The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.
Revision history for Perl extension Imager.

0.01  Thu May  6 15:25:03 1999
        - original version; created by h2xs 1.18

0.02  Mon May 10 20:20:20 1999
        - Added PPM writer, and a new return type to
        be used to return undef() from a function returning
        (int)0.

0.03  Thu May 20 23:23:23 1999
        - Added Pseudo image handling thingies, now
        you can use your own get and put pixel routines.

0.04  Mon May 24 22:14:55 1999
        - Rewrote parts of jpeg support.

0.05  Tue May 25 23:40:01 1999
        - Added png support with libpng
        fixed error reporting and return types from
        some format routines.

0.06  Wed May 26 19:05:39 1999
        - Fixed Data::Dumper being used when it shouldn't have
        added feature checking to see at run time if 
        certain formats are avaliable.
        removed some ancient C++ style comments that hadn't 
        been properly removed, some out of bounds errors in ppm.c
        fixed tests so missing features are skipped.

0.07  Thu May 27 03:15:00 1999
        - Fixed the typemap for i_img and i_color so that you can pass
        a null pointer to them without getting a warning when warnings
        are enabled. Maybe this is counter perl but it greatly increases
        the usability of the functions.

0.08   Fri May 28 14:13:21 1999
         - Added support for gif via, giflib or libungif.
         gif is so ridden with patent issues but the user
         can decide which library to use since they have the same
         include files and calling interface.

0.09   Mon May 31 17:52:32 1999
         - Added image scaling per axis - faster than doing both
         axes at the same time.  The current method uses lancoz
         kernel and filtering. But more types should be added -
         including: nearest neighbor, linear, and bicubic.

0.10   Mon Jun  7 03:25:54 1999
         - Added T1 postscript font handling. It's very primitive
         at the moment but creative people can do nice stuff 
         with it.  Currently it is not possible to have it 
         generate an image of the correct size for a given string 
         but that is on the way.

0.11   Mon Jun  7 14:43:35 1999
         - Added T1 features for direct to image rendering in 
         a single color.  fixed some debug messages going to 
         stderr instead of a log file.

0.12   Tue Jun  8 02:58:43 1999
         - Fixed bugs in jpeg.c when loading images. Also specified
         prototype behaviour to on in the Imager.xs file. The
         Makefile.PL step doesn't complain anymore so that is
         hopefully fixed.

0.13   Wed Jun  9 06:55:58 1999
         - Fixed Imager.xs for init_log call.  Minor fixes here
         and there.

0.14   Thu Jun 10 14:11:22 1999
         - Rewrote most of the logging stuff so that it is now
         possible to remove all logging from the XS part of the
         module.  Also some primitive check of memory usage was
         added. Claes Jacobsson sent in some filters, inverter,
         contrast and a noise filter.

0.15   Mon Jun 14 08:13:29 1999
          - Wrote minor enhancement on the calling syntax list.
          Started on the object interface - added better support
          for quering avaliable filetypes.  Fixed memory leaks in       
          most loaders. New filters from Claes in this version
          are bumpmap, postlevels and mosaic.

0.16    Wed Jun 16 20:54:33 1999
          - Added fixes to the BEGIN and END parts, added writer
          function for the OO interface. Also added basic IPTC 
          reading support for jpegs.  Also a few filters have been
          added to the OO interface.

0.17    Thu Jun 24 11:09:15 1999
          - Added dynamic loading stuff - It's still missing a nice
          global symbol table. This will be fixed in next release.
          also calling the plugins is not all to easy at the moment.


0.18    Mon Jun 28 12:31:33 1999
          - Added global symbol table - plugins now need a symbol
          table pointer in them. When the module is loaded it is
          set to point at the global symbol table. Also some barebones
          Makefile.PL has been made in the dynfilt directory - it works
          on my system - I hope it does on yours.


0.19    Fri Jul  1 15:00:03 1999        
          - Added a way new scaling method for creating easy previews.
          It only uses nearest neighbor, so it's doesn't look very nice
          but it may be better for applications like remote sensing.

0.20    Mon Jul  5 10:15:37 1999
          - Added and rewrote documentation.

0.21    Mon Jul  6 19:15:37 1999
          - Fixed a bug in the scaling routine - it wasn't
          handling 0< cases.

0.22    Sat Oct  9 07:04:14 1999
          - Added a new method to write gif images - now 
          it is possible to have a part of a palette fixed.
          this is very usefull if one needs to make sure that
          some color like white is in the pallete.  This method
          also allows some ditherding and gives better colormap
          than the mediancut from the gif libraries.  It does
          need much more cpu power though.  Hopefully later versions
          will be faster.

0.23    **************** Internal release only
          - Fixed the %instances bug - caused ALL memory to be leaked.
          Added real noise function - need feedback on how it should
          be used. Also box(), and polyline are now in place.  Polygon
          is missing but antialiased line drawing with integer endpoints are
          done but should be replaced with a version that can have 
          floating point endvalues. Two noise filters addded.

0.24    **************** Internal release only
          - Converted i_color into an object from a reference, so now it's
        giving an object instead of a void ptr or an integer or something.

0.25    **************** Internal release only
          - Added basic Truetype functionality - still needs a rewrite
        to be decent. Currently it's a port of a demo program that
        uses an awful amount of global variables and there is much IO since
        no caching of glyphs is done.

0.26    Tue Nov 23 03:57:00 1999 > Development release <
          - Added transformations so that an image can be wrapped.
        To achive decent speed a C based stackmachine is included. As a result
        transformations need to be specified in rpn (postfix) notation. It
        also can use the Affix::Infix2Postfix class to do the conversion for it.

0.27    Tue Dec 28 03:57:00 1999 > CPAN release <
          - This is a bugfix version mostly, thanks to claes for pointing
        out the problems - fixed palette saving wasn't working correctly after
        version 0.24 - rather surprised this didn't crash everything.
        Also fixed that for t1 fonts the bounding box wasn't being reported 
        unless the font had been used before.  This is either a bug in t1lib
        or a mistake in it's documentation.  Another lingering bug since 0.24
        what that $img->box() wasn't creating it's default color properly.
        Added i_tt_text() method and more debuging to the truetype routines.            
        truetype testcase fixed and old debug rubish removed.

0.28    Tue Jan  4 05:25:58 2000 > CPAN release <
          - Only fixes to truetype test and transformation tests.
        Thanks to schinder of cpan testers for testing and reporting.

0.29    Tue Jan  4 21:49:57 2000 > CPAN release <
          - fixes to get rid of warnings under pre 5.005, 
        Fixed broken preproccessor directives to work on non gnu
        compilers.  Fixed DSO test on HPUX - both code errors and
        HPUX creates .sl instead of .so so the tests were failing.

0.30    Sun Jan  7 05:00:06 2000 > Bunch of Alpha releases <
          - An attempt to automate installation.

0.31    Sat Jan 15 03:58:29 2000 > Fixes fixes fixes <
          - Fixed a bug pointed out by Leolo where loading gifs
        skips the first line of the imageload() has been
        by read() - for now load is an alias for read.  It will
        be removed in the future.  Also, fixes dynamic loading on
        systems that prepend an underscore to symbols.  At the present
        the only system that I know of that needs this is OpenBSD.
        BUT YOU MUST RECOMPILE ALL OF YOUR OLD MODULES AGAINST THIS BUILD.
        Added getchannels() method ( How did I manage to delay this
        untill now ).  Some document changes but nothing substantial.
        Also fixed the png read/write routines to handle all colorspaces
        and images with alpha information.  Also now it's possible to 
        have Imager guess the format of the files to load or save 
        when passing files to read or save from the filename.
        Also all of the tests except dynamic loading now pass on OS/2.

0.32    Tue Feb 29 17:26:00 2000 CPAN RELEASE
          - Added the getcolorcount method.  Fixed interlace handling
        on png images.  Fixed the missing channel count in crop()
        method. Rewrote most of t1lib database stuff - created color
        and font classes.  T1 stuff is mostly done - TT things were
        rewritten too and now include most of what is needed for
        pixmap caching.  Added documentation for fonts.  Comments have
        been added to some of the relevant c-routines.  Added a copy()
        function in Imager.xs and a corresponding method name.
        Changed the underlying data type for the raw XS images from
        pointers to objects - this will hopefully catch the most
        basic errors and keep the segfaulting down.  This means that
        all of the underlying XS calls for readjpeg, readgif, readpng 
        and readraw do not take the first parameter any more.  
        Made fixes to keep it not spewing warning on 5.004 perl.        

        **** If you had any code that didn't use the OO interface ****
        ****         It will probably not work any longer         ****
        
0.33    Beta -- No final
          - Fixed the end message from Imager 0.32.  Destroy called
        on an empty image.  Did some work on the polygon method.  
        Some clean up in the Makefile.PL script.  Fixed a buffer
        overrun in the t_transform in Imager.XS.  Fixed the 
        error handling in the jpeg loader.  It now correctly
        returns undef if a load on an image fails.  It also 
        sends the error messages to the log file.  Added errstr()
        method to the image object.  Added a new way to read()
        objects from scalars.  So far this is only implemented for
        jpeg, png and gif.  ppm and raw soon - as always if someone
        wants to do an overhaul on the ppm stuff feel free.  It seems
        like such a basic format that interfacing with a library is more
        work than implementing all of the needed routines instead.

0.34    Beta -- No final
          - Bunch of documentation fixes, backed out ppm code.
        Put in TonyC's giant transform2 patch.  Fixed the patch
        to make it ansi compliant.  Fixed a bunch of bugs in the
        Freetype code with regard to vertical and horizontal 
        positioning and bounding boxes.  Cleaned up a lot of the
        code so it runs under -Wall.  Code that is still in 
        development such as the polygon converter do not compile
        cleanly.  Fixed the non antialiased versions of truetype
        dump to image routines.  Also removed the FIXME for the 
        hardcoding of antialias in the Imager string method.
        Fixed sign error and a missing cache of the bounding box
        calculation for the rasterize function.  Removed some
        debugging code I forgot to remove.  Added iolayer.h
        and iolayer.c but they don't do anything for now.
        
0.35    pre2 -- No time yet
          - Fixed some compile warnings for various files under -Wall.
        Added functionality for jpeg reading of seekable files, it's not
        really working yet.  This version is pretty much *not* working.
        Do not install unless you intend to do a lot of development.
        Repeat - it doesn't even pass tests (but it compiles).  Ok now reading
        jpegs works from scalars, my guess is that it also works from non
        seeking sources such as sockets or pipes.

0.35  pre3 - No time yet
          - Added the *right* patch from Tony which combines
        the common code from i_readgif and i_readgif_scalar into
        i_readgif_low.  Added tiff reading support through iolayer.

0.35  pre4 - No time yet
          - Added tiff writing (no options) support through
        iolayer.  Also made some small fixes for the iolayer reading
        (was always doing two reads when one was needed).  Patched the
        Imager::read() call so that it now uses a mixture of old and new
        functions.  

0.35  pre5 - No time yet
          - Fixed various gnu'isms in the c code (some bugs in the link list
        implmentation).  Fixed missing #skip codes when gif format is not
        present in any form.  Added fixes for 5.004_04 in the transform2 function.
        Made sure it compiles cleanly with sun's cc.  Switched from a .jpeg
        for transform2 check to a .ppm file so it runs when jpeg is not 
        present.  Added a test for tiff files in t10formats.t.


0.35  pre6 - No time yet
          - Fixes to Makefile.PL.  Should find freetype includes on more
        distributions now.  Ran tests on Solaris and Hpux, minor fixes.
        Compiles with some warnings on with both hpux and solaris' cc.  
        Made some minor changes to the documentation.  Fixes to tiff.c log
        code.

0.35  pre7 - No time yet
          - Fixes 64 bit bug on freebsd.  While libtiff mirrors the effects of
        lseek it's toff_t is a uint32, while lseek uses off_t which can be a 64
        bit quantity.  Added the IM_LFLAGS environment variable to help
        people with broken libgifs (that want to link with X).

0.35  Sun Jan 28 19:42:59 EST 2001
          - More makefile fixes, fixed a few signedness warnings.  
        Checked to see if it compiled cleanly on Solaris and HPUX.
        Fixed a 5.004_04 warning and added more ENV flags for makefile.

0.36    Mon Jan 29 09:36:11 EST 2001
          - String as 0 or "" caused an error in $img->string().  Fixed a 
        documentation error invoving string() method syntax.  Merged a patch
        for non antialised truetype fonts.  Fixed an error in the Makefile.PL
        which caused a makefile to be generated that bombed with sgi's make.

0.37    Mon Tue 30 09:36:11 EST 2001
          - Several documentation fixes.  Pod documentation for almost every
        function in image.c.  Added sys/types.h include in iolayer which was
        causing problems on various linux systems.

0.38    pre1 - No time yet
          - Fixed a braindamaged fix on the Makefile.PL file.  Moved the 
        code for Imager::Color into lib/Imager/Color.pm.  Wrote some pod
        about how it works.  Made the names of Imager::Color XS routines
        all begin with ICL_ and used the prefix rules of XS to get nice names
        in perl.  Found a bug (not fixed) in how XS handles 
        returning an object to an object it had as a parameter (double
        free).  

0.38    pre2 - No time yet
          - Fixes lots of for documentation, patch for freetype bounding
        box handling.  Split put code for Imager::Font into Font.pm and added
        more documentation for Font.pm.  Added string local ascender and
        descender for tt fonts.  Note that this needs to be added to t1 fonts
        before 0.38 final.

0.38    pre3 - No time yet
          - Fixed an in consistency in the bounding box functions for t1
        fonts.  Now both versions get the 6 argument bounding_box method
        described in Imager::Font.  Started converting the comments in 
        font.c so that they are viewable by doco.perl.  Added two examples
        of filters.  Need to make them more usefull and then add more
        notes than are in compile.txt.


0.38    pre4 - No time yet
          - Completed adding pod comments to font.c, tiff.c and iolayer.c.
        Those along with image.c should now have every single function 
        described in pod format.

0.38    pre5 - No time yet
          - Replaced ppm.c with pnm.c which adds support for pbm/pgm/ppm
        files ascii and binary formats.  Added patches for the gif routines.
        Patched some of the color quantizing routines (Leolo and TonyC).
        There is one bomb and one warning in this test, and frankly I don't
        see why they are suddenly there.

0.38    pre6 - No time yet
          - Patch from Tony that fixes infix when Parse::RecDescent is present.
        Checked some cases where malloc/free were used instead of mymalloc/myfree.
        Added bufchain iolayer code.  You can now write to a chain of buffers and
        later slurp it all into a single perl scalar.  Found some oddity of t/t10
        test not giving the right return value when malloc debugging was enabled.
        Fixed some of the logging code and the malloc debugging options.  Added 
        more tests for tiffs.

0.38    pre7 - No time yet
          - Added i_gradgen code and put it into the filters hash.  Think a
        seperate pod for filters would be a good idea.  Also removed some of the
        debugging code from the iolayer.  Added pod comments to filters.c and
        looked over the code.

0.38    pre8 - No time yet
          - limited Win32 support, Imager installs and builds using VC++,
        but there's no image/font format support yet.

0.38  pre9 - No time yet
          - Added lots of color quantization code from Tony with benchmarks.
        Also fixes ugly stack overrun in old version.  Added fixes for the lmfixed
        problem.  Four of them, let's see which is fastest.  This version adds
        some voronoi partitioning - it's dog slow but it's a reference implementation
        to check if faster algorithms are doing the right thing [tm].  Added a check
        for giflib 3.
        - Win32 support for libpng
        - fixed set_internal creating 2 refs to one object
        - Win32 support for jpeg
        - Win32 support for tiff
        - base level error reporting code added, I still need to modify
          older code to use it
        - fix translate=>'giflib' handling of single-colour images.  
          Previously a single colour image would cause an error in 
          MakeMapObject().
        - fix t/t10formats.t tests against giflib3
        - added test for transparent gifs
        - minor doc fixes for transparent gifs
        - make it less error-prone to write transparent gifs
        - documented the options for reading raw images
        - Changes switched to use spaces for formatting
        - switch gif.c to new error reporting with related changes to
          Imager.xs, Imager.pm
        - each of the image formats now have their own test file,
          extracted from t10formats.t, usually with som extra tests
        - Added flip() and docs to Imager.pm and i_flipxy to image.c.
          Added testcases too.
        - Fixed an overflow bug in png.c
        - added the colors parameter to Imager::read() which receives 
          the arrayref of colours that i_readgif(_scalar) returns
        - fixed a problem in makemap_addi in quant.c where only the first
          image was compared against the palette when the palette was being 
          refined.  I screwed up on this originally (TC).  This is the other
          half of a fix for Leolo's lmfixed problem <sigh>.
        - makemap_addi() now only puts colours into the palette that it
          found while optimizing the palette (so we don't get 256 colour
          palettes for 2 colour images :)
        - initial implementation of a row-based interface to low-level
          images (for performance)
        - changed Imager::read() for GIF so that the arrayref pointed to 
          by the ref supplied in colors contains Imager::Color objects 
          instead of more arrayrefs, keep this interface stable until we 
          can make an informed choice about i_readgif()
        - incorporated Micksa's faxable tiff support, treating as a type
          of tiff rather than a separate format
        - t/t70newgif.t now checks for gif support
        - added the convert() method
        - Added support for transparent to alpha conversion during loading
          of png files.  Note that libpng 1.0.1 is broken in handling this,
	        we know that 1.0.5 works.
        - support for writing PGM (portable gray map) files
        - Fix for i_gifread_low() for images with only a local and no global
          colormap and an abort if no colormap is present (which is legal
          in gif).
        - several fixes to i_readgif_low()
	- Fixed crop() method so that it is consistent with documentation.
        - make clean now removes the testout directory
	- added tests for the OO interface for tiff, added an option to
	  choose the resolution for fax tiffs, removed some unused code,
	  fixed minor problems with Imager::write() handling of image types
        - changed README note on libgif bug to refer to t105gif.t instead
	  of t10formats.t

0.39  pre1
	- split Imager::Font into a base, *::Type1 and *::Truetype
        - writing faxable tiff now allows 2 and 4 channel images
        - virtual image interface - 8-bit/sample images
        - paletted images
        - 16-bit/sample images
        - masked images
        - writing non-8bit images to raw
        - writing   ''       ''   to tiff
        - i_convert support for high-bit images and paletted images
        - i_copy for high-bit and paletted images
        - tests for rubthru
        - rubthru can now rub a 2 channel image over a 1 or 3 channel 
          image
        - rubthru for high-bit images
        - i_readgif_multi functions, which return all the frames from a
          GIF file, along with a bunch of meta-information as tag
        - OO interface to tags
        - OO interface read_multi() to read multi-image files (just GIF
          so far)
        - documentation for the multi-image GIF reading and tags
        - rotate() method with rotate by steps of 90 degrees (and docs)
        - fixed a bug in i_img_pal_new_low()
        - added gaussian to the filters list
        - documented the individual filters
	- fixed the right-hand side of bounding boxes for TT fonts 
	  where the last character overlaps the advance width (eg. 
	  italic fonts)
        - added rotation by exact amounts in degrees or radians,
          implemented using the matrix idea from Addi.  The procedural
          matrix transformer is exposed, but still needs testing (as XS)
          and needs an OO interface, hopefully with some helper tools,
          like the preset interface with ->convert().
        - MY::postamble() in Makefile.PL included a broken rule 
          (Makefile.PL 1.13.2.5)
	- support for GDI fonts under Win32
	- made that work for cygwin too (needs the w32api package)
	- freetype1 support under Win32
	- t1lib support under Win32
	- fixed a minor bug in font.c (invalid font files caused a SEGV)
	- checked cygwin for both t1lib and freetype1
        - freetype 2 support
        - exposed the matrix_transform() function to the OO interface
        - added Imager::Matrix2d convenience class
        - support for setting the resolution when writing to PNG
        - retrieve physical resolution to tags when reading from PNG
        - found an XS bug in the interface to i_tags_add()
        - fixed handling of writing text to a channel with freetype2 
          (sometimes the edge of a character would damage the edge of the 
          previous character)
        - some utf8 support for freetype2
        - some vertical layout support for freetype2
        - named parameters for specifying colors, with quite a few options.
        - glyph size issues for freetyp2
        - minor problem in handling of canon option
        - low-level bmp writing (moving it to laptop)
        - Windows BMP reading and writing
        - added OO interfaces for the mosaic, bumpmap, postlevels and
          watermark filters
        - added t/t61filters.t to test the filters
        - fixed some problems in jpeg handling from the exp_represent merge
        - fixed buffer flushing for wiol jpeg code
        - added some tests that will hopefully catch it in the future
        - added the OO interfaces to the mosaic, bumpmap, postlevels and
          watermark filters, and documented them
        - fixed a sample size conversion problem in i_gpixf_d() etc.
        - added simple color representation conversion functions (used
          in i_fountain().)
        - added the fountain filter:
          - creates gradients similar to paint software
          - 90% support for GIMP gradient files
          - OO interface and documentation
          - Imager::Fountain for building/loading fill definitions
          - named value translation for filters
        - added a generic fill mechanism
          - created versions of i_box() and i_arc() that can fill using
            generic fills
          - solid generic fills (with alpha blending if asked for)
          - hatched generic fills (with some options)
          - fountain generic fills
          - sample code to generate an examples page
        - improved the scale* hatches a bit
        - add the flood_fill() method (using the existing i_flood_fill)
        - implement i_flood_cfill() based on i_flood_fill, and 
          add general fills to the flood_fill() method
        - generalize the combine parameter to get different ways of 
          combining the fill with the target
        - the tt driver (freetype 1) didn't handle the first character
          hanging off the left of the character cell
        - the tt driver produces some artifacts when characters
          overlapped
        - error handling for writing jpeg images
        - writing paletted images to GIF now uses the image palette
          if it's small enough
        - $img->arc() now calls i_circle_aa() if a complete circle is
          being drawn in a plain color
        - image based fills
        - unsharp mask
        - make i_conv() clamp the bottom end of the samples range too
          (makes it useful for sharpening)
        - adjust ascender/descender values for FT1.x to avoid losing
          descenders (specifically the bottom of "g" in ImUgly.ttf or
          arial.ttf at 14pixels)
        - added tga.c to read targa images
        - added i_bumpmap_complex to do more accurate bumpmapping
        - added an image type with doubles as samples
        - change i_copy() and i_sametype() to handle double/sample images
        - added basic POD to trans2.c
        - transform2 now uses the error interface
        - myrealloc() is implemented for malloc debug mode
        - now buffer chains are freed when destructor for Imager::IO
          is called
        - adjusted the Lanczos window width for the scaling code and
          added code to normalize the filter generated to preserve
          intensity scaling.
        - constant edge extension for scaling operations
        - better error checking of automatic fill conversions
        - fix some range checking for the fountain fill segment type

0.39  pre2
	- A few scattered mymalloc/free fixes
        - io_buffer implemented as a source for io layer and
          XS and perl OO code added.
        - Test for iolayer t/t07iolayer.t implemented.
	- Fixed known memory leaks.
	- gif screen was wasn't being calculated correctly

0.39    Released: Nov 2 2001.

0.40  pre1
        - anti-aliased polygon fill
        - add general fill to polygon fill
        - make color values smarter for the drawing functions
        - implemented reading and writing the TIFF text tags
        - added prototypes for some of the derivative tags functions
        - read paletted tiff images into Imager paletted images
        - on partial tiff image reads, set the i_incomplete tag
        - tiff reading now uses the error stack
        - use the error stack value from reading bmp files
        - fix an error message in bmp.c
        - added has_chars() method to Imager::Font::FreeType2
        - freetype 2 bounding box function didn't know UTF8 (doh!)
        - write paletted images as paletted to tiff
	- initialize counter for packing 4 bit data
        - don't allocate hashboxes as locals since it overflows the 
	  stack by default in Darwin
        - applied T1 afm patch from Claes Jacobsson
        - split IM_INCPATH and IM_LIBPATH with $Config{path_sep}, so they
          work on Windows
        - Added memory pools for easy cleanup of temp buffers
        - Added read support for sgi .rgb files.
        - io_new_fd() now creates a FDSEEK io object
        - implemented i_readgif_wiol()
        - Imager->read() now uses i_readgif_wiol();
        - extend callback iolayers at C and Perl levels
        - implemented i_writegif_wiol()
        - split out Perl iolayer initialization into private methods
        - add tests for each type of iolayer in t50basicoo.t
        - read/write multi-image tiff files
        - tests in t50basicoo.t for multi-image/file
        - ASCII PBM files were reading white as dark gray (255 vs 1)
        - modify the Freetype2 font code to pick it's own encoding
          rather than using whatever strange translation FT2 happens to 
          pick
        - modify the FT1 font code to use the first encoding as a default
          if we don't find a unicode encoding
        - use the glyph bbox to calculate the glyph bitmaps for rendering
          with FT1 rather than the global ascender/descender.  This helps
          with fonts that have broken metrics.
	- fix calculation of descender for FT2, it was calculating the
	  minimum decent rather than the maximum for the characters in 
	  the string
	- didn't set default for bounding_box() utf8 parameter (caused a
	  warning when calling bounding_box() on a FT2 font with no utf8
	  parameter)
	- Added lib/Imager/Draw.pod documentation of primitives.
	- Added lib/Imager/Transformations.pod, some docs of simple transforms.
	- Added lib/Imager/ImageTypes.pod, draft of ImageType pod.
	- Added lib/Imager/Filters.pod, draft of Filters pod.
	- Added lib/Imager/Engines.pod, draft of Engines pod.
        - added getpixel() and setpixel() methods
        - added Artur's OSX dlload() emulation, with minor changes
        - modified _color() to work around a 5.6.0 bug
        - replaced old gif options with tags
        - we now log which memory block is being freed before giving 
          an error on it being re-freed
        - fixed stupid bug in deleting tags
	- fixed minor pod errors involving &gt;

0.40pre2
        - make t1log optional, defaulting to off.  You can enable the log
          with Imager::init(t1log=>1) (Ticket #369)
        - quote a few hash key strings to prevent warnings on 5.004
        - modify quantization code to handle 1 channel images 
          correctly (Ticket #365)
        - make channel pointer to i_gsamp() const int * so we can pass
          const arrays
        - handle the presence of the default broken giflib better,
          by giving them some more prose about the problem, and skipping
          all but the first test that fails with the broken giflib
        - i_box_cfill() went into an infinite loop on fountain fills
          or fills on images with more than 8-bits/sample
        - hide more of our Darwin dlload emulation to prevent runtime
          symbol conflicts
        - use INT2PTR() and PTR2IV() macros instead of direct casts to
          prevent pointer vs int size warnings on 64-bit builds of perl.
        - add our own INT2PTR and PTR2IV definitions when perl doesn't
          supply them
        - difference() method
        - hide (with static) or rename many symbols that could possibly
          cause conflicts with other libraries or perl
        - Fix runaway cache problem for freetype1 font cache.
        - Added version logging when log is started.

0.40
        - difference() wasn't setting an alpha channel when the input
          image didn't have one
        - improve crop() documentation, see
          http://perlmonks.org/index.pl?lastnode_id=155767&node_id=155428
        - handle the first "buggy giflib" test more portably.  Previously
          it used fork() which caused problems on systems that didn't have
          a real fork().

0.41
        - skip() in testtools should be skipn() and it should use the 
          loop variable for the test number

0.42
        - quote the 'min' parameter to scale in Imager::Transformations
          and at least mention it in the docs beyond the examples
        - document the values for the read() and write() method type 
          parameter
        - support UTF8 text with Freetype 1.x 
        - second parameter to SvPV() must be of type STRLEN
        - Doc pathces from cogent.
        - Fixed out of bounds access on bitmap for flood fills.
        - some char * vs unsigned char * casts missing on OSF1 v4.0
        - some enums had , on the last item in datatypes.h, which OSF1
          didn't like
        - Compaq C 6.4 (OSF1) claims to be C99 but doesn't provide 
          stdint.h, don't try to use it under OSF
        - fix missing initialization in tags.c
        - Change i_draw to i_line and have it use Bresenham's line
          drawing algorithm.
        - support has_chars() method for Freetype 1.x
        - fixed log message for i_ft2_has_chars()
        - fixed some broken checking for UTF8 in FT2 code
        - handle UTF8 strings passed to T1 low-level functions
        - handle flags for underline, strikethrough and overline for T1
          low-level output functions
        - OO interfaces to UTF8 and flags, for now leaving the flags as
          specific to Imager::Font::Type1
        - mc_web_map was storing colors with an alpha of 0
        - update the URLs for libpng and zlib
        - Fixed empty string handling for tt font driver.
        - Fixed tiff handling of images all contained in a 
          single strip (rowsperstrip = -1) and added a hack
          for images missing the rowsperstrip tag.
        - Fixed default parameters for crop()
        - Added Preliminary specialized scaling code.
        - Added image type detection.
        - added smoke test for win32 font bounding_box() method, and
          fixed the problem reported in ticket #1748.
        - update t37w32font.t to use i_line() instead of i_draw()
        - replaced non-portable call to rint() with (int)(dy+0.5) (so we
          can build on Win32)
        - the default tifflib warning handler was crashing on Win32
          under ActivePerl, when an unrecognized tag was read from the 
          file.  For now we'll just drop the warnings in the bit bucket.
          (Ticket #1513)
        - the code to read multiple tiffs didn't handle files with more
          than five images correctly, causing a memory overrun.
        - fix some minor test code hiccups
        - implemented i_t1_has_chars(), tests for same
        - added ExistenceTest.{pfb,afm,ttf} for testing $font->has_chars
        - tests for Imager::Font::Type1::has_chars();
        - tests for Imager::Font::Truetype::has_chars();
        - internal and external bounding box calculations now use
          the same hint flags as text output for Freetype 2.x
        - made the i_foo_bbox() interface more expandable by using
          symbolic constants for the sizes and array offsets
        - added a / character to the ExistenceTest.foo fonts that 
          overlaps the right side of the character cell, to test the 
          advance width reporting.
        - added advance width to the i_foo_bbox() interface, and 
          implemented it for FT2, FT1 and Type 1
        - Imager::Font::bounding_box() now returns an Imager::Font::BBox
          object in scalar context.
        - implemented $font->align() text output method, for simple output
          of aligned text
        - created Imager::Font::Wrap::wrap_text to perform simple text 
          wrapping
        - FT1, FT2 and T1 fonts now support the face_name method
        - FT1, FT2 and T1 now support the glyph_names() method
        - Debian woody supplied FT2.0.9, which didn't support 
          FT_Has_PS_Names(), so we use the FT_HAS_GLYPH_NAMES() macro 
          instead.
        - some older FT1 don't define TT_MS_LANGID_ENGLISH_GENERAL,
          which we use, define it if freetype doesn't.
        - Added extra options to rubthrough() so only a subpart of
          source image is used.
        - the image fills didn't handle filling with source images of
          less than four channels correctly
        - added comment support the postfix transform2() expression
          parser
        - transform2() can now produce images with other than 3 channels.
        - added a correct T_AVREF input mapping to the typemap to 
          simplify parameter lists
	- shut off one of the align subtests in the ft2 test.  The triggering
	  is likely an ft2 bug.
        - removed some half written scaling optimization code.
	- added /sw/lib and /sw/include to the Makefile.PL code for osX.
	- removed all references to the www.eecs.umich.edu page and changed
	  them to imager.perl.org ones.

0.43
	- added log() and exp() functions to transform2()
	- change the getpN() functions in transform2() to set a 
	  reasonable alpha if the input image has no alpha
	- document the constants that transform2() defines
	- skip the right number of tests when FT2 isn't available
	- This version pushed to CPAN because of skip problem in FT2 test.

0.43_01 Fri 8 Oct 2004
	- only call FT_Get_Postscript_Name() on FT 2.0.6 and later
	- put the IM_LIBPATH and IM_INCPATH values first in the search
	  path so the builder gets their local versions if desired rather
	  than the system versions they might be trying to avoid
	- document the exp() and log() transform2() functions
	- document the constants normally set by transform2().
	- refer the user to appropriate documents in the example in 
	  Imager.pm
	- change the list of documents in Imager.pm to move the document 
	  names out of the =item lines so we can make them into links
	- the scale() method now produces a warning when called in
	  void context.
	- font.c now only uses the defined T1Lib error codes
	- update ppport.h and remove the duplicate definitions from
	  Imager.xs.  Had to mangle ppport.h to prevent duplicate global
	  function definitions.
	- newer versions of tifflib require that all of the function
	  pointers passed to TIFFClientOpen be non-NULL, the mmap() and
	  munmap() pointers were always NULL and the sizeproc was
	  sometimes NULL.
        - there would be a link or load time error on m_init_log() if
          Imager was built with IM_NOLOG, fixed by renamed init_log()
          to m_init_log() in log.c (thanks to Takumi Yamani)
        - moved some variable definitions to the right place
        - the Win32 font driver bounding box function now returns the 
          right number of values (both thanks to Takumi Yamani)
        - the Win32 font driver now uses DEFAULT_CHARSET rather than
          ANSI_CHARSET are the lfCharSet value for the LOGFONT,
          as suggested by Takumi Yamani.
        - fontfiles/ExistenceTest.{pfb,ttf} weren't marked as
          binary in the CVS repository (caused test failures if you
          built from CVS on Win32)
        - Makefile.PL should now handle INCLUDE or LIB with spaces in them
          correctly on Win32.
        - the pnm reader read maxval for ppm/pgm files and then ignored it,
          it's now validated (0 < maxval < 65536) and used to scale
          samples.  Note that binary ppm/pgm files (P6/P5) with maxval >
          255 result in an error, since I didn't want to add new features
          just yet, just get the code that's there working correctly.
          Thanks to Elthek on rhizo for reporting this and help in 
          tracking it down.
          Resolves https://rt.cpan.org/Ticket/Display.html?id=7465
        - added a bunch of tests for reading pnm files.
        - previously, if you supplied to_paletted and empty color map
          and set make_colors to 'none', quant_translate() would segfault.
          This was because it was making the reasonable assumption that
          you'd have colors to map to.  quant_translate() now checks there
          is at least one color and return NULL if there isn't.
        - i_img_to_pal() now does error checking of the value returned by
          quant_translate().
        - Imager::to_paletted() now checks for success/failure of 
          i_img_to_pal() correctly and does appropriate error handling.
        - i_writegif_low() did no error checking on the result of
          quant_translate(), it now does
        - we now test that trying to write a GIF image with no palette
          allowable by the quant options is a failure.
        - Imager::write() was doing nothing with the result of the call 
          to i_writegif_gen(), in particular it wasn't returning () on 
          failure.
        - added tests for paletted image handling and the methods
          specific to those images
        - the XS for i_setcolors() was missing the OUTPUT clause for
          RETVAL, and hence wasn't returning failure on failure.
        - supplying a sufficiently small scaling value could make the
          scale() method return an image with zero height or width.
          Several of the above together resolve 
          https://rt.cpan.org/Ticket/Display.html?id=7467
        - the void context warning for scale() now includes the callers
          filename/line (instead of the default of Imager.pm line 15xx)
        - Imager->new will now return undef if the dimensions or number of
          channels specified for an image are out of range.  An error
          message can be retrieved with Imager->errstr.
        - added the C<builtin> color specifier and the 
          Imager::Color::Table class which defines those colors.
          Resolves https://rt.cpan.org/Ticket/Display.html?id=2593
        - added the equals() method to Imager::Color.
          Resolves https://rt.cpan.org/Ticket/Display.html?id=2238
        - prevent a test warning introduced by the above
        - the rotate() and matrix_transform() methods now accept a 'back'
          parameter specifying a background color.  This is only used
          where there is no source data (think of the corners around an
          image rotated 45 degrees) and it not combined with transparent
          pixels from the source image.
          Resolves https://rt.cpan.org/Ticket/Display.html?id=6140
        - removed a warning generated by the new matrix_transform() test
        - added a method index to Imager.pm
        - corrected "flood fill" to "flood_fill" in Imager/Draw.pod
        - removed compose() method from Imager/Transformations.pod since
          it isn't implemented yet
        - the image resulting from a crop is now the same type as the
          source image (paletted vs direct, bits/sample)
          Resolves https://rt.cpan.org/Ticket/Display.html?id=7578
        - the parameters to crop() weren't handled correctly in most 
          cases other than supplying left,top,right,bottom.
        - clarified the documentation for crop() providing more detail
          and more examples
        - the edges of the cropped area are now cropped against the 
          edges of the source image
        - cropping to zero width/height is treated as an error (no
          image is returned and $src->errstr has a message)
          Resolves https://rt.cpan.org/Ticket/Display.html?id=7581
        - built 0.43_01 for testing

0.43_02 Mon 26 Oct 2004

- the changes to scale() had some problems with integer vs floating point
  calculations (only caught in tests under perl 5.8.5 <sigh>)
- the FT2 glyph_names() method didn't do correct error handling
  when the string parameter wasn't supplied
- i_ft2_glyph_name() accepted only an unsigned char as the
  character code to get the name for, which meant it
  didn't work for unicode characters \x{100} or above
- the XS for i_ft2_glyph_name() had a similar problem
- added NameTest.ttf to be used in checking unicode glyph
  names
- added reliable_only optional parameter to the glyph_names()
  method so you can ignore theresult of FT_Has_PS_Glyph_Names()
- handle errors given by i_ft2_glyph_name() a bit more
  correctly
- the FT1 glyph_names() method didn't do correct error handling 
  when the string parameter wasn't supplied
- some memory allocated when creating a callback IO object (io_new_cb)
  wasn't being released (detected with valgrind)
- the testtools.pl match[nx]() functions escapes the test string on 
  test failure a bit better
- the XS code for i_tt_glyph_name() used unsigned char to store a 
  unicode character when it should have used unsigned long.
- the XS code for i_t1_glyph_name() used unsigned char to store a 
  unicode character when it should have used unsigned long.
- resolves https://rt.cpan.org/Ticket/Display.html?id=7949
- the type 1 glyph_names() method didn't do correct error handling 
  when the string parameter wasn't supplied
- renamed io.h to imio.h to prevent problems building under cygwin.
  resolve https://rt.cpan.org/Ticket/Display.html?id=7948
- i_writegif_low() wasn't setting the returned global palette, which
  meant a rubbish palette was returned to the user (detected with valgrind)
- built 0.43_02 for testing

0.43_03 Wed 8 Dec 2004

- change the "double-include" protection macro that imio.h uses.
- updated download locations for the various libraries that Imager 
  depends on.  Added some advice for cygwin.
- more information on gif library versions in README and Makefile.PL
- creating an image where the size of the allocated image buffer would
  overflow an integer would cause too small a buffer to be allocated.
  This could potentially be a security hole.
  partly resolves https://rt.cpan.org/Ticket/Display.html?id=8213
- set i_format to jpeg for jpeg files and test for it
- set i_format to png when reading png files and test for it
- i_yres was being set to the xres when reading a png file
- added many bad BMP files to test various code paths in bmp.c, and 
  made many minor fixes to bmp.c to make them work:
  - it was possible for various types of read failures to SEGV, both
    as NULL pointer dereferences and buffer overflows
  - some errors, like palettes being too large for the number of bits
    per pixel, were not being caught
  - failing to read all of a packed data structure would not cause
    a read failure
  - invalid compression types were not always caught
  - error messages are more consistent (still not always great messages,
    but one step at a time)
- added bmp_compression_name, bmp_used_colors, bmp_filesize, bmp_bit_count 
  tags on reading a BMP file
- added tools/imager to the distribution.  This is still very 
  experimental and untested.  Patches welcome, if you write tests to go
  with them.
- the BMP reader now validates the bfOffBits value from the BMP header
  and skips to that offset before reading image data.  Previously this
  value was read but otherwise ignored.
- added --palette action to tools/imager
- i_img_pal_new() now releases the image object memory if creation
  fails.
- set i_format to gif when reading gif files and test for it
- set i_format to pnm when reading pnm files and test for it
- set i_format to tga when reading tga files and test for it
- handle image creation failures when reading tga files
- set i_format to tiff when reading tiff images and test for it
- handle image creation failures when reading tiff files
- test t/t35ttfont.t no longer requires TTFONTTEST to be set (and 
  hasn't for a long time,) removed that requirement from README
- updated home page URLs throughout the .pods
- added information on reporting bugs to the SUPPORT section of Imager.pm
- regops.perl now sorts the dumped data structure to produce minimal diffs
- quant.c now checks for integer overflow allocating its image data
- i_readraw_wiol() now checks for image creation failure
- i_readrgb_wiol() now checks for image creation failure
- i_writergb_wiol() was an empty stub, it now pushes an error message
  and explicitly returns failure.
- i_readrgb_wiol() now sets i_format to rgb.
- set i_format to raw when reading tga files and test for it
- document i_format tag
- some tests were using $Config{ivsize} when they should have been 
  using $Config{intsize}
  Resolves: https://rt.cpan.org/Ticket/Display.html?id=8645
- tools/imager has been removed from the MANIFEST, it's way too late
  to test/debug for 0.44.
- image.h had no prototype for i_ft2_bbox_r() and it was being called
  from Imager.xs
- giflib 4.1.3 still doesn't have all the bugs fixed, update the notes
  in README and in the buggy_giflib.txt file that t105gif.t produces
- make the inclusion of NO_META dependent on the version of EU::MM.
  I was going to include a META.yml but EU::MM's generation is too 
  simplistic (and misses the leading document header too).  For now
  I'll leave it out.
- built 0.43_03 for testing

0.44 Wed 15 Dec 2004
- modify the default include directories list to include the location
  FreeBSD keeps freetype1 headers after the freetype2 locations.
  This ensures that the -I options are generated with freetype2
  locations first, so that those directories are searched before
  the freetype1 directories.  This prevents problems compiling
  font.c on FreeBSD.
  Resolves: http://rt-cpan.fsck.com/Ticket/Display.html?id=1723
- finish off a sentence in the "Producing an image from a CGI script"
  example in Imager::Files
- method index didn't include errstr()
- document that you don't want the FT2 freetype.h in the include path

=================================================================

        For latest versions check the Imager-devel pages:
        http://imager.perl.org/

=================================================================