Release Notes: This release adds accumulated cleanups, minor bugfixes, and improvements.
Release Notes: The BMP codec source and memory allocation were improved and cleaned up. The vector font API was improved to indicate an error if loading of the font file failed.
Release Notes: This release makes reading plain RAW image data more robust. It fixes casting some escape values in the PDF codec. It fixes reading RAW images data in the econvert CLI frontend.
Release Notes: A problem with inverting 1-bit PBM ASCII files was fixed. Support was added for the latest Evas library version. Some code cleanups were done.
Release Notes: This release adds support for relative vector lines, offset vector text, and changing the stroke width multiple times with the econvert command line frontend. It further extends the new image iterator and improves the correctness of the neighbor de-noise filter. It fixes JPEG EXIF parsing, as well as the external API bindings to detect Perl 5.10 and the Python module path.
Release Notes: HSV transform was optimized. BMP and PCX codec were improved for certain color-maps and performance as well as certain colorspace conventions and image iterators. Various other fixes were made, including adaptions for various new compiler errors and warnings. The internal copy of the AGG library was removed (to avoid code bloat and fragmentation).
Release Notes: EXIF based rotation was implemented while loading JPEG and dcRAW images. Vector text drawing support was enhanced to support rotating the text, computing the geometry by specified alignment gravity, as well as stroking the text outline. The stroke width for vector geometry can now also be specified in the command line frontend. The optimized orthogonal rotation code was fixed for HDR images. Support for compiling with a non-GNU compiler was improved.
Release Notes: PCX and TGA (Targe) image writing was implemented. The palette unpacking library function was optimized to also recognize unsorted gray colormaps and thus optimize further processing in gray instead of RGB data.
Release Notes: A new image append option was added. Thumbnail down-scaling was optimized. RGBA pixel accessor get and set methods were added to the stable, external API. The lossless JPEG handling and loading of black and white TIFF images was optimized. Various algorithms were further sped up. It is now possible to specify font files in the CLI frontend. The image viewer class was refactored for better code reusability. The configure check for libtiff's C++ wrapper was improved and compilation with older libjpeg versions was fixed.