
Can we pin down the PPD variable name and values so I can prepare for this?
CUPS.org User: mike
Moving to 1.5.
As for the keyword, patience grasshopper, will propose something soon...
CUPS.org User: rickr.mn.rr
"Can we pin down the PPD variable name and values so I can prepare for this?"
Bump!
CUPS.org User: till.kamppeter
This is fixed now in the libusb-based USB backend, as I have implemented a bi-di support in it which is based on the way backend/usb-darwin.c (Mac OS X) does (see STR #3477 and STR #2890). This bi-di method has no restrictions to certain manufacturers and therefore it will work with all printers. The fix is in both 1.5.x and 1.6.x SVN repositories, so it will appear in the CUPS 1.5.2 release for the first time and it will also be in Ubuntu Precise (12.04 LTS). Important is that you build the libusb-based version of the backend. You do not need to blacklist the usblp kernel modules as the backend automatically unlinks it from the printer it connects to (and even relinks it after the connection).
So this STR can be closed as fixed, but it must be mentioned in the documentation that bi-di with the usblp-based backend is restricted and the build system should default to the libusb version. Perhaps the usblp version should be removed as it does not give any advantage any more.
CUPS.org User: mike
Duping to STR #2890 which tracked the addition of bidi support to the libusb-based backend. As Till points out, we now support it for all printers.
Version: -feature CUPS.org User: rickr.mn.rr
In 1.4.2 backend/usb-unix.c: / * Disable backchannel data when printing to Brother, Canon, or * Minolta USB printers - apparently these printers will return * the IEEE-1284 device ID over and over and over when they get * a read request... /
But a USB connected Konica-Minolta magicolor 1600W (and others, e.g. 1680MF, 1690MF, and Okidata C110) has bidi capability. Only the Oki C110 works right. I patched this on my copy of backend/usb, and no harm is done. The marker levels from /usr/lib/cups/filter/command2foo2lava-pjl are right now.
From: Rick Richardson Date: 08:03 Fri
I think this should be in a file (hostname/resource/blacklist/whitelist)and not compiled in.
From: Michael Sweet Date: 08:08 Fri
Please do. It would make sense to add a PPD keyword for this, such that the driver can specify whether bidi is wanted/needed.