Initial import of sane-backends version 1.0.24-1.2

1 files changed, 1172 insertions, 0 deletions

diff --git a/doc/umax/umax.CHANGES b/doc/umax/umax.CHANGES
new file mode 100644
index 0000000..77d8d62
--- /dev/null
+++ b/doc/umax/umax.CHANGES
@@ -0,0 +1,1172 @@
+CHANGES FOR UMAX BACKEND FOR SANE
+---------------------------------
+
+umax -> sane-0.6:
+-----------------
+
+    * umax-sane-backend created:
+      - umax.c splitted in:
+        umax-cli.c, umax-scsi.c, umax-driver.c, umax-text.c
+      - umax.h splitted in:
+        umax-scanner.h, umax-scsi.h, umax-scsidef.h, umax-struct.h
+      - outputs done by DBG(level,msg,args...) defined in debug.h
+      - scsi-calls changed to sane-format (of sane-0.57)
+      - created umax.c and umax.h for sane-backend
+
+    * changed search-path for scsi.h and sg.h, now needs /usr/include/scsi
+
+    * removed mistake in bound-function
+
+    * added "Vista S-6E" as supported
+
+    * added PowerLook-II (PL-II) as supported
+
+    * added Astra 600S as supported
+
+
+sane-0.6 / 0.61 / 0.62 -> 0.63
+------------------------------
+
+    * added exception to gpl-licence:
+    >   This program is free software; you can redistribute it and/or
+    >   modify it under the terms of the GNU General Public License as
+    >   published by the Free Software Foundation; either version 2 of the
+    >   License, or (at your option) any later version.
+    >
+    >   This program is distributed in the hope that it will be useful, but
+    >   WITHOUT ANY WARRANTY; without even the implied warranty of
+    >   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+    >   General Public License for more details.
+    >
+    >   You should have received a copy of the GNU General Public License
+    >   along with this program; if not, write to the Free Software
+    >   Foundation, Inc., 59 Temple Place - Suite 330, Boston,
+    >   MA 02111-1307, USA.
+    >
+    >   As a special exception, the authors of SANE give permission for
+    >   additional uses of the libraries contained in this release of SANE.
+    >
+    >   The exception is that, if you link a SANE library with other files
+    >   to produce an executable, this does not by itself cause the
+    >   resulting executable to be covered by the GNU General Public
+    >   License.  Your use of that executable is in no way restricted on
+    >   account of linking the SANE library code into it.
+    >
+    >   This exception does not, however, invalidate any other reasons why
+    >   the executable file might be covered by the GNU General Public
+    >   License.
+    >
+    >   If you submit changes to SANE to the maintainers to be included in
+    >   a subsequent release, you agree by submitting the changes that
+    >   those changes may be distributed with this exception intact.
+    >
+    >   If you write modifications of your own for SANE, it is your choice
+    >   whether to permit this exception to apply to your modifications.
+    >   If you do not wish that, delete this exception notice.
+
+    * split resolution in x-res and y-res, flag bind-resolution
+
+    * only sane_... functions will be exported from umax-backend,
+      all umax_... functions are made static
+
+    * removed umax-emergency-routine for SANE
+
+    * made program-flow (esp. with do_canel) SANE-specific
+
+    * fixed missing 3-pass parts
+
+    * made scanmodes, sources, analog gamma, gamma download, negative,
+      quality calibration, highlight, shadow only selectable if they are
+      available
+
+    * calibration-mode selectable in advanced options if available
+
+    * warmup selectable if available
+
+    * changed umax-scsi.c like described from Gordon Matzigkeit
+      in his diff/patch-file from 21.06.97 so that the umax-backend
+      should compile on non-linux-systems now - thanks!
+
+    * split highlight and shadow in R, G and B for color-mode.
+
+    * reset adf and uta flag when not used (in sane_start)
+
+    * added possibilty to change calibration-mode if available
+
+    * added send_halftone_pattern to umax-scsi.c
+
+    * changed umax-scsi.c like described from Gordon Matzigkeit
+      in his diff/patch-file from 21.06.97 so that the umax-backend
+      should compile on non-linux-systems now - thanks
+
+    * added support for each color (R,G,B) for highlight and shadow
+
+    * added "Astra 1200S" and "UC840" as supported 
+
+    * added bad bugfix for preview-problem with S6E
+      (in umax-scsi.c function start_scan)
+
+    * moved resolution-check from check_values to bound_scanarea,
+      now no floatingpoint-exception will occur if no resolution is given!
+
+    * changed range-check for width and height
+
+    * removed bug in set_window for three-pass-rgb-mode
+      -selected wrong buffer for some green and blue options
+      -scanmode now is greyscale instead of rgb
+
+    * changed start_scan(us) to better use with wid-list
+      and added missing 3 wid-bytes in scan.cmd
+
+    * removed parameter color in three_pass_function_headers,
+      the color ist taken from us->three_pass_color now.
+
+
+sane-0.63/0.64/0.65 -> 0.66
+---------------------------
+
+    * changed gamma-data-upload for 3-pass-color-scan
+
+    * corrected wrong calculation for highlight and shadow
+      selection for rgb-values
+
+    * changed gamma data to use 8 bit (256 bytes) or 10 bit (1024 bytes)
+
+    * included 10/30 bits per pixel mode
+
+    * changed bug in sane_exit (free()) that sometimes made a
+      segmentation-fault
+
+    * made patches runtime-selectable
+
+    * changed set_window_command to support 30 bits per pixel
+
+    * changed send_gamma_data to support three_pass_gamma and 10/30 bpp
+
+    * changed row_len in umax_bound_scanarea to 10/30 bit mode
+
+    * added halftone-support for ug630, uc630 and uc840,
+      but I don't know if all values are ok!
+
+    * changed CBHS-50-range to 78-178 (from 103-153)
+
+    * changed rangecheck for highlight and shadow
+
+    * fixed get_cbhs_range-bug in umax-scsidef.h
+
+    * added ADF-inquiry-bits to umax-scsidef.h
+
+    * added color_line_arrangement_mode-inquiry-byte to umax-scsidef.h
+
+    * added CCD_line_distance-inquiry-byte to umax-scsidef.h
+
+    * added all most unused bits and bytes in inquiry to umax-scsidef.h
+
+    * made patches runtime-selectable
+
+
+sane-0.66 -> 0.67
+-----------------
+
+    * removed bug in sane_start: ADF and UTA flags have not been set correct
+
+    * removed bug in init_options: value range for 10 gamma-input-bits corrected
+
+    * made custom gamma available for lineart and halftone for nonlinear effects
+
+    * added Vista-T630 as supported scanner
+
+    * removed test_warmup in wait_scanner because adaptec 1542 has problems
+      with request_sense-command in test_warmup
+      it was only used for printing a message, so it was not important.
+
+    * if a scanner has a full inquiry-block but is not known to be supported,
+      the driver will print a warning and continue instead of aborting,
+      option --continue (on umax-cli) will suppress the warning!
+
+    * changed ugly definition
+      char gamma[x] -> char *gamma; gamma=alloca(x);
+
+    * removed bug (wrong length for 10 bit-gamma) in send_gamma_data
+
+    * changed all buffer-definitions to unsigned char and insterted
+      type conversions, so there are no more warnings because of
+      typemismatches.
+
+    * corrected bad definition of set_DCF2_gamma_output_bits
+
+
+sane-0.67 -> 0.68
+-----------------
+
+    * changed open config to fp = sanei_config_open(UMAX_CONFIG_FILE)
+
+    * changed all set OPT...BIND commands, endless-loop happend in sane-0.67
+      =>ane-0.67-umax-fix.diff.gz
+      after that patch changed same for ten_bit_mode
+
+    * changed some strdup() to (SANE_Char *)strdup()
+
+    * added UMAX Astra 610S as supported scanner. it has the same problems
+      in color-mode like the Astra 600S.
+
+
+sane-0.68 -> 0.69
+-----------------
+
+    * changed maximum y-resolution from optical_res to y_res
+
+    * changed sane_cancel (waitpid) like suggested by David Mosberger
+
+    * added #include <lalloca.h> for AIX-support
+
+    * replaced call to umax_wait_scanner with umax_do_calibration
+      after umax_start_scan in umax.c
+
+    * moved sense_handler to umax-scsi.c
+
+    * removed call to umax_get_data_buffer_status in reader_process
+
+    * added umax_send_shading_data, umax_send_gain_data
+      and   umax_read_shading_data, umax_read_gain_data
+
+    * changed name: umax_read_data_block -> umax_read_image_data
+
+    * changed length offest in do_inquiry from 5 to 4
+      and changed set_inquiry_length-offset from 5 to 4
+
+    * added calib_lines to structure US and added it to
+      set_window - command in umax-scsi.c
+
+    * changed in umax-scsi.c routine identify_scanner:
+      vendor and product is tested for each model instead of
+      vendor=="UMAX " and only model tested!
+      Because of that the format of scanner_str has changed
+      from product,product,...
+      to vendor,product, vendor,product,...
+
+    * added support for Linotype Hell Jade (LinoHell Office)
+
+    * added empty routine umax_do_calibration in umax-scsi.c, it shall be called
+      after umax_start_scan!
+
+    * replaced all umax_*scsi with sanei_*scsi, so the umax-driver
+      now directly calls the sanei_scsi-commands.
+
+    * moved sense-handler from umax.c to umax-scsi.c
+      and added detection for driver-calibration-request
+      sense_handler got a new parameter (void *arg).
+
+    * changed wait_scanner for SUN: usleep(0.1sec) -> sleep(1s)
+      this is only a bad bugfix that will be replaced one day!
+
+
+sane-0.69 -> 0.70
+-----------------
+
+    * moved reader_process to umax-scsi.c
+
+    * changed include <sane/config.h> for _AIX and non-_AIX
+
+    * added missing "," in definition line for UMAX Astra 1200S in umax-scanner.h
+
+    * added in umax-scsi.c in function identify_scanner a test that only
+      warnings and errors are printed if vendor="UMAX "
+
+    * added routine umax_order_line_to_pixel and umax_order_line
+
+    * moved reader_process from umax.c to umax-scsi.c
+      and added routine for color ordering.
+
+    * reset length offest in do_inquiry from 4 to 5
+      and set_inquiry_length-offset from 4 to 5
+      because umax-S6E didn't work with it!
+
+    * corrected error in identify-scanner in vendor-test
+
+    * changed end-detection of list of known scanners from
+      #define known_scanners 15 to last item "END_OF_LIST"
+
+    * added routine that corrects wrong inquiry-data from
+      Astra 600S and Astra 610S
+
+    * added test and selection for do_color_ordering in check_values
+
+    * changed __sun-code in umax_wait_scanner because the scsi-driver
+      answers with IOERROR instead of DEVICE_BUSY
+      (this is only a bugfix for a little time)
+
+
+sane-0.70/0.71 -> 0.72
+----------------------
+
+    * inactivated selection of quality calibration (until it is supported
+      by the driver)
+
+    * changed maximum x_resolution from optical_res to max_x_res
+
+    * reordered DBG-levels and defined them by symbols
+
+    * changed all sizes that used type int to type size_t
+
+    * changed wait_scanner for testing !sane_status_good,
+      because testing sane_status_device_busy does not work with
+      all scsi-drivers!
+
+    * changed ccd-order in umax_order_line, the order is now
+      specified by 9 values (before by 3 values), changed
+      order tables and added a table for Astra 600S and
+      another table for Astra 610S
+
+    * changed ccd-order, now based on optical_resolution
+      and added new routine umax_forget_line
+
+    * added inquiry-block for UC1200S: NOT TESTED!
+
+    * added inquiry-block for UC1260: NOT TESTED!
+
+    * included print_inquiry into umax-scsi.c
+
+    * added correction-routine for x_res > 600 dpi, phys_width is
+      corrected in umax-scsi.c: bound_scanarea
+
+    * changed umax-scsi: sense_handler:
+      changed test for valid sense-block
+      added test for button pressed
+
+    * added output of sense-code and error-code in sense_handler
+
+    * added options to select calibratin and scan exposure times
+
+    * added options to select calibratin and scan lamp density
+
+    * moved all option-names/titles/descr. to saneopts.h
+
+    * combined bind for highlight, shadow, analog-gamma and new
+      exposure time functions to rgb-bind
+      
+    * added set_WD_line_arrangement(buffer_r, WD_line_arrengement_by_driver);
+      if driver is doing color-ordering from line to pixel
+
+    * included umax-scsi.c into umax.c
+
+    * included umax-struct.h into umax.h
+
+    * included US-structure into Umax_Device-structure
+
+    * added free(scanner->device->buffer) to sane_close
+
+    * added free(scanner->device->pixelbuffer) to sane_cancel
+
+    * added free(scanner->device->pixelbuffer) to do_cancel
+
+    * added SANE_UNIT_MICROSECOND to exposure-times
+
+    * added output of status-string in wait_scanner
+
+    * moved some options for Scan-mode-block to Enhancemant-block
+
+sane-0.72 -> 0.73
+-----------------
+
+    * disabled RGB_BIND if no such value is available
+
+    * included readproc
+
+    * included bound_scanarea into check_values, problem
+      with wrong scanarea for UTA should be solved now!
+
+    * all scanmode- and scansource-strings are #DEFINEd now
+
+sane-0.73 -> 0.74
+-----------------
+
+    * added support for 12/36 bits per pixel (for Astra 1220S)
+
+    * added support for Astra 1220S
+
+    * included inquiry data for UG80 (Escom Image Scanner 256)
+
+    * moved WAIT_SCANNER from umax_read_data to umax_read_image_data,
+      umax_read_shading_data now works!
+
+    * added umax_start_scan at the end of calibration by driver,
+      the scanner does not hang any more if quality calibration is enabled!
+      There must'nt be a "wait_scanner" in calibration by driver routine!!!
+
+    * added usage of driver internal values for calibration by driver if
+      the scanner does not give back the informations about the shading data:
+      width = dev->width * dev->inquiry_optical_res ( * 3 in RGB-mode )
+
+    * added 2 calibration-routines:
+      - use last line for shading data
+      - calculate average value for shading data
+      - prepared routine for complex calculation
+
+    * added gamma download curve format type 0 and 1
+
+    * defined gamma DCF = 1 for UC1200S and UC1260
+
+    * defined gamma DCF = 0 for UC1200SE and ARCUS PLUS
+
+    * changed sense_handler: if bad sense_key => DEVICE_BUSY
+      problem with black area at top of scan for some SCSI-adapters is solved now
+
+    * changed wait_scanner: waits 0.5 sec now
+
+    * disabled quality calibration in preview mode
+
+    * made upper_left_x and upper_left_y even values
+      it looks like even resolutions > 600 dpi work now!
+      (not ok!!!)
+
+    * added sigterm_handler for umax_reader_process
+      changed sane_canel, do_canel and do_eof(=close_pipe),
+      cancel now works!
+     
+    * changed trim_rowbufsize that made problems with SGI,
+      there has been an overflow because of big buffersize
+
+      
+sane-0.74 -> 1.00 (build 0)
+---------------------------
+
+    * added output of umax-backend-version in sane_init
+
+    * added support for Linotype Hell Jade2 (Vobis ScanBoostar Premium)
+      SCSI-ID "LinoHell","Office2"
+
+    * changed base of calculation of scanwidth and scanlegth from
+      pixel to resolutionbase (1/300, 1/600 or 1/1200 inch)
+      problem with wrong pixelcount at higher resolutions should be
+      solved now
+
+    * changed selection for calibration mode
+
+    * changed calculation for width in pixels for lineart and halftone mode
+
+    * added inquiry for UMAX UC1200SE
+
+    * corrected 12/36-bit-mode in send_gamma_data and array in umax.h
+
+    * added support for Gemini D-16
+
+
+sane-1.00-build-0 -> build-1
+-----------------------------
+
+    * added support for Nikon AX-210
+
+    * changed all grey to gray (David Mosberger asked for it)
+
+    * changed bug in gray gamma table conversion
+
+    * added Linotype Hell Saphir 2 as supported scanner
+
+    * changed umax-uc1260.h
+
+    * added "- 1" in umax_order_line:
+      else if (dev->pixelline_opt_res < length * 3 - CCD_distance * 3 - 1)
+      else if (dev->pixelline_opt_res < length * 3 - CCD_distance - 1)
+
+    * changed all allloca() to malloc() + free()
+
+    * removed options speed and smear, I got the info from umax that
+      these values should never be changed!
+
+    * removed ten_bit_mode and twelve_bit_mode options, added option bit_depth
+      and added support for 9, 14 and 16 bpp
+
+    * replaced u10_range and u12_range by dynamically set scanner.output_range
+
+    * removed static allocation for gamma_table[0..3], this is done dynamically
+      in sane_open and freed in sane_close now.
+
+    * added an empty string to the end of scan_mode_list and source_list
+
+build-1 -> build-2
+------------------
+
+   * added "*info |= SANE_INFO_RELOAD_PARAMS;" in control_option set bit_depth
+
+   * added inverted calibration modes for Astra 6X0S scanners.
+
+
+build-2 -> build-3 (sane-1.0.1)
+-------------------------------
+
+   * inversion of calibration is automatically set for Astra 6X0S and Vista S6E
+
+   * check for machine endianess is done only at backend start now!
+
+   * made umax_do request_sense static
+
+
+build-3 -> build-4
+------------------
+
+   * Added option name and parameter in debug output in option set value
+
+   * Added option name in debug output in option get value
+
+   * Changed calculation of pixels/line and lines, may be the problem
+     with shaered images at resolutions>600 dpi is solved now
+     later Remark: This error does still occur, but not so often!
+
+
+build-4 -> build-5
+------------------
+
+   * Changed calculation in umax_order_line, color images should
+     have the correct length now for Astra 6X0S and Vista S6E
+
+   * Added DOR resolutions and area
+
+
+build-5 -> build-6
+------------------
+
+   * Bits for ADF and UTA in driver internal inquiry block (old UMAX scanners)
+     are set dynamically now.
+
+   * Added Genius ColorPage-HR5 (vendor KYE) (=Astra 1200S) as supported scanner
+
+   * Added Astra 2400S as experimentally supported scanner
+
+   * Added pauses for Astra 2400S
+
+   * Removed reposition_scanner at start of scan, Astra 2400S does not work with it!
+
+   * changed debug output in correct_inquiry
+
+   * for UMAX S12 the driver does not wait any more until scanner is repositioned
+     this is experimental, it may work with other scanners too!
+
+
+build-6 -> build-7
+------------------
+
+   * changed some int -> unsigned int (resolution and scanwidth etc)
+
+   * calculate_pixel is called with right values in UTA mode now.
+
+   * if DOR resolution is not defined and DOR is available
+     the DOR resolution is set to double standard resolution
+
+   * removed wait_scanner after set_window-command
+
+   * added Mirage D16-L as supported scanner
+
+   * Astra 2400S works stable now
+
+
+build-7 -> build-8
+------------------
+
+    * for Mirage D-16L the driver does not wait any more until scanner is repositioned
+      this is experimental, it may work with other scanners too!
+
+    * Mirage D-16L does not give back informations for calibration by driver,
+      added new modus that "guesses" the values for Mirage D-16L.
+
+    * Added read_shading_data to 16 bit calibration routine and made a routine for
+      averaging with low byte first and one for high byte first
+
+    * Byte order (HBT) is set always (also in lineart and halftone)
+      because it is used in driver calibration
+
+    * Added relevant_*_res the is set to normal or double resolution values
+
+    * Added test if scsi buffer is large enough for one shading line
+
+    * selection for Calibration mode, Shading type and Speed/Smear
+      are only compiled in if define is set in umax.h
+
+
+build-8 -> build-9
+------------------
+
+    * slow speed options is preset for Vista T-630
+
+    * speed/smear can be set now (forgotten #ifdef)
+
+    * selection of exposure times now is only available if it is enabeld
+
+    * gamma table is disabled in lineart and halftone modes
+
+    * if image is transmitted completly sane_read (close_pipe)
+      returns with SANE_STATUS_GOOD in adf mode and in any other case
+      with SANE_STATUS_EOF
+
+
+build-9 -> build-10
+-------------------
+
+    * removed change about sane_read returns SANE_STATUS_GOOD in ADF mode
+
+    * transformed all gif files into jpeg format because of all the
+      trouble about the gif licence. Will never use gif again!
+
+    * added correction for inquiry-ADF-bit for Astra 2400S
+
+    * added second test method to detect ADF error in sane_start:
+      inquiry block is read and cover open, jammed and no paper is tested
+
+
+build-10 -> build-11
+--------------------
+
+    * test for adf errors (sane_start) is now only done if adf is selected
+
+
+build-11 -> build-12
+--------------------
+
+    * moved adf error tests from sane_start to umax_start_scan
+
+    * sane_control_option: printing option name is set to "(none)" if pointer is NULL
+
+    * added error message after each call of sanei_scsi_cmd
+
+    * changed manipulation of ADF bit for Astra 2400S in umax_correct_inquiry
+      from dev->adf to set_inquiry_sc_adf()
+
+
+build-12 -> build-13
+--------------------
+
+    * changed inquiry data of UC1200SE
+
+    * added set_inquiry_CBHS(50) in correct_inquiry for UC1200SE
+
+    * correct_inquiry is called for scanners with short inquiry now
+
+    * dev->inquiry_max_calib_lines is preset to 66,
+      problem with Vist S6E V1.6 should be solved this way.
+
+    * inquiry block parameters > 0x93 are only used if block is long enough
+
+
+build-13 -> build-14
+--------------------
+
+    * changed inquiry_cbhs_[min,max] -> inquiry_[contrast, brightness, threshold,
+      highlight, shadow]_[min,max] and set correct values for CBHS_255 and CBHS_50
+
+    * changed cbhs_str[CBHS_50] from "78-178" to "0-50"
+
+    * changed umax-uc1200se.h inquiry byte 0x64 bit 5 to 0 (gamma download format != 2)
+      and wdb len from 0x30 to 0x31
+
+    * disabled halftone pattern download, and halftone dimension selection - not ready
+      preselected is a 8x8 built in halftone pattern
+
+    * removed set_inquiry_CBHS(50) from correct_inquiry for UC1200SE
+
+    * added "Astra 2200 " as supported scanner
+
+    * added cast from (void *) to (unsigned char*) in umax_send_gamma_data
+
+
+build-14 -> build-15
+--------------------
+
+    * add "(unsigned)" to option number test in sane_control_option
+
+    * for Astra 1200 the driver does not wait any more until scanner is repositioned
+      this is experimental, it may work with other scanners too!
+
+
+build-15 -> build-16
+--------------------
+
+    * OPT_SOURCE: removed change of width/height and "*info |= SANE_INFO_RELOAD_PARAMS;"
+      from get_value to set_value and added "*info |= SANE_INFO_RELOAD_OPTIONS;"
+     
+
+build-16 -> build-17
+--------------------
+
+    * changed calibration mode from SHADING_TYPE_AVERAGE_INVERT to SHADING_TYPE_AVERAGE
+      for UMAX S-6E, UMAX S-6EG and Vista-S6E
+
+    * changed several int -> unsigned int
+
+    * added support for sanei_scsi_open_extended and variable scsi buffer size
+
+
+build-17 -> build-18 (sane-1.0.2)
+---------------------------------
+
+    * some cosmetic changes of the form
+        if (..) {...}
+      =>
+        if (..)
+        {
+          ...
+        }
+
+    * removed unnecessary SHADING_TYPE-handling routines for UMAX S-6E, UMAX S-6EG and Vista-S6E
+
+    * scsi buffer is only reallocated when size has changed
+
+
+build-18 -> build-19
+--------------------
+
+    * added UMAX PowerLook III as supported scanner
+
+    * added EPSON Perfection600 as supported scanner (is a Astra 1200S)
+
+    * type of option 0 (Option numbers) is set to SANE_TYPE_INT now (undefined before)
+
+    * removed bug that mixed 12 bit and 14/16 bit gamma table when 14 or 16 bit are supported
+      and changed value range of grey custom gamma table:
+      - maximum output value is 255 for 8 bits/pixel
+      - maximum output value is 65535 for 9-16 bits/pixel
+      the custom gamma table does work now with 9-16 bits/pixel output mode
+
+    * added routine that reverses bit order in lineart mode when scanner sends LSB first in lineart
+
+
+build-19 -> build-20
+--------------------
+
+    * added variable calibration_width_offset, Mirage D-16L offset of 308 is defined there
+      and added offset of 28 for PowerLook III
+
+    * umax_do_calibration: guessed calibration data vaules use 2 bytes/color if scanner
+      supports gamma input bit depth supports 2 bytes/color
+
+    * calibration by driver now does work for PowerLook III
+
+    * added function umax_set_rgb_bind() that sets the sensitivity of some options
+      that make available different settings for RGB. This function now is called
+      when scanmode is selected.
+
+    * changed some int to unsigned int
+
+    * made dev->pause_after_reposition = -1; as default, so as default sane-umax does not wait
+      any more while the scanhead is moving back - this may break the support of some scanners
+      so it may have to be changed (dev->pause_after_reposition = 0;) for some scanners.
+      reported to work for:
+      - Supervista S12
+      - PowerLook III
+      - Astra 2200
+
+
+build-20 -> build-21 (sane-1.0.3)
+---------------------------------
+
+    * Added Linotype "JADE" as supported scanner
+
+    * Removed bug in sane_close that caused segfault when
+      sane_close was called twice with the same scanner handle
+
+build-21 -> build-22
+--------------------
+
+    * replaced fgets() by sanei_config_read()
+
+    * added command queueing for scsi read commands
+      - removed WAIT_SCANNER scsi command call after read command
+      - the additional buffers are allocated in the reader process
+      - the number of buffers can be defined in umax.conf
+        with option scsi-maxqueue [1..8] default:2
+      (did some tests with Supervista S12 and PowerLook III
+       on the same scsi controller (ncr53c810):
+       command queueing works, it is possible to start
+       both scanners at the same time. The scanner with the
+       higher scsi-ID (=priority) does scan without stops,
+       the other scanner has to wait long time between the
+       scsi-read commands, it also is possible to give
+       bothe scanners different scsi command queueing depth).
+
+    * prepared replacement of pipe by sanei_ipc_* routines 
+
+    * added routines umax_get_lamp_status and umax_set_lamp_status
+
+    * updated inquiry output (new bits from powerlook scsi protocol 1.0 X101)
+
+    * added new (powerlook scsi protocol 1.0 X101) window descriptor block entries to umax-scsidef.h
+
+    * added button protocol (3 buttons, from powerlook scsi protocol 1.0 X101)
+
+    * added button "lamp off" to turn off scanner lamp
+      (this option may be remoeved again in later version, may be
+       it will be replaced by an autmatic lamp turn off function)
+
+    * added options to umax.conf
+      lamp-off-at-exit [01] default:0
+      lamp-off-time [10..720] (in minutes) default:30
+
+    * if  lamp_off_at_exit == 1 and lamp_control_available == 1
+      the scanner lamp is turned off in sane_close
+
+    * renamed umax-scanner.h and umax-"scannername".h to umax*.c
+      because these files create data code
+
+
+build-22 -> build-23
+--------------------
+
+    * scsi commands set/get lamp control are send with a command length of
+      10 bytes instead of 6 bytes and control bit #7 is in byte 0x03
+      and not in byte 0x04 (it is described incorrect in the umax doc)
+
+    * enabled lamp control for Astra 2200
+
+    * removed umax.conf options lamp-off-at-exit and lamp-off-time
+
+    * added frontend option lamp-off-at-exit
+
+    * added frontend option lamp-on
+
+    * forcing calibration_bytespp = 1 for Astra 2200 - quality calibration does work now
+
+    * for Astra 2200 selection for different x/y resolutions is disabled
+
+    
+build-23 -> build-24 (sane-1.0.4)
+--------------------
+
+    * forcing calibration_bytespp = 1 for Astra 2100S
+
+    * Added Astra 2100S as supported scanner
+
+    * enabled lamp control for Astra 2100S
+
+    * Added warning in sane_control_option if option could not be set
+ 
+    
+build-24 -> build-25
+--------------------
+
+   * dev->lamp_control_available is preset to 0
+
+   * dev is erased (0) after it is allocated
+
+   * added special options for PowerLook 3000:
+     - calibration is done for each CCD pixel
+     - calibration_bytespp is set to 1
+     the scanner still is not listed as supported scanner because
+     the changes are not tested
+
+   * added options that are read from umax.conf:
+
+     - execute-request-sense:
+       values: 0=disabled, 1=enabled, default=0
+       If set to 1 umax_do_request_sense is called in umax_do_calibration.
+       This can hang the system (but has been enabled until this version)
+
+     - scsi-buffer-size-min, scsi-buffer-size-max:
+       values: 4096-1048576, default min=65536, max=131072
+
+     - scan-lines, preview-lines:
+       values: 1-65535, default scan-lines=40, preview-lines=10
+       define the maximum number of lines that are scanned into one buffer
+
+     - force-preview-bit-rgb:
+       values: 0=disabled, 1=enabled, default=0
+       set preview bit in rgb real scan
+
+     - slow-speed, care-about-smearing:
+       values: -1=auto, 0=disabled, 1=enabled, default=-1
+       dangerous options, needed for some scanners
+
+     - calibration-full-ccd:
+       values: -1=auto, 0=disabled, 1=enabled, default=-1
+       do calibration for each pixel of ccd instead of selected image
+
+     - calibration-width-offset:
+       values: -1=disabled, >=0 set value
+       add an offset width to the calculated with for image/ccd
+
+     - calibration-bytes-pixel:
+       values: -1=disabled, 0=not set, 1=1 byte/pixel, 2=x bytes/pixel
+       use # bytes per pixel for calibration
+
+     - invert-shading-data:
+       values: -1=auto, 0=disabled, 1=enabled, default=-1
+       invert shading data before sending it back to the scanner
+
+     - lamp-control-available:
+       values: 0=automatically set by driver - if known, 1=available, default=0
+
+   * updated html-documentation
+
+   * updated sane-umax.man
+
+   * added pre focus options (PowerLook 3000):
+     - FP  - fix focus position
+     - FDC - lens calibration in doc position
+     - PF  - disable pre focus
+     - LCL - use 0.6/0.0mm holder focus position
+
+   * added #define SANE_I18N(text) text to umax.c. This is to mark the backend
+     specific texts that shall be translated by the frontend
+
+    
+build-25 -> build-26
+--------------------
+
+   * added umax.conf option handle-bad-sense-error #
+
+   * added Mirage II as supported scanner
+
+   * added routine umax_set_max_geometry that defines available surface
+     and resolution in dependance of scansource and double optical resolution
+     (before this bugfix the change of dor mode overwrote the settings of
+      scansource and vice versa)
+
+   * added ESCORT Galleria 600S as supported scanner, is an Astra 600S
+
+   * removed if (config_line[len - 1] == '\n')... while reading config file
+
+   * changed some static unsigned int to static int
+
+   * sane_exit does free devlist 
+
+   * added some options for Powerlook 3000 - still is not completly supported
+
+    
+build-26 -> build-27
+--------------------
+
+   * added Linotype Saphir4 as supported scanner (is a Powerlook III)
+
+   * increased length of vendor, product and version by 1 to make sure the
+     string always ends with '\0'
+
+   * changed default minimum scsi buffer size from 64KB to 32 KB
+
+    
+build-27 -> build-28
+--------------------
+
+   * removed "#include <assert.h>" from umax.c, assert() is never called
+
+   * added UMAX Powerlook 3000 as supported scanner
+
+   * bugfix for 3 pass scanning:
+     reposition_scanner is called in sane_start for 2nd and 3rd pass of 3 pass scan
+
+
+   * bugfix for UTA and DOR geometry offsets:
+     - added inquiry_uta_x_off and inquiry_uta_y_off
+     - geometry ranges are defined from xoff..xoff+dx, yoff..yoff+dx now
+     - check_values also does test for this range
+     - removed dev->maxwidth and dev->maxlength, values are calculated in place now
+
+   * bugfix for DOR:
+     - bottom-right edge keeps in position if possible when DOR is switched on
+
+   * bugfix for Astra 600S and 610S color mode:
+     - scale_y is set to 0.5 if y_resolutions is smaller than optical_res/2
+     this does fix the color scan problems for Astra 610S.
+
+   * added Heidelberg Linoscan 1400 "HDM LS4H1S" as supported scanner
+
+    
+build-28 -> build-29
+--------------------
+
+   * removed bug in umax_set_max_geomentry: now left y keeps in position when possible
+
+   * rgb bind keeps unchanged when scanmode != color (only gray selection active in this case)
+
+   * in function static int umax_give_scanner(Umax_Device *dev) moved umax_reposition_scanner(dev);
+     to the end of the function
+
+   * set calibration area for astra 2200 to full ccd and calibration_bytespp to 2
+
+   * UC840 does wait in reposition_scanner until scan head stops moving
+
+   * set minimum resolution to 5 (was 1 before what could make problems)
+
+
+build-29 -> build-30
+--------------------
+
+  * added EDGE KTX-9600US as supported scanner
+
+  * Changing scansource does not change lower left selection edge
+    if not necessary
+
+
+build-30 -> build-31 (sane-1.0.6)
+---------------------------------
+
+  * sane-umax.man uses @LIBDIR@ and @CONFIGFIR@
+
+  * changed automatic calibration_width_offset marker from -1 to -99999
+
+  * changed
+      dev->calibration_area to 1 and
+      calibration_width_offset to -1 and
+      inquiry_max_calibration_data_lines to 66
+   for UMAX Supervista S-12, LinoHell Office and LinHell JADE
+
+
+build-31 -> build-32 (sane-1.0.7)
+---------------------------------
+
+  * added Nikon AX-110 as supported scanner (Vista S6E)
+
+  * added umax.conf option "gamma-lsb-padded" with options:
+    -1 = auto
+    0 = inactive (msb padded)
+    1 = active (lsb padded)
+    activated for astra 1220s, internal 12 bit gamma correction does work now
+
+
+build-32 -> build-33
+--------------------
+
+  * added USB support for Astra 2200
+    thanks to Frank Zago <fzago@greshamstorage.com>
+    for creating umax-usb.c and suggesting several changes to umax.c
+   
+  * added umax.conf option "connection-type", values: 1=scsi, 2=usb
+
+  * added umax.conf device selection "usb 0x1606 0x0230" (usb vendor_id product_id) 
+
+  * removed #ifdef HAVE_SANEI_IPC parts (SANEI_IPC was never defined)
+
+
+build-33 -> build-34 (sane-backends-1.0.8)
+--------------------
+
+  * removed bug in reader_process / umax_queue_read_image_data:
+    - the integer value that keeps the length information is kept valid
+      as long a queued scsi command is still running now
+    - when the scsi command returns with less data than requested
+      then the routine does handle this correct now
+      (before the routine handled this as if the requested amount of
+       data was returned)
+
+  * backend waits until scanhead has repositioned after scan for UMAX UC1260
+
+  * y_dpi quanitsation is set to 5 (same as x_dpi quantisation - was 1)
+
+
+build-34 -> build-35
+--------------------
+
+  * changed ID string "LS4H1S  " to "LS4H1S " (remove one space)
+
+  * activated lamp control for Astra 1220S V1.5
+
+
+build-35 -> build-36
+--------------------
+
+  * added PowerLook 2100XL as supported scanner
+
+  * settings for PowerLook 2100XL:
+    calibration_width_offest = 52
+    (calibration_bytespp = 2, default)
+    (calibration area = image, default)
+
+  * force_quality_calibration defined for PowerLook 2100XL:
+    (normal calibration mode expects different calibration_width_offsets
+     for PowerLook 2100XL, so it is not possible to calculate the correct
+     number of calibration pixels in normal calibration mode)
+    - disable quality calibration option for frontend
+    - always set quality calibration bit when scanning
+
+  * changes for exposure times:
+    - Inquiry does print correct maximum value now
+    - changed variable type from SANE_FIXED to SANE_INT
+    - exposure times setting for colormode does work correct now
+    - as default there is only one common setting for exposure times in rgb mode,
+      added umax.conf option exposure-time-rgb-bind to enable different settings
+      for each color
+    - added option select-calibration-exposure-time, disabled by default.
+      if disabled then the calibration expousre times are the same as the scan exposure times
+
+  * added UMAX PS-2400X as supported device
+
+  * added batch scanning support:
+    Options:
+    --batch-scan-start:		calibration, scan, scanhead is not moved into home position
+    --batch-scan-loop:		no calibration, scan, scanhead is not moved into home position
+    --batch-scan-end:		no calibration, scan, scanhead is moved into home position
+    --batch-next-tl-y:		defines the y start position of the bext batch scan
+
+    - calibration for batch scan needs a different offset and always is done for full ccd.
+      PowerLook 2100XL: calibration_width_offset_batch = 1052
+      Powerlook III:	calibration_width_offset_batch = 828
+
+
+build-36 -> build-37
+--------------------
+
+  * added Linotype Hell SAPHIR3 as supported scanner
+
+  * if batch-next-tl-y is not set then the backend uses br_y
+
+
+build-37 -> build-38 (sane-backends-1.0.9)
+--------------------
+
+  * added Power Look 2000 as supported scanner
+
+
+build-38 -> build-39 (sane-backends-1.0.10)
+--------------------
+
+  * Power Look 2000: corrected wrong calibration-width-offset from 52 to 22
+    and added calibration-width-offset-batch to 24
+
+  * initialized some global pointers and num_devices in sane_init
+
+  * replaced fork() by thread for OS2
+
+build-39 -> build-40 (sane-backends-1.0.12)
+--------------------
+
+  * UC630: dev->pause_after_reposition = 0
+
+
+build-40 -> build-41 (sane-backends-1.0.13)
+--------------------
+
+  * Astra 2200 SU via USB connection: quality calibration is reset and not selectable
+
+
+build-41 -> build-42
+--------------------
+
+  * replaced fork() routines by sanei_thread routines
+
+  * added call sanei_scsi_req_flush_all(); to do_cancel - in threaded case this is not done when reader_process is cancled
+
+  * in reader_process: marked freed scsi buffers as freed(NULL). This is necessary when reader_process is run as thread!
+
+  * Backend is compatible to Mac OS-X now.
+
+
+build-42 -> build-43
+--------------------
+
+  * DOR mode does test for non negative upper_left_x values, set to 0 if negative
+
+
+build-43 -> build-44
+--------------------
+
+  * the sense_hanlder stores the sense data in buffer[0] when valid sense data is avialable
+
+  * do_calibration:
+	- corrected bug in reading calibration width from scanner (added /bytespp)
+	- when additional request sense is not called then the buffer is not erased,
+	  the sense data from the sense_handler is used
+	- when additional request sense is called then the buffer is erased at first
+
+build-44 -> build-45
+--------------------
+
+   * added defualt options for Linotype Opal2
+   * changed bug in highlight-blue function (using green value)
+