.help revisions Jun88 pkg.dataio
.nf

dataio$lib/mkpkg
    The getdatatype.x and ranges.x files were duplicates of those in the
    XTOOLS library which is linked in dataio.  These files should be the
    versions actually linked in the binary, making XTOOLS unnecessary, but
    I think the intent was to use XTOOLS.  The getdatatype.x is identical
    but there are slight changes in ranges.x.  These files were left in
    place in case there are problems found but the binary should now be
    using the XTOOLS versions.  (12/5/08, MJF)

dataio$import/ipproc.gx
    Fixed a type flag being used to determine byte-swapping. (12/5/08, MJF)

dataio$import/t_import.x
    Fixed a type declaration (1/21/08, MJF)

dataio$export/exraster.gx
    Fixed a bug in computing the number of output pixels (1/5/04, MJF)

dataio$import/ipproc.gx
    An operand point was possibly being freed twice, once in the ip_wrline()
    procedure and again in the evvfree() call when processing completed.  
    This could cause a segfault on some system (9/27/02, MJF)

dataio$export/exraster.gx
dataio$export/bltins/exppm.x
    There was a bug in the generation of PPM files when using images with
    and odd number of columns causing the line to be too long by one byte.
    The output image will now truncate the last column to avoid this since
    we cannot write byte data. (8/9/02, MJF)

dataio$export/export.h
    Changed the zscale sampling parameters to use more points spread out
    over more of the image.  The old values would sometimes find an
    innappropriate z1/z2 range causing problems when doing many images in
    batch mode.  (3/20/02, MJF)

dataio$fits/t_rfits.x
dataio$imtext/t_rtextimage.x
    Changed the clgetc calls to to clgstr calls for the datatype parameter
    in rfits and the otype parameter in rtextimage. This change is required
    to avoid an "ERROR: Parameter not a legal character constant (parname)"
    error introduced by recent changes to the CL. Basically "" is no longer
    a legal argument for clgetc. (6/15/01 LED)

dataio$fits/fits_rheader.x
    Fixed a bug in the MEF file reading error recovery code that can cause
    a segvio due to a too many open file descriptors condition. (5/1/01 LED)

dataio$export/bltins/expgm.x
dataio$export/bltins/exppm.x
    Fixed a bug in writing the header for these formats on PCIX.  (6/23/00 MJF)

dataio$export.par
    Made the 'format' parameter automatic mode (5/16/00 MJF)

dataio$export/expreproc.x
    Modified so that the 'overlay' colors are not scaled.  (5/16/00 MJF)

dataio$reblock/t_reblock.x
dataio$reblock/reblock_file.x
    Fixed a bug in the output block writing code caused by a failure to
    check the status return of the awaitb call. At the same time worked
    around a longstanding  problem in tape to tape copies caused by
    trying to open and close a magtape file for writing when there is
    no data and fixed an initialization bug in the record trimming and
    padding code that has been there a long time. (2/9/00 LED)

dataio$import/ipdb.gx
    Fixed a string overflow bug causing segvios on PCs (12/13/99 MJF)

dataio$import.par
dataio$export.par
    Changed query param modes to auto to avoid prompt from epar :go 
    command. (11/4/99 MJF)

dataio$export/bltins/exeps.x
    Fixed an array overrun when writing EPS trailer comments. (10/25/99 MJF)

dataio$export/mkpkg
dataio$export/*/mkpkg
dataio$import/mkpkg
dataio$import/*/mkpkg
    Fixed missing/extra file dependencies  (9/20/99, MJF)

dataio$export/exrgb8.x
    Fixed a bug causing the cmap() function to write a garbage line at the
    top or bottom of the output image. (8/20/99 MJF)

dataio$export/exzscale.x
dataio$import/t_import.x
dataio$import/bltins/ipgif.x
    Cleaned up some missing sfree() calls.  (7/28/99 MJF)

dataio$fits/wfits.h
dataio$fits/fits_cards.x
dataio$fits/fits_params.x
    Modified wfits to write the DATE keyword value in the new format (including
    a time field) and in units of GMT. (5/8/99 LED)

dataio$export/bltins/exhdr.x
    Changed the binary file header to use a 4-digit year (5/5/99, er..1999 MJF))

dataio$export/bltins/exras.x
dataio$export/bltins/exxwd.x
    Fixed a bug in which rasterfiles and XWD file aren't swapped on LSB
    machines when requested by the user. (12/14/98 MJF)

dataio$import/t_import.x
dataio$import/ipproc.gx
    Fixed a bug preventing 1-D data from being converted.  (12/10/98 MJF))

dataio$import/t_import.x
    Modified to initialize the 'use_cmap' flag to on by default so 8-bit
    colormap images will be converted correctly.  (7/14/98 MJF)

dataio$fits/fits_cards.x
    Added a check for pre-existing IRAFNAME keywords to the wfits task
    (6/18/98 LED)

dataio$export/exzscale.x
dataio$export/t_export.x
dataio$export/exobands.gx
dataio$export/bltins/exeps.x
dataio$export/bltins/exgif.x
dataio$export/bltins/exrgb.x
dataio$export/bltins/exxwd.x
    Removed some unused variable and fixed type clashes (3/28/98  MJF)

dataio$export/expreproc.x
    Fixed a bug where the setcmap function was finding the colormap name
    incorrectly when '@' params were used in a zscale function (2/2/98  MJF)

dataio$import/t_import.x
dataio$import/ipobands.gx
    Fixed a bug in which use of the red()/green()/blue() functions caused
    the size of the output image to be computed incorrectly. (12/12/97 MJF)

dataio$fits/fits_read.x
    Fixed a bug in the header listing code which resulted in rfits reading
    through the whole image after listing the data when EXTENSION = N.
    (11/3/97 LED)

dataio$fits/fits_read.x
    Fixed a bug in the global header handling code that could result a
    the global image header being left in the tmp$ directory.
    (8/6/97 LED)

dataio$wfits.par
dataio$fits/t_wfits.x
dataio$doc/wfits.hlp
    Added a new parameter call fextn whose default value is "fits" to the
    wfits task. A ".fextn" suffix is appended to the output disk file names
    to ensure that they are compatable with the fits kernel.
    (6/20/97 LED)

dataio$rfits.par
dataio$wfits.par
dataio$doc/rfits.hlp
dataio$doc/wfits.hlp
dataio$fits/fits_cards.x
dataio$fits/fits_files.x
dataio$fits/fits_params.x
dataio$fits/fits_read.x
dataio$fits/fits_rheader.x
dataio$fits/fits_rimage.x
dataio$fits/fits_rpixels.x
dataio$fits/fits_wheader.x
dataio$fits/fits_wimage.x
dataio$fits/fits_wpixels.x
dataio$fits/fits_write.x
dataio$fits/mkpkg
dataio$fits/rfits.com
dataio$fits/rfits.h
dataio$fits/t_rfits.x
dataio$fits/t_wfits.x
dataio$fits/wfits.com
dataio$fits/wfits.h
    Installed new versions of rfits and wfits. The new rfits and wfits
    include support for: 1) reading and writing multi-extension fits files,
    2) reading and writing global header, 3) reading and writing ushort
    images by default when appropriate.
    (6/9/97 LED)


dataio$doc/export.hlp
dataio$export/excmap.x
dataio$export/cmaps.inc
    Added the 'overlay' cmap as a builtin cmap. (6/6/97)

dataio$export/expreproc.x
    Removed a call to scale the colormaps when using the default values.
    Cmaps are now only scaled when a brightness/contrast value is set in
    the setcmap() function.  (6/6/97  MJF)

dataio$export/bltins/exgif.x
    Fixed a small error in the output of GIF files causing some display
    programs to complain.  GIF images which would now be an odd number of
    bytes have an extra trailing ';' delimiter.  This should be harmless
    as all processing is supposed to stop once that char is found. (6/6/97 MJF)

dataio$mkpkg
dataio$dataio.cl
dataio$dataio.hd
dataio$dataio.men
dataio$x_dataio.x
dataio$import/			+
dataio$import.par		+
dataio$export/			+
dataio$export.par		+
dataio$doc/import.hlp		+
dataio$doc/export.hlp		+
    Installed the IMPORT/EXPORT task for general use.  The images database
    used by the IMPORT task is currently defined to be dataio$import/images.dat.
    (3/31/97 MJF)


dataio$fits/fits_rheader.x
    Explictly set SIMPLE(fits) to YES the first time this card is encountered
    so that duplicate fits SIMPLE cards are properly filtered out. Duplicate
    SIMPLE cards are illegal fits so this should never happen but ...
    (6/28/96 Davis).

dataio$fits/fits_cards.x
    Added some wfits code to filter any "END     " keywords out of the image
    header userarea.
    (3/17/95 Davis)

dataio$fits/t_wfits.x
    Wfits was using the name of the @file instead of the first file  as the
    root output fits file name if the number of output files was 1.
    (1/18/95 Davis)

dataio$fits/fits_wheader.x
    The autoscaling code was failing  in the case bitpix=16 and pixtype=ushort.
    Bscale and bzero values were being set to 1.0 and 0.0 respectively,
    resulting in truncation of data values. The code has been modified to
    set bscale and bzero to 1.0 and 32768 instead.(10/18/94 Davis)

dataio$imtext/t_rtextimage.x
dataio$imtext/t_wtextimage.x
dataio$imtext/rt_cvtpix.x
dataio$rtextimage.par
dataio$wtextimage.par
dataio$doc/rtextimage.hlp
dataio$doc/wtextimage.hlp
    A parameter "pixels" was added to select whether to read or write
    the pixel values.  This is complementary to rfits/wfits and allows
    use of these tasks to store and restore image headers.  (10/22/93, Valdes)

dataio$reblock/t_reblock.x: Davis, Jan 20, 1993
    Added support for multiple disk file input and output to the reblock
    task.

dataio$fits/fits_read: Davis, Apr 27, 1992
    Modified rfits to use the fe parameter to skip to EOF for devices
    e.g. cartridge tapes which seem to be having problems with file
    skips.

dataio$fits/t_rfits.x: Davis, Mar 25, 1992
    Rfits was using the name of the @file instead of the first file  as the
    root output image name if the number of output files was 1.

dataio$fits/fits_read: Davis, Feb 27, 1992
    Changed the interpreation of the fe parameter as read from dev$tapecap
    from MB to KB.

dataio$fits/fits_write: Davis, Feb 18, 1992
    Replaced a call to imgimage with one to imgcluster to extract the root
    image name minus cluster and section.

dataio$fits/t_rfits.x: Davis, Feb 18, 1992
dataio$fits/t_wfits.x: Davis, Feb 18, 1992
    Changed the maximum sequence number that can be appended to an output
    root image of fits file name from 999 to 9999.

dataio$fits/rfits.com: Davis, Feb 18, 1992
dataio$fits/t_rfits.x: Davis, Feb 18, 1992
dataio$fits/fits_read.x: Davis, Feb 18, 1992
    Implemented a scan mode in rfits so that devices which have a slow
    single-file file skip function (e.g. dat drives) can be used more
    efficiently with the rfits make_image=no option.

dataio$fits/fits_params.x: Davis, Feb 17, 1992
    Modified wfits so that string parameters that are 1) written explictly
    by wfits, and 2) <= 20 characters long including quotes, will have the /
    in column 33 instead of 2 spaces past the end of the string. The
    affected keywords are OBJECT, ORIGIN, DATE, IRAFNAME, IRAF-BPX, and
    IRAFTYPE.

dataio$fits/fits_wheader.x: Davis, Feb 17, 1992
    Modified the short_header=yes option in wfits so that the image pixel
    type, fits bitpix, and the scaling parameters are printed on the
    standard output.

dataio$fits/fits_rimage.x: Davis, Feb 17, 1992
dataio$fits/fits_wimage.x: Davis, Feb 17, 1992
    Modified rfits so that the ieee +/-NaNs, and +/-Infs are correctly
    mapped to a user specified native floating point number. Underflow values
    are automatically converted to 0.0. A warning message is printed on
    the terminal if any bad pixels were replaced.  A warning message is
    also printed if valid floating point pixel values > MAX_REAL or <
    -MAX_REAL were detected. Imreplace can be used to replace these
    explicitly.

dataio$fits/fits_rheader.x: Davis, Feb 14, 1992
    Modified rfits to replace control characters decimal 0 (00X) to
    31 (1FX) and decimal 127 (7FX) with the blank character. The
    new fits standard now explicitly defines these illegal in fits
    headers.

dataio$wfits.par: From, Davis, Feb 13, 1992
dataio$doc/wfits.hlp: From, Davis, Feb 13, 1992
dataio$fits/wfits.h: From, Davis, Feb 13, 1992
dataio$fits/t_wfits.x: From, Davis, Feb 13, 1992
dataio$fits/fits_write.x: From, Davis, Feb 13, 1992
dataio$fits/fits_wheader.x: From, Davis, Feb 13, 1992
dataio$fits/fits_wimage.x: From, Davis, Feb 13, 1992
dataio$fits/fits_wpixels.x: From, Davis, Feb 13, 1992
    1. Modified wfits to fetch the default fits blocking factor for a device
    from the dev$tapecap file. The user can still overrride this value
    (which is usually set to 10) for variable blocked devices, but is no
    longer required to know or set the block size for fixed block devices
    like cartridge tapes.

dataio$mtexamine/t_mtexamine.x: From, Davis, Jan 6, 1992
dataio$cardimage/t_rcardimage.x: From, Davis, Jan 6, 1992
dataio$cardimage/t_wcardimage.x: From, Davis, Jan 6, 1992
dataio$reblock/t_reblock.x: From, Davis, Jan 6, 1992
dataio$fits/t_rfits.x: From, Davis, Jan 6, 1992
dataio$fits/t_wfits.x: From, Davis, Jan 6, 1992
    1. Modified the mtexamine, rcardimage, wcardimage, reblock, rfits, and
    wfits tasks to accept the new magtape file name syntax.

dataio$reblock/t_reblock.x: From, Davis, Dec 11, 1991
    1. Modified reblock so that character constants like '\n' can be
    used as record padding characters.

dataio$t2d/mkpkg: From Davis, Dec 3, 1991
    1. Removed the printf.h file dependency from the mkpkg.

dataio$imtext/mkpkg: From Davis, Dec 3, 1991
    1. Added missing files dependencies for the files rt_rheader.x
    (imio.h) and wti_wheader.x (imio.h).
    
dataio$fits/mkpkg: From Davis, Dec 3, 1991
dataio$fits/fits_read.x
    1. Added missing files dependencies for the files fits_rheader.x
    (ctype.h) and fits_wimage.x (error.h).
    2. Removed unused "include <imset.h>" statement from fits_read.x

dataio$cardimage/mkpkg: From Davis, Dec 3, 1991
    1. The entries for t_rcardimage.x and t_wcardimage.x were missing
    several file dependencies.

dataio$fits/fits_cards.x: From Davis, Oct 15, 1991
    1. Changed the name of the IRAF-B/P keyword to IRAF-BPX to conform to the
    new FITS standard.

dataio$fits/fits_rpixels.x: From Davis, Oct 9, 1991
    1. The rfits task has been modified to permit a short last record, i.e.
    a last record that has not been padded out to 2880 bytes, without
    generating any warning messages.

dataio$fits/wfits.h: From Davis, Jun 11, 1991
dataio$fits/t_wfits.x: From Davis, Jun 11, 1991
dataio$fits/fits_write.x: From Davis, Jun 11, 1991
dataio$fits/fits_wheader.x: From Davis, Jun 11, 1991
    1. The wfits task has been modified to write IEEE format FITS files
    (fits bitpix = -32 and -64), instead of scaled integers if the input
    image pixel type is real or double respectively and if the wfits parameter
    bitpix is left at the default. If the user overrides the default and
    elects to scale the data, a warning message with an estimate of the
    precision loss is provided.

dataio$fits/fits_write.x: From Davis, Jun 10, 1991
dataio$fits/fits_read.x: From Davis, Jun 10, 1991
    1. Modified the fits writer and reader so that the IRAFNAME parameter
    can deal with image sections. The fits writer will now record image
    sections in the IRAFNAME parameter instead of inserting a blank.
    The directory specification is still stripped. The fits reader will
    now strip off any section notation before attempting to rename
    the output image.

dataio$fits/fits_rimage.x: From Davis, Jan 17, 1991
dataio$fits/fits_wimage.x: From Davis, Jan 17, 1991
    1. Modified the scaling routines in rfits and wfits to minimize
    the precision lost when converting from real pixels to fits integers
    and vice versa. 

dataio$cardimage/t_rcardimage.x: From Davis, Jan 3, 1991
dataio$doc/rcardimage.hlp: From Davis, Jan 3, 1991
    1. Modified rcardimage so that the error message encountered
    when an odd-blocked rcardimage tape is encountered is less obscure.
    2. Modified the rcardimage help page to include an example of how to
    reformat and odd-blocked cardimage tape with reblock.

dataio$fits/t_rfits.x: From Davis, Dec 6, 1990
    1. Modified rfits so that it will supply a temporary root output file
    name if old_irafname="yes" and quit with a clear error message if
    old_irafname="no", in the case where the user sets the output file
    to the null string "".

dataio$fits/fits_rheader.x: From Davis, Sept 11, 1990
    1. Changed rfits so that history cards know go into the user area
    instead of the history area where they get truncated.

dataio$fits/wfits.h: From Davis, August 15, 1990
dataio$fits/t_wfits.x: From Davis, August 15, 1990
dataio$fits/fits_write.x: From Davis, August 15, 1990
dataio$fits/fits_wimage.x: From Davis, August 15, 1990
dataio$fits/fits_wpixels.x: From Davis, August 15, 1990
dataio$fits/fits_rpixels.x: From Davis, August 15, 1990
    1. Wfits will now permit FITS blocking factors up to and including
    22 although a warning message will be issued if a blocking factor
    > 10 is requested.

    2. Wfits occasionally crashed with a segmentation violation if a
    non-standard fits blocking factor was selected. This error was
    triggered if the unused portion of the output block to be blank
    filled was greater than 2880 bytes. This bug has been fixed in
    2.10

    3. Rfits was not reading FITS data with a block size < 2880 on
    the 9-track drives correctly. This bug has been fixed in 2.10.

    4. Wfits.rfits now sets the length of the user area to the maximum of
    the default of 28800 chars and the value of the environment variable
    "min_lenuserarea".

dataio$fits/fits_wheader.x: From Davis, July 3, 1990
dataio$fits/fits_cards.x: From Davis, July 3, 1990
dataio$fits/fits_rheader.x: From Davis, July 3, 1990
    Fixed a problem in the way bscale and bzero were computed that was
    causing floating point errors for some double precision images,
    basically because the precision was worse than I thought.
    The really problem is that the min and max of a double precision
    image are  stored as reals in the image header do they do not
    quite correspond to what is in the image. The solution was to extend
    the values of the min and max to slightly lower and higher values
    respectively. These problems will go away when ieee becomes more
    accepted. (See messages below for history of this problem)

    Wfits now checks for the presence of the FITS keywords SIMPLE, BITPIX,
    NAXIS and NAXISn in the user area and filters them out before writing
    the FITS header.

    Rfits will now ignore FITS keywords that are duplicates of SIMPLE,
    BITPIX, NAXIS and NAXISn. A warning message is issued if any of
    these keywords are duplicated.

dataio$fits/fits_wheader.x: From Davis, April 21
dataio$fits/fits_write.x: From Davis, April 21
dataio$fits/fits_read.x: From Davis, April 21
    The original scaling algorithm was restored due to problems encountered
    with the new one. This will be looked into more fully in version 2.10.
    These means that a problem with double precision images may remain.

    Since I had to make the above change at the last minute I added some
    code to flush the STDOUT  after in input and output file names
    are computed and written to STDOUT. This avoids a problem with
    output not being flushed when an error condition occurs and output
    is being redirected causing confusion for the user who may not be able
    to tell where the error occured in that case.

dataio$fits/fits_wheader.x: From Davis, Mar 24, 1990
    Fixed a problem with the scaling routines in wfits. Images with a
    minimum which was negative and distant from the majority of the
    data values could cause a problem in the scaling.

dataio$fits/fits_rimage.x: From Davis, Mar 10, 1990
    Recoded the routine that computes the mins and maxs of an image
    slightly to remove a problem with the STF kernel. The mins and maxs
    of the image were being reset to 0 when the first data was written
    over-riding the program initialization of MAX_REAL and -MAX_REAL
    and defeating the minimum calculation for all positive data.

dataio$fits/fits_wheader.x: From Davis, Mar 9, 1990
    Fixed a problem in the way bscale and bzero were computed that was
    causing floating point errors for some double precision images,
    basically because the precision was worse than I thought.
    The really problem is that the min and max of a double precision
    image are  stored as reals in the image header do they do not
    quite correspond to what is in the image. The solution was to extend
    the values of the min and max to slightly lower and higher values
    respectively. These problems will go away when ieee becomes more
    accepted.

dataio$fits/fits_cards: From Davis, Jan 20, 1990
    Added a filter to remove duplicate IRAF-MIN, IRAF-MAX, IRAFTYPE,
    and IRAF-B/P keywords from the user area. 

dataio$fits: From Davis, Jan 19, 1990
    1. Support was added for the IEEE floating point format to both
    the rfits and wfits tasks. Rfits now recognizes -32 and -64 to
    be legal values of bitpix representing respectively real and
    double IEEE floating point format. Values of bscale and bzero
    are applied if present in the header. By default wfits still
    writes integer format FITS tapes with autoscaling. However if
    the users selects a bitpix of -32 or -64 the appropriate floating
    point format is written. In this case scaling is disabled.

    2. Rfits was modified to take a list of output images names or
    an output image root name.

    3. Wfits was modified to take a list of output fits file names
    or an output fits file root name.

dataio$fits/fits_wheader.x: From Davis, Nov 20, 1989
    1. Fixed wfits so it would write out type "ushort" images correctly
    with the default parameters. The chosen bitpix is 32 instead of
    16.

dataio$fits/t_rfits.x,t_wfits.x: From Davis, May 29, 1989
    1. Changed both these tasks so the STDOUT is only flushed on a newline
    if has not been redirected. Changed the remaining eprintf statements
    to printf statements.

dataio$fits/fits_rheader.x: From Davis, May 9, 1989
    1. Fixed abug in the code which decodes hms format numbers. The
    problem was caused by leading blanks and the fact that the ctoi
    routine does not recognize the plus character.

dataio$fits/t_wfits.x: From Davis, Mar 31, 1989
dataio$fits/wft_wimage.x: From Davis, Mar 31, 1989
dataio$fits/wft_wpixels.x: From Davis, Mar 31, 1989
    1. Changed wfits so that a warning message is printed if the fits
    long blocks option is used. The previous version only warned the
    user if an illegal fits block size was used.
    2. Changed wfits so that the record structure written is printed
    out on the standard output if short_header = yes as well as if
    long_header = yes.
    3. Changed the error trapping code so that the number of records
    actually written is printed out when wfits terminates prematurely
    with an error condition.


dataio$fits/t_wfits.x: From Davis, Mar 14, 1989
dataio$fits/fits_rpixels: From Davis, Mar 14, 1989
    1. Changed wfits so that warning messages are printed if the user
    overrides the default value of bitpix or turns of autoscaling.
    These messages will be printed to the standard output along with
    the output log.
    2. I have fixed a problem in the error checking code in rfits.
    Too many reads were being done after an error recovery resulting
    in the data in the output image being skewed. This needs to be
    rechecked on DRACO where the original error recovery was done.

dataio$reblock: From Davis, Jan 27, 1989
    1. Fixed a problem in reblock for tape to tape copies. The copyn parameter
    was being ignored if no reblocking was occurring. This problem has been
    fixed. A minor problem with the record counter was also fixed.

dataio$fits: From Davis, Apr 14, 1988
    1. Added an option in wfits to write a non-standard physical block size
    of blocking_factor > 10. This permits users with restricted block
    size devices to read and write fits tapes.

    2. Changed the cl file name template commands inside wfits to images
    name template commands.

dataio$fits: From Davis, Mar23, 1988
    1. Fixed a bug in the fits string parameter trimming routine in which
    the newline was being overwritten if the string was exactly 80
    characters long.

dataio$rfits: From Davis, December 11, 1987
    1. Fixed a small bug in the rfits disk handling code. If a user
    successfully read a fits disk file, for example fitsdat, and then
    tried to read a list of files using a template which did not match
    any of the disk files, rfits would try to reread fitsdat. Rfits was
    not handling the 0 length disk file list condition correctly.

dataio$rfits: From Davis, December 3, 1987
    1. Rfits now checks for valid bscale and bzero values. If it cannot
    decode bscale or bzero it sets them to 1.0 and 0.0 respectively.
    2. Rfits and wfits no longer flushes STDOUT on a newline if the output
    has been redirected to a file. This greatly improves the speed of rfits
    and wfits when the long_header parameter is set to yes especially for
    VMS systems.

dataio$rfits: From Davis, September 3, 1987
    Rfits will now print out the ol irafname if short_header = yes, make_image
    = no and old_name = yes. This makes it easier for users to list their
    IRAF fits tapes.
    
dataio$reblock/t_reblock.x: From Davis, August 12, 1987
    The offset parameter in reblock was not being queried for by
    the code.
	
dataio$imtext/rt_rheader.x,wti_wheader.x: From SJacoby, June 10, 1987:
    Tasks RTEXTIMAGE and WTEXTIMAGE no longer limit the image user area
    being written to or read from to 2880 chars.  The size of the
    user area to be created or accessed by these tasks is controlled 
    by the IRAF environment variable `min_lenuserarea', and is not
    limited by the code.

dataio$fits/fits_rheader.x: From Davis, June 4, 1987:
    The code for reading fits cards into the user area has been changed.
    Rfits opens a new image with the default min_lenuserarea and
    reads cards into it  until it is filled. If the user area is
    completely filled rfits issues a warning message along with the
    number of fits parameters it was not able to completely store.

dataio$fits/fits_write.x,fits_read.x,fits_rpixels.x: From Davis, May 15, 1987:
    1. I changed the error checking code so that it would work correctly
    with the fits long blocks option. Wfits now does a call to fstat
    to find out the number of bytes in the last read and uses this number
    to validate the buffer if a read error occurs. There is no way to
    recover from a read error in the tape record containing the header info.
    2. I added a check for the maximum permitted buffer size in wfits.
    The program will abort if the device cannot suuport the length of
    the output record requested.

dataio$mtexamine/t_mtexamine.x: From Davis, May 15, 1987:
    Error checking code has been added to MTEXAMINE. The task will now
    print out a warning message for each bad record encountered and
    continue reading the file instead of skipping to the next file.
    The correct record count is preserved.

dataio$imtext/wti_wheader.x, dataio$imtext/rt_rheader.x:  Hammond, Mar 24, l987.
    The FITS format header written by task WTEXTIMAGE has been changed.  It
    no longer contains the keywords SIMPLE=T and NAXIS=0.  The output of
    WTEXTIMAGE is a simple text file which makes no attempt to conform to
    FITS standards.  Task RTEXTIMAGE was modified to read both the old and 
    new format headers.

dataio$t2d/t_t2d.x: From Lytle, Mar 20, 1987:
    1. T2D now deletes the zero length file left over when the program
    encounters the end-of-tape and opens and closes an empty file.
    2. I also changed the verbose output format somewhat to make it
    more logical.

dataio$t_wcardimage: From Davis, Mar 19, 1987:
    1. WCARDIMAGE now checks that the input files are not binary
    files before trying to write them to tape.

dataio$fits: From Davis, Mar 19, 1987:
    1. More extensive error checking has been added to the FITS code.
    Rfits attempts to recover from a read error in the data matrix.
    Instead of terminating with a partially written image RFITS will
    try to skip over the bad data. The resulting output image will
    have the correct number of rows and columns but may contain one
    or more records of bad data. The results of the error checking
    may be tape drive dependent.
    2. RFITS now prints a warning message if no pixel file is
    written (NAXIS = 0).
    3. RFITS now checks the first 6 characters of the first header record
    to see if they are equal to SIMPLE.

dataio$fits_wheader.x: From Davis, Jan 28, 1987:
    1. The scaling routine for determining bscale and bzero
    introduced by Skip for the MV10000 was found to have problems
    with some low dynamic range data. The symptom was that
    the min value of an image restored from a FITS data would
    have larger than expected roundoff errors. I have changed it back
    to my original scaling algorithm.

dataio$cardimage/: From Davis, Jan 20, 1987:
    1. Rcardimage has been modified to accept a list of disk files as
    input as well as a list of tape files.

dataio$fits/: From Davis, Jan 20, 1987
    1. The FITS longblocks option has been added to the FITS readers and
    writers. RFITS will read long-blocked FITS data. The redundant
    len_record parameter has been removed. IRAF mtio handles the tape
    record buffering transparently.
    2. The len_record parameter has been removed from WFITS and
    replaced with a blocking parameter factor, which specifies the number
    of 2880 byte FITS records can be written in a single tape block.
    The maximum legal FITS blocking factor is 10.
    3. RFITS has been modified to accept a list of disk files as well as
    a list of tape files. This should facilitate file transfers over
    the ethernet.

dataio$reblock/reb_reblock_file.x: From Davis, Dec 12, 1986
    1. A bug in the seek option on disk binary files has been fixed.
    This bug would cause the number of bytes read to be incorrectly
    computed.

dataio$lib/addcards.x: From Hammond, Oct 27, 1986
    Header cards containing real values are now written with a %g rather 
    than %f format.  This change affects task wtextimage.

dataio$fits/fits_wheader.x: From Davis, Oct2
    1. The way the scaling routine wft_set_scale computes the data range
    has been changed. Instead of adjusting maxdata and mindata individually
    for machine precision; the data range is first computed and then adjusted
    for the machine precision. This change was made in response to precision
    problems encountered on the MV10000.

dataio$fits/wfits.h: From Davis, Sep12, 1986
    1. The integer constants BYTE_MIN, BYTE_MAX, BYTE_BLANK etc have been
    changed to type double to avoid compiler generated integer overflows.
    The data type of TAPEMIN and TAPEMAX in the WFITS structure has also
    been changed to double. This change was made in response to compiler
    errors encountered on the MV10000 and should be transparent to the users.

dataio$imtext/rt_cvtpix.x: From Hammond, Sep2, 1986
    A typographical error was corrected in a call to patmake.  The
    pattern "[DdEd]" has been replaced with the correct pattern "[DdEe]".
    This means numbers written with a lower case 'e' in the exponent field 
    are recognized as floating point numbers when read from the text file.  
    This procedure is called by task RTEXTIMAGE.

dataio$mtexamine/t_mtexamine.x: From Davis, Aug20, 1986
    1. MTEXAMINE on the SUN  was outputting an array of zeroes when asked to
    dump records with output_type = c. The problem was that a long integer
    was being passed to the routine ctocc instead of a char. This error
    was not being detected on the VAXES.

dataio$fits/fits_params.x: From Davis, Aug20, 1986
    1. The boolean parameter param in routine wft_encodeb was changed to type
    integer. Wft_encodeb was being called with param = YES which caused
    portability problems on the MV10000.

From Davis July 16, 1986:

1. RFITS has been modified so that imbedded blanks in the UT, ZD, ST, RA and
DEC keywords are replaced by zeros. For example the mountain fits writers
produce hms numbers of the form 20: 6: 3. RFITS will convert this to 20:06:03.

-----------------------------------------------------------------------------

From Davis June 13, 1986:

1. TXTBIN and BINTXT have been modified so that the file number is
appended to the output file name. Previous versions appended the extensions
".txt" and ".bin" to the input file name.

------------------------------------------------------------------------------

From Davis June 12, 1986:

1. WCARDIMAGE and WFITS now append a file number to the output file name
if multiple disk files are being written. In the old writers a suffix
was appended to the input file name (.fit for WFITS and .crd for
WCARDIMAGE) and the output images were being copied to the input
directory.

------------------------------------------------------------------------------

From Davis June 8, 1986:

1. WFITS has been modified to store only the root of the image name. All
pathname information has been stripped off.

------------------------------------------------------------------------------

From Davis May 28, 1986:

1. The output of WFITS has been changed to print the file number, input file
name, output file name, title and dimensions if long_header = no and
short_header = yes.

-------------------------------------------------------------------------------

From Davis May 22, 1986:

1. RFITS now writes 80 character records into the user area instead of
trimming trailing whitespace from each record. This was changed to facilitate
the image database interface.  This change was also made to WTEXTIMAGE.

-------------------------------------------------------------------------------

From Davis May 21, 1986:

A bug in the record trimming code of REBLOCK has been fixed. REBLOCK was
computing the offset in the input block of data incorrectly.

--------------------------------------------------------------------------------

From Davis May 13, 1986:

A bug in the error checking code in MTEXAMINE has been fixed. If there is an
error on mtopen the program will abort instead of trying to open the next file.

--------------------------------------------------------------------------------

From Davis May 1, 1986:

A problem with the autoscaling option in WFITS has been fixed. In order
to avoid wrap around problems WFITS now assumes that the number of digits
of machine precision is 1 less than the number in mach.h. In the case of
the Vax the number is actually 6.? not 7 as quoted.

-----------------------------------------------------------------------------

From Davis Apr 17, 1986:

Changed boolean == false constructs in files t_wcardimage.x and t_reblock.x
to ! boolean.

----------------------------------------------------------------------------

From Davis Apr 4, 1986:

The format of the RCAMERA DATE-OBS parameter has been changed form
dd-mm-yyyy to dd/mm/yyyy to bring it into conformity with FITS standard.

---------------------------------------------------------------------------

From Hammond Mar 25, 1986:

Task RTEXTIMAGE has been fixed so it properly skips over non standard fits
headers, the number of lines being specified by the parameter nskip.

___________________________________________________________________________

From Davis Mar 9, 1986:

The order of the REBLOCK parameters  outfiles and file_list has been switched
in order to preserve the correct command line sequence

----------------------------------------------------------------------------

From Davis Mar 3, 1986:

The error checking in WFITS has been corrected so that WFITS terminates if
it encounters a file write error instead of continuing to the next file
as done previously.

----------------------------------------------------------------------------

From Davis Feb 19, 1986:

1. Rfits and rpds have been fixed so that attempting to delete the last
empty image does not generate a cannot delete protected file message.

----------------------------------------------------------------------------

From Davis Feb 3, 1986:

1. A mysterious bug in which the date of observation card would sometimes
not appear in the header has been fixed. A newline was missing from the
proceeding header card.

----------------------------------------------------------------------------

From Davis Jan 16, 1986:

1. Wfits no longer needs write permission to work. However as a consequence
wfits no longer updates the image min and max.

2. The scaling routines in rfits and wfits fits have been replaced by
appropriate vector operators.

3. The coordinate transformation parameters are now stored in the user
area and are available to hedit, imgets etc.

4. Scaled data is now read into real images regardless of the value of
bitpix. 

-----------------------------------------------------------------------------

From Davis Jan. 5, 1986:

1. Rfits, rpds and rcamera now open dev$null instead of a temporary disk
file for option make_image = no. This eliminates a lot od disk access overhead
and should speed up these readers considerably.

2. The default parameter options are now long_header=no and short_header=yes.
Setting the long_header parameter to yes will over-ride the short header
parameter.

---------------------------------------------------------------------------

From Davis Dec. 3, 1985:

1. Rcamera will now print and store the header parameters ccdpicno and airmass
if defined.

2. A bug in the fringe scaling parameter calculation in rcamera was fixed.
Currently the mountain programs store this number in floating point format.
This will be changed in future necessitating a corresponding change in
rcamera.

-----------------------------------------------------------------------
From Valdes Oct. 10, 1985:

1.  Defined widstape from ONEDSPEC package in the DATAIO package.  The
source and executable, however, still reside in ONEDSPEC (x_onedutil.e).
Widstape and widsout should be combined and the source put in DATAIO
at some point.
.endhelp