sane-pint(5) | SANE Scanner Access Now Easy | sane-pint(5) |
sane-pint - SANE backend for scanners that use the PINT device driver
The sane-pint library implements a SANE (Scanner Access Now Easy) backend that provides generic access to hand-held and flatbed scanners using the PINT (PINT Is Not Twain) device driver. The PINT driver is being actively developed on the OpenBSD platform, and has been ported to a few other *nix-like operating systems.
PINT is designed to provide an ioctl(2) interface to many different scanner types. However, this backend has only been tested with flatbed single-pass scanners, and more work will probably be required to get it to use other scanner types successfully.
If have successfully used the PINT driver with your scanner, but it does not work using this SANE backend, please let us know. To do this, send a mail with the relevant information for your scanner to sane-devel@lists.alioth.debian.org. Have a look at http://www.sane-project.org/mailing-lists.html concerning subscription to sane-devel.
This backend expects device names of the form:
Where special is the UNIX path-name for the special device that corresponds to the scanner. The special device name must be a PINT device or a symlink to such a device. For example, under NetBSD or OpenBSD, such a device name could be /dev/ss0 or /dev/scan0.
The contents of the pint.conf. file is a list of device names that correspond to PINT scanners. Empty lines and lines starting with a hash mark (#) are ignored. A sample configuration file is shown below:
Gordon Matzigkeit, adapted from existing backends written by David Mosberger.
There are minor roundoff errors when adjusting the ranges, since PINT uses units of 1/1200 of an inch, and SANE normally uses millimeters. Symptoms of these errors are skewed images. This should really be fixed (no pun intended) as soon as possible, but I just don't know/care enough about fixed-point representation and roundoff errors to do this correctly. Workaround: use inches as the scanning unit, and everything usually works fine.
The PINT 0.5e interface does not provide a way to determine valid ranges for DPI, modes, and scan sizes. So, the SANE backend queries the PINT device, and dynamically discovers valid ranges by doing a binary search. This means that the driver takes longer to initialize than seems necessary.
Resetting the scanner does not seem to work (at least not on my HP ScanJet 4p). For that reason, the driver sends a SCIOCRESTART, then gobbles up any remaining input until it hits EOF.
Not all of the scanners have been identified (i.e. whether they are flatbed or handheld).
X and Y resolutions are assumed to be the same.
No testing has been done on three-pass or handheld scanners, or with Automatic Document Feeder support.
14 Jul 2008 | @PACKAGEVERSION@ |