Category : Printer + Display Graphics
Archive   : DTA18G-2.ZIP

Output of file : WHATSNEW.DTA contained in archive : DTA18G-2.ZIP
What's new in DTA:

Rel 1.8g (02/26/92)
o Fixed some bugs in reading archive files... if COMMAND.COM
wasn't in the root directory of the current drive, DTA
would be unable to execute the archiver.
Also, it wasn't allocating enough memory in the real-mode
version to run an archiver, regardless of where
o Added comments to @files... just prefix comments with a ";".
o DTA now ignores blank lines in @files.
o Fixed assorted bugs.
o Added grayscale TIF output
o Added /I option, which works like /K (skip), except instead
of specifying how many frames to skip, specify the total
number of frames you want in your flic. So if you have
50 files... X000.TGA, X001.TGA, on up to X049.TGA, and
you type:
dta x*.tga /i10
then DTA will create a 10-frame flic from X000, X004, X009,
Rel 1.8f (12/30/92)
o Fixed some more bugs.
o Added support for 8-bit grayscale TGA files (type 3).
(To create one, use /B8 /FT).
o Added picture rescaling. The /SC option, which only works
when you're creating flics, rescales an image so it's the same
size as the screen resolution.
The /SW option rescales an image so it's the same width as the
screen resolution (or, if you enter a number, like /SW100,
to the width that you specify... if you use /SW for an output
format other than a flic, then a specified width is required)
The /SD option rescales an image so it's the same depth as the
screen resolution (or, if you enter a number, like /SD100,
to the depth that you specify... if you use /SD with an output
format other than a flic, then a specified width is required)
Rel 1.8e (12/17/92)
o For some reason the /NM option didn't work anymore. Fixed.
o Fixed a bug in FLC compression... it caused the wrong pixels
to be updated when there was a change in a pixel row after a
run of more than 254 unchanged pixels.
o Fixed a bug that caused a page-fault (runtime error 216)
when creating GIF or TGA files in the protected mode
o There used to be a limit on the width of pictures that DTA
could process... if you made it read a picture wider than
1280 pixels, you'd either get a spectacular crash requiring
a reboot (real mode) or an error message (protected mode).
Now it should be able to process pictures with a width up
to 16K pixels or so without crashing.
Rel 1.8d (11/30/92)
o Fixed one bug, which prevented DTA from creating 1-frame
flics correctly.
o Including two versions of DTA... DTA.EXE (real mode) and
DTAX.EXE (protected mode). The memory management stuff
that I included in 1.8b and 1.8c is now separated into
another archive, DTAMEM.ZIP.
The real-mode version runs faster than the protected-mode
version, but only gives you access to 640K of memory.
Rel 1.8c (11/27/92)
o Whoops, introduced some new bugs in 1.8b. Got ZIP, LZH, and
ARJ working fine, but it wouldn't read pictures outside of
archives anymore.
Rel 1.8b (11/27/92)
o Fixed a bug that caused a runtime error 204 whenever a frame
couldn't be compressed and had to be stored raw.
o Flics of resolutions other than 320x200 are now saved with the
extension 'FLC' instead of 'FLI'... to be consistent with
conventions established by Autodesk's Animator Pro.
o This is the first version of DTA to be compiled in 286 protected
mode. This means that DTA can access up to 16MB of extended
memory without all that messy EMS/XMS/virtual memory management.
It also means DTA doesn't recognize EMS or page to disk any more.
If you require EMS or virtual memory, stick with the original
version 1.8.
Two extra files (DPMI16BI.OVL and RTM.EXE) are distributed
along with DTA to support protected mode. You must keep these
files either in the same directory as DTA, or in the DOS path.
When DTA is run from inside a Microsoft Windows enhanced-mode
DOS window, it'll use Windows' DPMI services instead of the
Borland DPMI server.
o MAYBE fixed a bug in FLC compression.
o DTA can read pictures stored in ZIP and ARJ archives as well
as the previously supported LZH archives ... IF you've got the
appropriate de-archiver (PKUNZIP.EXE, ARJ.EXE, LHA.EXE) in your
DOS path.

Rel 1.8 (11/08/92)
o DTA couldn't locate PCX, IMG, or GIF files in LZH archives... fixed.
o DTA can now read 24-bit PCX files (as well as the previously
supported 8-bit type).
o Added some additional error-checking so DTA will stop with an error
message if it runs out of disk space while creating output files.
Previously it just kept on trying to write to a full disk.
o In a previous version, made some changes in the GIF-writing
code to speed it up and use less memory. As a result, upside-
down TGAs weren't getting flipped rightside-up like they're
supposed to... fixed.
o Removed some useless experimental options: /L, /E, /O
If anybody ever actually found a use for any of these, let me know
and I'll put 'em back.
o Removed /W, since it was only added as a workaround to a bug that's
now finally fixed.
o Added GIF89a input. (extension blocks are IGNORED)
o Fiddled with output-file parameters. You now specify the output
filename with /O, and the format with /F instead of specifying
both with /F
o Added /H, which lets you adjust brightness. /HA50 increases the
values of red, green, and blue by 50%. /HR200 /HG-25 increases the
value of red by 200%, and decreases green by 25%.
o Added /C, which lets you skip frames when scanning for a palette.
If you use /C5, then DTA will only scan 1 picture for every 5
in your animation.
o Added /K, which lets you skip input files when generating a
flic. If you use /K3, DTA will ignore 2 frames out of every 3.
If you use both /K2 and /C2, then DTA will skip every other
picture, and create a palette out using only 1 of every 4
o Fiddled with the time blurb so it writes in minutes and seconds
instead of milliseconds.

Rel 1.7g (08/27/92)
o Fixed a rarely-encountered bug in 320x200 FLI compression.
o Reworked the documentation (actually, Dan Farmer did most all of
this task).

Rel 1.7f (07/19/92)
o Fixed some weird bugs in selecting output filenames for GIF files.
o Added /PO option, which causes DTA to create a single optimal
palette even when it's creating GIF files.

Rel 1.7e (07/02/92)
o Removed all the logging code, since the memory allocation
problems seem to be licked.
o First attempt at creating hi-color and true-color FLIs.
(To create a hi-color FLI, use the /b16 parameter. The
resulting file with have an extension of FLH. To create
a true-color FLI, use /b24. The resulting file will have
an extension of FLT.)
So far there's no player that will display these new files,
but Steve Enns is working on one.
o The color-mapping speedups in the previous version resulted
in bad color choices in some cases. Added an extra
step between palette selection and color-mapping that
should improve it.
o Added /RA parameter (auto-resolution)... tells DTA to use
the size of the first picture to decide how big to make
an FLI. Note: use this parameter only with pictures with
even horizontal dimensions.
o When DTA used XMS to store pictures, it wasn't always clearing
the picture buffer when it was supposed to... resulting in
some background garbage when animating pictures smaller than
the FLI.
o Added /DR (random dither) parameter. Adds/or subtracts a
random value from each pixel. Random values are the same
across frames, so FLIs shouldn't explode in size like they
sometimes do with error-diffusion dithers like Floyd-Steinberg
and Sierra Lite.
Specify the range of the random number by appending a number
to the /DR parameter, like this:
dta *.tga /dr4
That will result in random numbers ranging from -4 to 4.
Note: I stole the idea from Stephen Coy's IMG2GIF program.
o Got rid of /M parameter, which affected the order of
color-tree pruning. It was a failed experiments in
improving color-selection.
o Changed /MI (max internal colors) parameter to /M
o Changed RLE TGA output so that line boundaries aren't
crossed within a packet... (to avoid a VPIC bug)

Rel 1.7d (04/25/92)
o Fixed some more memory allocation problems.
o Speeded up color-mapping. When dithering, there should
be a big speed increase. When not dithering there should
be a lesser increase.
o Added Sierra Lite dithering (/DS parameter). SL dither
another error-diffusion-type dither, somewhat quicker
than Floyd-Steinberg.

Rel 1.7c (04/15/92)
o Fixed a bug in grayscale output.
o Reduced the amount of memory required for the color tree.
o Fixed some other memory-related problems.
o Got rid of V (verbose), and added /LOG (which causes
assorted memory allocation info and other boring
debugging-type data to be written to a file called "DTA.LOG".
o Added /MI parameter (which means "maximum internal colors").
The octree color routine allows only 256 colors to exist
in the tree at any time. Whenever the number of colors
exceeds 256, parts of the tree get truncated.
With /MI, you can change that maximum number of colors
to something other than 256.
If you change it to a number larger than 256, then the palette
will not be reduced to 256 colors until DTA is done scanning
This can result in a nicer palette, but it uses more
memory. I've had some nice results using "/mi2048".

Rel 1.7b (04/01/92)
o Rel. 1.7 was missing a last-minute .FLI bug fix and
a document update or two.

Rel 1.7 (03/31/92)
o Reads 256-color PCX files.
o Reads and writes 16, 24, and 32-bit TGA files, compressed
or uncompressed.
The default is compressed TGA-24. /B16 selects 16-bit,
/B32 selects 32-bit. /NC selects no rle compression.
Dithering works for TGA-16 files... But if you use ordered
dithering, DTA ignores any 'strength' value and uses a
default (because TGA-16 color reduction is regular).
o When scanning for a palette, it's a lot faster than before
when reading 256-color pictures.
o Fixed EMS bugs. Added support for XMS.
o Added /332 (3/3/2 palette) option
(If you use ordered dithering, DTA ignores any supplies
'strength' value and uses a default).
o Added /NM (no-remapping) option. When reading 8-bit
colormapped files, this tells DTA to use the original
palette that's included in the file instead of creating
a new palette from the colors in the picture, and to skip
re-mapping the picture to the palette.
If you're creating an .FLI from a bunch of .GIFs or .PCXs,
then DTA will use the palette from the first input file,
and it will assume that all of the pictures use the same
palette. This will result in slightly better-looking output,
and DTA will generate the .FLI *MUCH* faster.
If these files do not use the same palette, then do not
use this option.
/NM does not work with any sort of dithering, frame-averaging,
or anything else that requires fiddling with colors.
If you're converting .PCX to .GIF, then the single palette
restriction does not apply...

Rel 1.6 (01/25/92)
o Can write uncompressed Targa-16 files (use the /B16 switch).
o Can read compressed and uncompressed Targa-16 files (like
those you can produce with POLYRAY).
o Added /M option, which causes the palette tree-reduction
routine to prune the most popular branches instead of
least popular. The resulting palette should be biased
a bit more toward detail in a picture instead of smoothly
shaded areas.
o Added /L option, which allows you to set the initial
(pre-reduction, that is) color resolution for one or
more of the color components (r,g,b) to something other
than the default 6 bits. It doesn't make sense to set
it higher than 6 bits, since that's the most VGA can display.
But you could, for example, tell DTA to use fewer bits for
green and blue so that the reds get more representation,
like so:
dta *.tga /lb2 /lg3
(that sets the blue resolution to 2 bits and the green
resolution to 3 bits).
o When averaging pictures, DTA used to use an equal percentage
of each of the original pictures. That's still the default,
but with the /E option, you can change this. /E lets you
specify two numbers: (a) the amount of weight to give the
first picture, and (b) how much to increment that weight
for each successive picture. So, assuming you start with
three TGA files, if you use this command:
dta *.tga /a5 /e2;1 /fg
the first frame will get a weight of 2, the second will get
3, and the third will get 4. Which means the resulting
GIF file will be made up 22% from the first picture, 33%
from the second, 44% from the third.
This can be used with /t (trail) as well as /a (average).
The increment number can be negative... if you want
the first picture to stand out more than the others.
The default for the initial weight is 1. The default
for the increment is 0. If you use /e without specifying
an increment, then the increment is set to 1.
If you don't specify either number, then both numbers are
set to 1.
o Fixed a bug in averaging GIF files.
o When creating a new TGA file, DTA forgot to use the resolution
of the input file like it does with GIF files. Fixed.
o Can now ignore comments in TGA files properly.
o When creating GIF files, DTA now saves at the same
time it performs the mapping step, instead of saving to
a buffer and saving it at the end.
For pictures that would require virtual memory (bigger
than 320x200), this'll save some time.
o Added 640x400 resolution for low-end SVGA.

Rel 1.5f (12/13/91)
o DTA can now read IMG files created by the Stephen Coy's
Vivid raytracer.
o Speeded up some input functions that were slowed down
because of some recently added features.
o Another probably unsuccessful attempt to fix the same bug
that 1.5d was supposed to fix.
o Fixed a bug in dithering (sometimes DTA wouldn't dither
the whole picture).
o DTA wasn't deleting a temporary file that it sometimes
creates. Now it does.

Rel 1.5e (11/24/91)
o DTA can now read GIF files as well as TGAs. GIF87 only,
no interlacing allowed.
o Added a /V (verbose) command line option. If you
specify this, DTA will tell you exactly how many
bytes each frame in an animation take up.
o Added the /W option, which forces DTA to use the
SVGA-type compression method, even if generating
a 320x200 .FLI.
o By the way the fix in 1.5d didn't help. However, the
problem is pretty rare. Maybe next time.
Rel 1.5d (11/18/91)
o Hopefully fixed a bug that caused problems with Trilobyte's
PLAY program.
Rel 1.5c (11/11/91)
o Attempting to support 800x600, 1024x768, and 1280x1024
SVGA resolutions. Note: This is completely untested.
I haven't the foggiest idea if it works or not.
Rel 1.5b (11/09/91)
o Added "expansion" feature, which inserts averaged frames
between regular frames.
o Fixed some bugs in some weird combinations of 3d, "pingpong",
and "trail".

Rel 1.5 (11/05/91)
o Ordered dithering
o VGA "mode x" .FLI resolutions
o fixed a bug in 320x200 delta encoding
which would sometimes cause animations to crash
o "averaging" and "trailing" (where individual frames are
created by averaging the colors across multiple input
files). The difference between "averaging" and "trailing" is:
If you run an "average" of 2 on 10 .TGAs, frame 1 will be
an average of files 1 and 2, frame 2 an average of files 3
and 4, and so on.
If you do a "trail" of 2 on 10 .TGAs,
then frame 1 will be an average of 1 and 2, frame 2 an average
of files 2 and 3, frame 3 an average of files 3 and 4, and
so on.
o Red/blue-type 3d
o TGA output (pretty useless unless you're also averaging
input files or creating 3d images). DTA does not compress
TGAs just yet.
o Removed .PCX output option.

Rel 1.4 TEST 2 (10/12/91)
o DTA can now read compressed .TGAs.
o DTA can new create .GIFs or .PCXs.
o Before 1.4, DTA would give up on compressing a frame if the
compressed version required more than about 64K to store.
This was fine for 320x200 animations, where an uncompressed
frame only requires 64K anyway, but it left something to
be desired for 640x480 animations, where an uncompressed
frame takes up about 300K.
Now, 640x480 frames will be compressed unless the compressed
version takes up more than about 300K.
Rel 1.3c (09/14/31) (second bug fix)
o .MAP files were missing carriage returns between
Rel 1.3b (09/13/31) (bugfix release)
o Fixed bugs from the disastrous Rel. 1.3
(including 640x480 .FLIs)
o Big speed increase in palette creation

Rel 1.3 (09/03/91)
o 640x480 .FLIs
o Better .FLI compression.
o In pre-1.3, the animation was stored in a temporary file,
then at the end a new file was created with the .FLI header,
and then the temporary file was copied onto the end of that.
Now, only one file is used to store the animation, and the
header gets modified at the end of the process.
o Frame #1 is now stored in a temporary file so that it doesn't
need to be recreated for the "ring frame".
o Added /G32 option for creating 32-level grayscale animations
for laptops.
o Fixed some bugs.

Rel 1.2 (08/06/91)
o Added support for exit codes returned by LHA 2.13...
If LHA returns a non-zero exit code, DTA will abort.
o If you hit the escape key while DTA is running,
DTA will abort as soon as it's done processing
an input file.
o Changed the way input files are sorted.
o Added /p (pingpong) option. (makes animation go in
0-1-2-3-4-3-2-1-0 order instead of 0-1-2-3-4-0).
o Added support for lists of files to process.
Tell DTA that a file is a list by preceding the filename
by an "@" character. A list file may contain
the names of .TGA files or .LZH files. You can't put
the name of another list file in a list file.
o Will now clip .TGA files with a resolution greater than
320x200 instead of choking on them.
o Added /O (sort order) option. When DTA has to sort filenames,
like when you use a wildcards or when you just give the name
of an .LZH file, "/O+" (the default) will tell DTA to sort
in ascending order, "/O-" in descending order.
This option is positional... it will only affect files
specified AFTER the option on the DTA command line.
You can use this more than once on a DTA command line.

Rel 1.1 (07/15/91)
o Can now extract .TGA files from inside of .LZH files.
(by shelling to LHA.EXE) Takes longer, but can save plenty
of disk space.
o Added /s (to specify playing speed).
o Doesn't choke on truncated .TGA files any more.
o Rewrote some I/O routines so scanning palettes is quicker.

  3 Responses to “Category : Printer + Display Graphics
Archive   : DTA18G-2.ZIP

  1. Very nice! Thank you for this wonderful archive. I wonder why I found it only now. Long live the BBS file archives!

  2. This is so awesome! 😀 I’d be cool if you could download an entire archive of this at once, though.

  3. But one thing that puzzles me is the “mtswslnkmcjklsdlsbdmMICROSOFT” string. There is an article about it here. It is definitely worth a read: