Skip to content


Subversion checkout URL

You can clone with
Download ZIP
tree: 1a091ac3c9
Fetching contributors…

Cannot retrieve contributors at this time

129 lines (116 sloc) 5.69 kB
Unreleased changes
- Removed the --with-sample-depth configuration option.
- Removed CMake support.
- Renamed "boxmv" filter to "boxavg".
- Changed how the -encoding option works. On Windows, write UTF-8 text by
default if not writing to a console. Removed the "utf8raw" and "utf16raw"
Version 0.9.10 - 2 Jun 2012
- Support reading and writing BMP files with more than 8 bits per sample.
- Support writing v5 BMP files (use -bmpversion 5).
- Support for reading some v4 and v5 BMP files.
- Support for writing OS/2-style BMP files.
- Added -reorient option.
- Support for reading Orientation tags from JPEG/Exif files.
- Ability to rotate and mirror images.
- Added "-precision 32" option.
- Added an alternate box filter, "boxmv".
- Better support for -grayscale when writing BMP images.
- Support "-cc <r>,<g>,<b>" syntax.
- [Windows only] Support for reading images from, and writing images to, the
- Changes to allow data streams other than regular files. Filenames that may
contain a colon now need to be prefixed with "file:".
- Support for custom bit depths (sBIT chunks) when writing PNG images. Added
-depthalpha and -depthgray options.
- Support for custom bit depths when writing 16-bit BMP images.
Version 0.9.9 - 29 Mar 2012
- Various changes to how non-square pixels are handled. Added -noresize
- Support for writing 16-bit BMP images.
- Support for reading 16- and 32-bit BMP images.
- Support for RLE-compressed BMP images with transparency.
- Support for reading BMP files.
- Support for writing BMP files with RLE4 compression.
- Support for writing BMP files with RLE8 compression.
Version 0.9.8 - 18 Feb 2012
- Support for building using GNU autotools.
- Support for two types of custom grayscale formulas.
- Added -gsf as an alias of -grayscaleformula, and made it unnecessary to
specify "-grayscale" if you set the grayscale formula.
- Support reading CMYK/YCCK JPEG images.
Version 0.9.7 - 30 Dec 2011
- Support for custom memory allocation functions.
- Renamed -pngcmprlevel option to -zipcmprlevel, and made it apply whenever
zip/deflate compression is used.
- Added Zip compression support for MIFF images, and made it the default.
- Added "-compress" option.
- Added support for transparent virtual pixels ("-edge t").
- Added "-translate s" option, so you can specify the translation amount in
source pixels instead of target pixels.
- Enabled transparency support for WebP images.
- Made the "edge policy" setting be per-dimension.
- Removed the -filteralpha feature.
- Stopped disallowing the "s" edge policy when -translate or -offset is used.
- Changed the default downscaling filter to catrom.
- Changed the box filter to have more predictable behavior.
Version 0.9.6 - 1 Dec 2011
- Added -translate option, to fine-tune the position of the image.
- Added "-blur x" option, to make it easy to do simple interpolation.
- Support -blur when using the "mix" filter.
- Added a way to request a specific density label.
- Added an option to write JPEG files using arithmetic coding.
- Don't clamp alpha samples before using them to apply a background color.
- Added -noincludescreen option, for more control over how GIF images are
- Added -page option, to read arbitrary images from GIF files.
- Changed the default bit depth to always be 8, instead of having it depend
on the depth of the input file.
Version 0.9.5 - 8 Oct 2011
- Use a binary search to find the nearest available color. Improves speed
significantly in many cases.
- Random dithering now uses a private pseudorandom number generator.
- Changed the default PNG compression level to 9.
- Updated the format of grayscale MIFF files to be compatible with the
current version of ImageMagick.
- Added -encoding option, to set the character encoding for text output.
- Allow -cs to set the output colorspace even when IW doesn't think it's a
good idea.
- No longer attempting to make the output colorspace be the same as the
input colorspace.
Version 0.9.4 - 9 Aug 2011
- Added -density option, to give some control over how the density of the
target image is calculated.
- Added -nowarn and -noinfo options.
- Support for reading GIF files.
Version 0.9.3 - 30 May 2011
- Support for rational numbers like "1/3" in the command line.
- Basic support for Google's WebP file format.
- Partial support for ImageMagick's MIFF file format.
- Support for image sizes that are relative to the original image size.
- Moved strings to string tables.
Version 0.9.2 - 7 Apr 2011
- Made the library use only UTF-8 for character encoding.
- Improved resampling speed.
- Read the background color metadata from PNG files, and use it as the
background color in certain circumstances.
- Detect input file formats based on the contents of the file, instead of the
file name.
- Added "sinc" filter.
- Support for writing uncompressed TIFF files.
- Replaced "-nobinarytrn" option with "-noopt".
- Added a "halftone" ordered dither.
- Support for reduced input bitdepths; respect PNG sBIT chunks.
- The image format modules now use abstracted I/O.
- Support for writing BMP files.
- Added option to set the PNG compression level.
- Improved speed of colorspace conversion.
Version 0.9.1 - 22 Feb 2011
- Added a crop feature.
- Support for optimizing to an image with binary transparency.
- Added "-interlace" option.
- Added "-jpegsampling" option.
- Retain pixel density information if the image isn't being resized.
- Support for optimizing to a 4- or 16-color grayscale image.
Version 0.9.0 - 6 Feb 2011
- Initial public release.
Jump to Line
Something went wrong with that request. Please try again.