diff options
author | Jörg Frings-Fürst <debian@jff-webhosting.net> | 2014-10-06 14:00:40 +0200 |
---|---|---|
committer | Jörg Frings-Fürst <debian@jff-webhosting.net> | 2014-10-06 14:00:40 +0200 |
commit | 6e9c41a892ed0e0da326e0278b3221ce3f5713b8 (patch) | |
tree | 2e301d871bbeeb44aa57ff9cc070fcf3be484487 /doc/sane-apple.man |
Initial import of sane-backends version 1.0.24-1.2
Diffstat (limited to 'doc/sane-apple.man')
-rw-r--r-- | doc/sane-apple.man | 252 |
1 files changed, 252 insertions, 0 deletions
diff --git a/doc/sane-apple.man b/doc/sane-apple.man new file mode 100644 index 0000000..2d388bf --- /dev/null +++ b/doc/sane-apple.man @@ -0,0 +1,252 @@ +.TH sane\-apple 5 "11 Jul 2008" "@PACKAGEVERSION@" "SANE Scanner Access Now Easy" +.IX sane\-apple +.SH NAME +sane\-apple \- SANE backend for Apple flatbed scanners +.SH DESCRIPTION +The +.B sane\-apple +library implements a SANE (Scanner Access Now Easy) backend that +provides access to Apple flatbed scanners. At present, the following +scanners are supported from this backend: +.PP +.br +.ft CR +.nf +--------------- ----- ------------------ ------ +AppleScanner 4bit 16 Shades of Gray +OneScanner 8bit 256 Shades of Gray +ColorOneScanner 24bit RGB color 3-pass +.fi +.ft R + +.PP +If you own a Apple scanner other than the ones listed above that +works with this backend, please let us know by sending the scanner's +model name, SCSI id, and firmware revision to +.IR sane\-devel@lists.alioth.debian.org . +See http://www.sane\-project.org/mailing\-lists.html for details on how to subscribe to +sane\-devel. + +.SH "DEVICE NAMES" +This backend expects device names of the form: +.PP +.RS +.I special +.RE +.PP +Where +.I special +is either the path-name for the special device that corresponds to a +SCSI scanner. For SCSI +scanners, the special device name must be a generic SCSI device or a +symlink to such a device. Under Linux, such a device name could be +.I /dev/sga +or +.IR /dev/sge , +for example. See sane\-scsi(5) for details. +.SH CONFIGURATION +The +.I apple.conf +file is a list of options and device names that correspond to Apple +scanners. Empty lines and lines starting with a hash mark (#) are +ignored. See sane\-scsi(5) on details of what constitutes a valid +device name. +.PP +Options come in two flavors: global and positional ones. Global +options apply to all devices managed by the backend, whereas positional +options apply just to the most recently mentioned device. Note that +this means that the order in which the options appear matters! + +.SH SCSI ADAPTER TIPS +SCSI scanners are typically delivered with an ISA SCSI adapter. +Unfortunately, that adapter is not worth much since it is not +interrupt driven. It is sometimes possible to get the supplied card +to work, but without an interrupt line, scanning will put so much load on +the system that it becomes almost unusable for other tasks. +.SH FILES +.TP +.I @CONFIGDIR@/apple.conf +The backend configuration file (see also description of +.B SANE_CONFIG_DIR +below). +.TP +.I @LIBDIR@/libsane\-apple.a +The static library implementing this backend. +.TP +.I @LIBDIR@/libsane\-apple.so +The shared library implementing this backend (present on systems that +support dynamic loading). +.SH ENVIRONMENT +.TP +.B SANE_CONFIG_DIR +This environment variable is list of directories where SANE looks +for the configuration file. Under UNIX directory names are +separated by a colon (`:'), under OS/2 by a semi-colon (`;'). +If SANE_CONFIG_DIR is not set, SANE defaults to +searching the current working directory (".") and then /etc/sane.d. +If the value of $SANE_CONFIG_DIR ends with the separator +character, the default directories are searched after the directory list. +For example, setting +.B SANE_CONFIG_DIR +to "/tmp/config:" would result in directories "tmp/config", ".", and +"@CONFIGDIR@" being searched (in that order). +.TP +.B SANE_DEBUG_APPLE +Controls the debug level. A value of 255 prints +all debug output. Smaller values reduce verbosity. Requires a +library compiled with debug support. + +.SH CURRENT STATUS +The apple backend is now in version 0.3 (Tue Jul 21 1998). Since +I only have the AppleScanner and not the other models (OneScanner, +ColorOneScanner) I can only develop/test for the AppleScanner effectively. +However with this release I almost completed the gui part of all scanners. +Most of the functionality is there. At least OneScanner should scan +at the AppleScanner's compatible modes (LineArt, HalfTone, Gray16). My +personal belief is that with a slight touch of debugging the OneScanner +could be actually usable. The ColorOneScanner needs more work. AppleScanner +is of course almost fully supported. + +.SH MISSING FUNCTIONALITY +Currently all three models lack upload/download support. +.TP +.B AppleScanner +Cannot up/download a halftone pattern. +.TP +.B OneScanner +Cannot up/download halftone patterns or calibration vectors. +.TP +.B ColorOneScanner +Cannot up/download halftone patterns, calibration vectors, +custom Color Correction Tables (CCT) and of course custom gamma tables. +.TP +.B Park/UnPark (OneScanner, ColorOneScanner) +.PP +The above functionalities are missing because I don't +have the hardware to experiment on. Another reason is my lack +of understanding as to how or if the SANE API provide means +to describe any array type besides gamma. + + +.SH UNSUPPORTED FEATURES +The following "features" will never be supported, at least while I maintain +the sane\-apple backend. +.TP +.B NoHome (AppleScanner) +The scanner lamp stays on and the carriage assembly remains where it stops +at the end of the scan. After two minutes, if the scanner does not receive +another SCAN command, the lamp goes off and the carriage returns to the home +position. +.TP +.B Compression (AppleScanner) +The Scanner can compress data with CCITT Group III one dimensional algorithm +(fax) and the Skip White Line algorithm. +.TP +.B Multiple Windows (AppleScanner) +AppleScanner may support multiple windows. It would be a cool feature +and a challenge for me to code if it could intermix different options +for different windows (scan areas). This way it could scan a document +in LineArt mode but the figures in it in Gray and at a different resolution. +Unfortunately this is impossible. +.TP +.B Scan Direction (OneScanner) +It controls the scan direction. (?) +.TP +.B Status/Reset Button (OneScanner) +This option controls the status of the button on the OneScanner model. You can +also reset the button status by software. + +.SH BUGS +SANE backend bugs are divided in two classes. We have +.B GUI +bugs and +.B scanner specific +bugs. +.PP +We know we have a GUI bug when a parameter is not showing up when it +should (active) or vice versa. Finding out which parameters are active +across various Apple modes and models from the documentation +.B ftp://ftpdev.info.apple.com/devworld/Technical_Documentation/Peripherals_Documentation/ +is an interesting exercise. I may have missed some dependencies. For example +of the threshold parameter the Apple Scanners Programming Guide says +nothing. I had to assume it is valid only in LineArt mode. +.PP +Scanner specific bugs are mostly due to mandatory round-offs in order to +scan. In the documentation in one place states that the width of the +scan area should be a byte multiple. In another place it says that the +width of the scan area should be an even byte multiple. Go figure... +.PP +Other sources of bugs are due to scsi communication, scsi connects and +disconnects. However the classical bugs are still there. So you may +encounter buffer overruns, null pointers, memory corruption and +.B SANE +API violations. +.TP +.B SIGSEGV on SliceBars +When you try to modify the scan area from the slice bar you have a nice +little cute core dump. I don't know why. If you select the scan area from +the preview window or by hand typing the numbers everything is fine. The +SIGSEGV happens deep in gtk library (gdk). I really cannot debug it. +.TP +.B Options too much +It is possible, especially for the ColorOneScanner, for the backend's +options panel to extend beyond your screen. It happens with mine +and I am running my X Server at 1024x768. What can I say? Try smaller +fonts in the X server, or virtual screens. +.TP +.B Weird SCSI behaviour +I am quoting David Myers Here... + +>> OS: FreeBSD 2.2.6 +.br +>> CC: egcs-1.02 +.br +Just wanted to follow up on this... I recently changed my SCSI card from +the Adaptec 2940UW to a dual-channel Symbios 786 chipset. When I started up +SANE with your driver, I managed to scan line art drawings okay, but Gray16 +scans led to a stream of SCSI error messages on the console, ultimately +hanging with a message saying the scanner wasn't releasing the SCSI bus. +This may be that the Symbios is simply less tolerant of ancient +hardware, or may be bugs in your driver or in SANE itself... + +.SH DEBUG +If you encounter a GUI bug please set the environmental variable +SANE_DEBUG_APPLE to 255 and rerun the exact sequence of keystrokes +and menu selections to reproduce it. Then send me a report with the +log attached. +.PP +If you have an Apple Macintosh with the AppleScanners driver installed, +reporting to me which options are grayed out (inactive) in what modes +would be very helpful. +.PP +If you want to offer some help but you don't have a scanner, or you +don't have the model you would like to help with, or you are +a SANE developer and you just want to take a look at how the apple backend +looks like, goto to apple.h and #define the NEUTRALIZE_BACKEND +macro. You can select the scanner model through the APPLE_MODEL_SELECT +macro. Available options are APPLESCANNER, ONESCANNER, COLORONESCANNER. +.PP +If you encounter a SCSI bus error or trimmed and/or displaced images please +set the environment variable SANE_DEBUG_SANEI_SCSI to 255 before sending me +the report. + +.SH TODO +.TP +.B Non Blocking Support +Make sane\-apple a non blocking backend. Properly support +.B sane_set_io_mode +and +.B sane_get_select_fd +.TP +.B Scan +Make scanning possible for all models in all supported modes. +.TP +.B Missing Functionality + +.SH "SEE ALSO" +sane(7), sane\-scsi(5) + +.SH AUTHOR +The sane\-apple backend was written not entirely from scratch by +Milon Firikis. It is mostly based on the mustek backend from +David Mosberger and Andreas Czechanowski |