- method index didn't include errstr()
- document that you don't want the FT2 freetype.h in the include path
-0.44_01
+0.44_01 Tue 24 May 2005
- the plugins Makefile has long produced an error after all tests were
complete on Win32, finally tracked down why
- Makefile.PL now checks the directories specified by $Config{locincpth}
and right bearing values.
- Imager::Matrix2d->rotate() would only rotate around the supplied
centre point if both 'x' and 'y' were non-zero.
+- the internal i_tags_get_string() function now acts correctly for
+ integer only tags.
+- the FT1.x driver now supports the align parameter correctly.
+ Tests were added to each driver to check correct handling of the align
+ parameter.
+- the Win32 font driver bounding_box() method now puts accurate values
+ in the ascent and descent values, previously they were set to the
+ font ascent/descent rather than the values specific to the string.
+- supplying align=>0 to the win32 font driver string() function
+ now aligns in the same way as the other drivers.
+- removed the bug reporting email address to prevent spammers stripping
+ it. The URL is still there and if someone knows how rt.cpan.org works
+ they can still figure out the email.
+- the internal function used to probe file formats if the type
+ parameter wasn't passed to the read() method would produce
+ diagnostics on stdout, precluding use in web applications. Silenced
+ it.
+- the setcolors() had a fencepost error making it impossible to
+ use it to set the last color in the palette.
+ http://rt.cpan.org/NoAuth/Bug.html?id=12676
+- fix the filter examples in Imager::Filter so they don't indicate that
+ a new image is returned (the filter() method works in-place)
+- add examples for the fountain filter
+- prevent warnings from Makefile.PL when there's no libraries found
+- Imager no longer assumes that MSVC is the only native Win32
+ compiler, MinGW can be used to build Imager.
+- fixed an inverted condition checking the giflib version, and
+ replaced a manual test skip in t/t105gif.t
+- add tests to exercise hatch fills for floating images. This code is
+ mis-generated with the gcc 4.0 that ships with OS X 10.4
+- work around junk that pkg-config sends to stderr on FreeBSD 4.10
+- testimg/base.jpg was missing from the MANIFEST, causing test failures
+- use a temp file to avoid messing with echo on Win32 when building
+ META.yml
+- add some Mac OS X notes to the README
+
+0.45 Mon 30 May 2005
+- t/t105gif.t wasn't handling buggy giflibs correctly
+
+0.45_01
+- give the colorcount() and maxcolors() methods their own entries and
+ add them to the method index.
+- added tiff_bitspersample and tiff_photometric tags to tiff images
+ when read
+- loading filter DLLs/DSOs had an off-by-one error allocating a buffer
+ for the filename of the library (does anyone use this functionality?)
+- remove old #if 0 code from Imager.xs
+- convert t/t15color.t to Test::More
+- prevent warnings when looking up a GIMP color with no palette
+ parameter and $ENV{HOME} not set.
+- added VERSION numbers to most .pms
+- convert t/t104ppm.t to Test::More
+- convert t/t107bmp.t to Test::More
+- convert t/t108tga.t to Test::More
+- error messages generated reading JPEG or PNG images are now
+ available via errstr()
+- implement/test/document set_file_limits()/get_file_limits() methods,
+ which allow limits on the size of image files read.
+- new example for convert() method based on Leolo's query
=================================================================
projects / imager.git / blobdiff