1 Imager release history. Older releases can be found in Changes.old
3 - update 1.002 release notes to include the center change for filled
6 - flood_fill() would escape beyond a 4-connected space under some
8 Added many more flood_fill() tests.
9 https://rt.cpan.org/Ticket/Display.html?id=103786
11 Imager 1.002 - 3 Apr 2015
14 - drawing anti-aliased filled circles is now 10 to 50 times faster
15 depending on the size of the circle.
16 This also changed the center from being the center of the pixel to
17 being the top left of the pixel to match the filled arcs drawn by
19 https://rt.cpan.org/Ticket/Display.html?id=101682
21 - enhancements to polygon filling:
23 - added a mode parameter to control how overlapping regions behave
25 - added a polypolygon() method to fill more than one polygon at a
28 - polygon filling is now exposed through the API.
30 - added colormodel(), alphachannel() and colorchannels() methods.
31 These were added for two reasons:
33 - a future version of Imager may allow the number of channels in
34 an image to not directly represent the color model of an image.
35 eg. a greyscale TIFF image with multiple alpha channels.
37 - a future version of Imager may allow an image to be read without
38 translation, for example a TIFF file that contains measurements
39 from an instrument. Currently Imager transforms the samples into
40 the range 0.0 ... 1.0 which may means the user has to translates
43 An untranslated image would be unusable as image data, so
44 colormodel() would return "unknown" in this case.
46 Similarly a CMYK image might be returned as an "unknown" color
47 model image, if the caller chooses to disable translation to
50 Imager 1.001 - 2 Jan 2015
53 - both Imager and perl 5.21.3 define my_strerror(), prevent a conflict
54 Thanks to Slaven Rezic for the report and the patch.
55 https://rt.cpan.org/Public/Bug/Display.html?id=98234
57 - GIF: clean-up a test file if the test for the giflib 4.2.0 file
60 - fix Imager::Matrix2d::rotate()'s centre point handling
61 https://rt.cpan.org/Public/Bug/Display.html?id=99959
63 - ICO: don't apply the icon mask to images with an alpha channel by
65 https://rt.cpan.org/Public/Bug/Display.html?id=99507
67 - make verbose probing output more verbose
69 - use Imager::Probe to probe for freetype 1.x
70 https://rt.cpan.org/Ticket/Display.html?id=100502
72 - The --enable and --disable parameters to the top-level Makefile.PL
75 - update the bundled/modified Devel::CheckLib to handle the gcc-4
78 Imager 1.000 - 28 Jul 2014
81 There's nothing special about Imager 1.000, it's just another release.
83 - fix the skip check for the iolayer qr// buffer test
85 - improve error reporting for the iolayer test failing on a small
86 number of Win32 CPAN testers
88 Imager 0.99_02 - 21 Jul 2014
91 - Imager::Filter::Mandelbrot (and dynfilt/mandelbrot.c) - initialize
92 the blue channel in the generated palette, and allow each color
93 component to be in the range 100..255 instead of just 100..254.
94 Thanks to Ralf Neubauer.
95 https://rt.cpan.org/Ticket/Display.html?id=97086
97 - revert "improved the XS for i_io_read() and i_io_raw_read()"
98 This caused problems with older perls and didn't provide much of a
99 performance improvement.
101 - support Inline 0.57 and later.
102 Inline 0.57 changed the "with" interface.
103 https://rt.cpan.org/Ticket/Display.html?id=97108
105 - don't define our own MAXINT macro, it conflicts with windows header
106 files and in a few places it was the wrong value to use anyway.
107 https://rt.cpan.org/Ticket/Display.html?id=96425
109 Imager 0.99_01 - 29 Jun 2014
112 - GIF: add support for giflib 5.1.0, which added error code pointer
113 parameters to EGifCloseFile() and DGifCloseFile().
114 https://rt.cpan.org/Ticket/Display.html?id=96756
116 - GIF: avoid a double-free when do_write() fails.
118 - fix SV type probing to work on perl before 5.12. Broken in 0.99.
119 https://rt.cpan.org/Ticket/Display.html?id=96761
121 - PNG: skip the benign error test before libpng 1.6.0, since the
122 error we're testing didn't exist before 1.6.0.
123 https://rt.cpan.org/Ticket/Display.html?id=94717 (continued)
125 Imager 0.99 - 25 Jun 2014
128 - Imager::IO->new_buffer() (and hence Imager->read()'s data
129 parameter) now accepts a reference to a scalar as well as just a
131 https://rt.cpan.org/Ticket/Display.html?id=92785
133 - Imager::IO->new_buffer() now always makes a copy of the passed in
134 buffer to avoid problems with temporary objects used for the return
136 https://rt.cpan.org/Ticket/Display.html?id=92785
138 - improved the XS for i_io_read() and i_io_raw_read()
139 https://rt.cpan.org/Ticket/Display.html?id=92738
141 - load plugins from absolute paths on Android
142 Thanks to Brian Fraser.
143 https://rt.cpan.org/Ticket/Display.html?id=93272
145 - added the jpeg_optimize parameter for writing JPEG files. This can
146 significantly reduce file sizes, but uses more memory and time.
147 https://rt.cpan.org/Ticket/Display.html?id=94292
149 - the autolevels filter now works on the luminosity of the image
150 rather then working per channel. The old autolevels filter is
151 still available as "autolevels_skew".
152 https://rt.cpan.org/Ticket/Display.html?id=94413
154 - Imager::File::PNG now supports libpng 1.6.10.
155 1.6.10 changed CRC errors from benign errors to normal errors,
156 which broke the test which used CRC errors to check for benign
157 error support. Switched to using a 1-bit grey-scale image with a
158 palette to test for benign errors.
159 https://rt.cpan.org/Ticket/Display.html?id=94717
161 Imager 0.98 - 3 Jan 2014
164 Incompatible changes:
166 - the return value of setpixel() has changed.
168 Previously the return values for undrawable pixels vs caller errors
169 changed depending on whether you used the multiple pixel calling
170 mechanism, or the single pixel mechanism.
174 - for an invalid parameter, such as an unknown colour, or missing
175 parameter, an empty list (or undef in scalar context) is
176 returned, and errstr() is set,
178 - otherwise the number of pixels drawn is returned. If none of
179 the pixels could be drawn (they were all outside the image), "0
180 but true" is returned.
182 https://rt.cpan.org/Ticket/Display.html?id=87650
186 - when drawing on an image with an alpha channel where the source
187 minimum is greater than zero, Imager would read from beyond the end
188 of a malloc() allocated buffer. In rare circumstances this could
189 lead to some of the source image not being written to the target
190 image, or possibly to a segmentation fault.
191 I don't believe this has any security concerns beyond that.
193 - if the first text drawn with Imager::Font::T1 is not anti-aliased,
194 text drawn would be nonsense. This would also read beyond the end
195 of a malloced buffer.
197 - non-AA text drawn with an Imager::Font::TT (Truetype 1) font would be
199 https://rt.cpan.org/Ticket/Display.html?id=88993
201 - Imager::Font::Wrap no longer requires the image parameter.
202 https://rt.cpan.org/Ticket/Display.html?id=87338
204 - a documentation typo fix in Imager::Transformations
205 Thanks to Adrian Yee.
206 https://rt.cpan.org/Ticket/Display.html?id=88598
208 - Image::Probe, Imager's internal module for probing library
209 locations now searches the directories specified by LD_RUN_PATH,
210 LD_LIBRARY_PATH, DYLD_LIBRARY_PATH and LIBRARY_PATH for libraries,
211 and the corresponing s/\blib$/include/ directories for header
213 https://rt.cpan.org/Ticket/Display.html?id=86428
215 Imager 0.97 - 15 Jul 2013
218 No changes from 0.96_02.
220 Imager 0.96_02 - 8 Jul 2013
223 - PNG: treat a version mismatch between headers and library as a
225 https://rt.cpan.org/Ticket/Display.html?id=86659
227 - PNG: the check for benign error support is more complex than a
228 simple library version check, check for the appropriate macro.
229 Thanks for Slaven Rezic for following up on my CPAN Testers result
231 https://rt.cpan.org/Ticket/Display.html?id=86659
233 - W32: add a missing head1 AUTHORS to the old Win32.pm
234 https://rt.cpan.org/Ticket/Display.html?id=86658
236 Imager 0.96_01 - 1 Jul 2013
239 - TIFF: handle SampleFormat = 2 by translating the signed integer
240 values to unsigned by flipping their sign bits.
242 Handle SampleFormat = 3 where possible.
244 SampleFormat is ignored for paletted images.
246 Mixed SampleFormat are handled incorrectly, since libtiff returns
247 only the first SampleFormat value, and an image with both an alpha
248 channel and SampleFormat = 2 for color channels probably has a
249 different SampleFormat for the alpha channel.
251 https://rt.cpan.org/Ticket/Display.html?id=74540
254 https://rt.cpan.org/Ticket/Display.html?id=69243
256 - Imager::Color's rgba() method now returns its values as integers
257 instead of floating point. (IV instead of NV).
259 - The XS for i_map() and i_matrix_transform() now use the AV *
260 typemap instead of rolling their own.
262 - The XS for DSO_call() now uses the HV * typemap instead of
265 - The XS for i_poly_aa(), i_poly_aa_cfill(), i_transform() and
266 i_gradgen() now use a new T_AVARRAY typemap that greatly
267 simplifies the XS code.
269 - many other minor XS changes
271 - some XS code formatting
273 - TT (Freetype 1.x) and FT2 (Freetype 2.x) non-antialiased rendering
274 now renders in alpha-combining mode, to match antialiased output.
275 https://rt.cpan.org/Ticket/Display.html?id=73359
277 - add sample code and cookbook notes for drawing a blurred drop-shadow.
279 - add support for libpng 1.6, which changed the defaults for
280 reporting some types of error.
281 https://rt.cpan.org/Ticket/Display.html?id=85746
283 - FT2: use gsamp/psamp() to transfer pixels from the work image to
284 the output image instead of gpix/ppix.
285 https://rt.cpan.org/Ticket/Display.html?id=83478
287 - eliminate various GCC warnings, mostly initialized but otherwise
290 Imager 0.96 - 19 May 2013
293 - rearrange Imager's test files.
294 https://rt.cpan.org/Ticket/Display.html?id=84596
296 - Imager::Probe::_gcc_lib_paths now forces C locale.
297 A sufficiently recent gcc prints localized keys for the search
298 paths, which avoids that localization.
299 https://rt.cpan.org/Ticket/Display.html?id=84963
301 - fix a pod error detected by Pod::Simple 3.27 and laterm and skip
302 pod tests unless AUTOMATED_TESTING or IMAGER_AUTHOR_TESTING is set
304 https://rt.cpan.org/Ticket/Display.html?id=85262
305 reported by Andreas König
307 - add a test for unclosed POD in C source, and fix the errors it
309 https://rt.cpan.org/Ticket/Display.html?id=74875
311 - fix some spelling errors detected by the newer aspell in Debian
314 - remove a trailing ' from lib/Imager/Font/Test.pm's POD
315 thanks to gregor herrmann <gregoa@debian.org>
316 https://rt.cpan.org/Ticket/Display.html?id=85413
318 - add a test for standard =head1 commands and fix the errors found
320 Imager 0.95 - 19 Apr 2013
323 - handle the SVf_UTF8 flag correctly on magic strings passed to
324 Imager::IO's write() and raw_write() methods.
325 This may misbehave in perl 5.6.x since the UTF8 flag may not be
326 visible in the main SV after SvGETMAGIC().
327 https://rt.cpan.org/Ticket/Display.html?id=83438
329 - document that bounding_box() ignores the transformation matrix
330 supplied to tranform() for fonts.
331 https://rt.cpan.org/Ticket/Display.html?id=84106
333 Imager 0.94_02 - 5 Apr 2013
336 - enable debug logging for the standard font tests
338 - skip the overloaded UTF-8 text output checks on 5.6.x and earlier,
339 due to a bug in that version of perl.
341 - don't test for read() failing on a write-only handle in 5.8.x and
342 earlier, this was fixed in change 862083f7e4 in perl.
344 - report the version of Inline found during testing (an attempt to
345 diagnose some intermittent failures.)
347 Imager 0.94_01 - 2 Mar 2013
350 - NOTE: possibly backward incompatible:
351 support reading from/writing to perl filehandes that aren't raw
354 This allows Imager's I/O to honour handles with layers such as
355 gzip, scalar file handles or tied file handles.
357 This is backward incompatible in that previous Imager would simply
358 use fileno() to retrieve the fd for the file and call write(2) etc
360 https://rt.cpan.org/Ticket/Display.html?id=78843
362 - moved most of README to lib/Imager/Install.pod which should make it
363 more accessible to the "web" generation, also significantly updated
366 - updated README's for the separately distributed modules to refer to
367 Imager::Install, and that they need -dev versions of packages
368 https://rt.cpan.org/Ticket/Display.html?id=81265
370 - the JPEG test code now reports the compile-time library version
372 - avoid a possible compiler optimization issue on CentOS 5.9 i386 by
373 rearranging (and mildly optimizing) some code.
374 https://rt.cpan.org/Ticket/Display.html?id=83212
376 - fix a POD error in Imager::Fill (detected on new Pod-Simple)
377 https://rt.cpan.org/Ticket/Display.html?id=83434
379 - fix a broken link to Graphics::Magick
380 https://rt.cpan.org/Ticket/Display.html?id=82743
382 - drawing text to a channel with FT2 would draw with random coverage
383 due to an uninitialized alpha channel.
385 - marked the function pointer underlying the mm_log() API with the
386 correct gcc magic and fixed the resulting warnings.
388 - fixed some other compiler warnings
390 - Imager::Font::W32 now properly reports text drawing errors
391 https://rt.cpan.org/Ticket/Display.html?id=70098
393 - handle the SVf_UTF8 flag correctly on magic (eg. overloaded)
394 strings passed as text to draw(), bounding_box(), has_chars(),
395 glyph_names() (where supported) in each of the font drivers.
396 This may misbehave in perl 5.6.x since the UTF8 flag may not be
397 visible in the main SV after SvGETMAGIC().
398 https://rt.cpan.org/Ticket/Display.html?id=83438 (partial)
400 Imager 0.94 - 15 Dec 2012
401 =========================
403 Variations on some of these changes were included in development
406 - improved thread safety
407 - the internal error stack and log file handle are now in a per-thread
409 - JPEG now captures IPTC information in a thread-safe way
410 - avoid globals where possible for warning capture in libtiff
411 - use a mutex to avoid re-entering thread-unsafe giflib
412 - use a mutex to avoid re-entering thread-unsafe tifflib
413 - use a mutex to avoid re-entering thread-unsafe T1Lib
414 - use a library handle per thread for freetype 2.
415 - use an engine handle per thread for freetype 1.x.
416 - originally these changes broke ABI compatibility, this has been
419 - clarify the return value of getpixel();
420 https://rt.cpan.org/Ticket/Display.html?id=81198
422 - fixed a race condition in parallel testing for T1
424 - fixed a bug in handling yoff for untransformed image-based fills
426 - documentation improvements for Imager::Fill
428 - FT2: report the library version while testing.
430 Imager 0.93 - 15 Oct 2012
435 - previously a transparency enabled GIF write (the default) would
436 always use a GIF89a header even if none of the images were
439 - previously the probe step for freetype-config would fail on cygwin
441 - catch an invalid matrix parameter to convert() instead of crashing
442 https://rt.cpan.org/Ticket/Display.html?id=79922
444 - remove the 16-bit/sample limitation from the documentation for
445 setsamples(), it hasn't applied for many releases.
446 https://rt.cpan.org/Ticket/Display.html?id=79989
448 - don't copy setsamples() data parameter, it may be a large scalar
449 https://rt.cpan.org/Ticket/Display.html?id=79990
451 - clean up .dSYM directories generated performing probes on OS X
454 - pass the --verbose command-line option through to Imager::Probe in
455 sub-module's Makefile.PLs
456 https://rt.cpan.org/Ticket/Display.html?id=75878
460 - support for giflib 5.0. (5.0.1 and later)
461 https://rt.cpan.org/Ticket/Display.html?id=79029
462 The giflib API Imager uses doesn't have a mechanism to set the
463 header version in 5.0.0.
465 - update the GIF library probe code to check for the new giflib 5.0.1
466 EGifSetGifVersion() function, and to check for the giflib 4.2.0
467 uninitialized gif89 bug.
468 https://rt.cpan.org/Ticket/Display.html?id=79679
469 http://sourceforge.net/tracker/?func=detail&aid=3574283&group_id=102202&atid=631304
471 - avoid static variables when capturing IPTC data from JPEG files
473 - match Imager::Font;:T1's error message translations to those from
474 later versions of T1Lib.
476 - for libtiff versions that support extended warning handlers (3.8.0
477 or later), use them to avoid some global variables.
479 - note Image::ExifTool for better metadata support.
480 https://rt.cpan.org/Ticket/Display.html?id=79251
482 Imager 0.92 - 14 Aug 2012
485 - giflib 4.2 eliminates the GIF_LIB_VERSION macro, handle that
486 correctly for both probing and runtime.
487 https://rt.cpan.org/Ticket/Display.html?id=77672
489 - allow building JPEG/imexif.c on C89 compilers.
491 - allow building GIF/imgif.c on C89 compilers.
493 Imager 0.91 - 4 Jun 2012
498 - The size of rotated images is no longer rounded up so aggressively.
499 Added rounding to the linear interpolation done for rotate() and
500 matrix_transform() for 1 and 3 channel 8-bit images.
501 Adjusted the two tranlate matrices used to build the final rotation
502 matrix to use the distance between the outlier pixels rather than
503 the pixel dimensions (ie. dimension-1).
504 With all of this the output of rotate(degrees => 270) on an 8-bit
505 image exactly matches the output of rotate(right => 270).
506 https://rt.cpan.org/Ticket/Display.html?id=77063
510 - eliminate the old IIM_new(), IIM_DESTROY() names from Imager's
511 internals, those names only matter for the XS interface.
513 - improve error reporting when PERL_INITIALIZE_IMAGER_CALLBACKS finds
514 the API level compiled into a loadable module such as
515 Imager::File::GIF doesn't match that of Imager. Previously it
516 could be difficult to determine exactly which module was failing to
518 https://rt.cpan.org/Ticket/Display.html?id=77173
520 - added Imager->check_file_limits() as an interface to the
521 i_int_check_image_file_limits() API.
523 - Imager->set_file_limits(reset => 1) now resets the limits to the
524 defaults rather than setting them all to unlimited.
526 - wrote a brief security note as Imager::Security
527 https://rt.cpan.org/Ticket/Display.html?id=68910
529 Imager 0.90 - 30 Apr 2012
534 - Imager::Test::is_imaged() attempted to process an "epsilon"
535 parameter but the prototype didn't allow for the extra parameter.
536 Corrected the prototype.
538 - when downconverting samples (eg. from floating point to 8 bit)
539 Imager now rounds the sample value rather than attempting to
540 allocate input values over output values equally. The old
541 behaviour had the probably surprising effect of converting a
542 floating point 2.1/255.0 into an 8-bit 3 rather than the expected
544 This may result in slightly different 8 or 16-bit output images.
546 - BI_BITFIELD BMP images were handled incorrectly, both in
547 incorrectly calculating the space required for the masks and in
548 processing the image data itself.
549 https://rt.cpan.org/Ticket/Display.html?id=76736
551 - some odd width BMP BI_RLE4 images use run lengths that run one off
552 the edge of the image. Imager now allows this, discarding the
555 - odd length RLE4 runs in BMP images were decoded incorrectly.
557 - pkg-config could sometimes return a library that was in a directory
558 EU::MM / $Config{libpth} doesn't know about, but the compiler did.
559 If no -L is included in the pkg-config library information check if
560 EU::MM can find the library, and if not, search our configured
561 directories and insert that into the library flags.
562 https://rt.cpan.org/Ticket/Display.html?id=75869
564 - Imager::Probe can now probe for libraries with dependent libraries,
565 common for static linking, eg. libpng requires libz.
566 https://rt.cpan.org/Ticket/Display.html?id=74043
568 - libpng 1.5 specific probes were looking for libpng 1.4 filenames.
570 - added alternative probe configurations that try to link libz, to
571 handle a statically linked libpng.
572 https://rt.cpan.org/Ticket/Display.html?id=74043
574 - if a probe includes testcode, Imager::Probe now checks that code as
575 part of the process of checking each configuration rather than as a
576 post test of the found configuration. This allows alternate
577 configurations to be checked if a matching but non-working
578 configuration is found.
579 https://rt.cpan.org/Ticket/Display.html?id=74043
583 - when reading or writing images via callbacks, the default callbacks
584 now push an error message indicating that a required callback was
585 called but not supplied.
586 https://rt.cpan.org/Ticket/Display.html?id=76782
588 - clarify which callbacks are required for reading and writing TIFF
591 - improve logging for creation of callback I/O layers.
593 - a little more documentation for Imager::Probe.
595 - the i_get_file_background() and i_get_file_backgroundf() APIs now
596 return int to indicate whether the i_background tag was found.
599 - improve error reporting
600 - add png_interlace, png_bits tags
601 - read paletted images as paletted images, including transparency
602 - read 1 bit greyscale images as a type suitable for other file
603 handlers to write as bilevel
604 - read 16 bit/sample PNG as 16-bit/sample Imager images
605 - write "bilevel" paletted images as 1 bit grayscale images
606 - write paletted images as paletted images
607 - write 16-bit (or higher)/sample images as 16-bit/sample PNG
609 - improved metadata support
610 https://rt.cpan.org/Ticket/Display.html?id=29268
612 Imager 0.89 - 18 Mar 2012
617 - getpixel(..., type => "float") and the API i_gpixf() have been
618 broken on paletted images since they were implemented.
619 https://rt.cpan.org/Ticket/Display.html?id=75258
623 - links in the METHOD INDEX now point at the method documentation
624 rather than the heading you can find them under.
625 https://rt.cpan.org/Ticket/Display.html?id=71495
627 - Imager (and the bundled dynamic modules) no longer fallback to
628 using DynaLoader if loading via XSLoader fails.
629 For the bundled modules this could hide useful error messages.
630 https://rt.cpan.org/Ticket/Display.html?id=75560
632 - IM_DEBUG_MALLOC mymalloc() no longer sn?printfs() a string to a
633 buffer in the array of allocations, but just stores the filename
634 pointer and line number.
635 https://rt.cpan.org/Ticket/Display.html?id=70388
637 Imager 0.88 - 22 Feb 2012
640 - describe how to build libgif etc on OS X in such a way as to be
641 compatible with a fat binary perl (such as the system perl), in
643 https://rt.cpan.org/Ticket/Display.html?id=73371
645 - update the change notes for 0.77_01 to indicate that libungif
648 - add some other imaging modules to SEE ALSO
649 https://rt.cpan.org/Ticket/Display.html?id=73906
651 - note that the generator of the apparently non-DFSG-free postscript
652 in MMOne.pfb is a Debian package.
654 - setsamples() is now a true complement to getsamples() - it can
655 write 8-bit or floating point samples from a scalar or array
656 reference. This adds i_psamp() and i_psampf() to the C level API.
658 - the XS interfaces to i_gsamp(), i_gsampf() and i_gsamp_bits() have
659 changed to make better use of the typemap, but these aren't part of
660 the perl level API anyway. There were no changes to the C level
661 interfaces to these functions.
663 - getpixel() and setpixel() now accept a mix of scalar and array
664 references for the x and y parameters, and unequal array lengths is
666 https://rt.cpan.org/Ticket/Display.html?id=73697
670 - correctly calculate the size of a rotated image
671 https://rt.cpan.org/Ticket/Display.html?id=69261
673 - fix incorrect rounding of color values for matrix_transform() and
675 https://rt.cpan.org/Ticket/Display.html?id=69261
677 - Win32 text output is now done in normal combine mode, the alpha
678 component of the color is now significant.
679 https://rt.cpan.org/Ticket/Display.html?id=70014
681 - remove long unused gif case from read()
682 https://rt.cpan.org/Ticket/Display.html?id=69244 (partial)
684 - the getsamples() target parameter was being treated as a hashref
685 when it's meant to be an array reference.
686 https://rt.cpan.org/Ticket/Display.html?id=74882
688 - getpixel() and setpixel() now returns an empty list when invalid
689 parameters are supplied.
690 Invalid values for type now result in an error for getpixel().
691 https://rt.cpan.org/Ticket/Display.html?id=73697
693 Imager 0.87 - 03 Jan 2012
696 - document the return value of the filter() method.
697 https://rt.cpan.org/Ticket/Display.html?id=72369
699 - document i_gsamp_bits() and i_psamp_bits().
700 https://rt.cpan.org/Ticket/Display.html?id=68815
702 - properly increment the Imager::Matrix2d $VERSION.
704 - actually include the Imager::Test tests in the dist
706 - correctly read and write 256x256 pixel ICO files
707 https://rt.cpan.org/Ticket/Display.html?id=69599
709 - make the error message from read() or read_multi() when they can't
710 identify the file type match reality.
711 https://rt.cpan.org/Ticket/Display.html?id=72475
713 - read() now uses $FORMATGUESS if it can't determine the file type
714 based on the file header, to match read_multi().
716 - re-work and add tests for def_guess_type(). def_guess_type() no
717 longer returns any random file extension as the file type.
719 - add gray4, gray16 and gray as presets for make_colors.
720 https://rt.cpan.org/Ticket/Display.html?id=67911
722 - add make_palette() method that produces a palette from one or more
725 - fix the Imager dependency for the separately distributed font
727 https://rt.cpan.org/Ticket/Display.html?id=72643
729 - fix i_render_color() to properly draw in "normal" mode - ie. when
730 writing to a 1 or 3 channel image the second or fourth channel of
731 the source color was being ignored, it is now significant.
732 https://rt.cpan.org/Ticket/Display.html?id=71564
734 Imager 0.86 - 31 Oct 2011
737 - improve error reporting for W32 tests
739 Imager 0.85_02 - 24 Oct 2011
744 - eliminate unused i_gif_opts type (clean-up)
745 https://rt.cpan.org/Ticket/Display.html?id=69245
747 - fix combine=0 fill color anti-aliasing on the double/sample path
748 https://rt.cpan.org/Ticket/Display.html?id=71309
750 - make default text color non-transparent
751 https://rt.cpan.org/Ticket/Display.html?id=71469
753 - apply the last of the Debian unforwarded spelling fixes
754 https://rt.cpan.org/Ticket/Display.html?id=70656
756 - the log() method used its message parameter as a C level format
758 https://rt.cpan.org/Ticket/Display.html?id=71653
760 - provide our own STRLEN typemap entry for older perls.
761 https://rt.cpan.org/Ticket/Display.html?id=71641
763 - add extra ppport.h configuration to support older perls.
765 - depend on Scalar::Util, since we use it and older perls don't have
768 - add overloaded eq to Imager::Matrix2d, since older perls don't seem
769 to synthesize it from overloaded "".
771 - use T1_StrError() for error messages on modern libt1
772 https://rt.cpan.org/Ticket/Display.html?id=69879
774 - actually load the font rather than just adding it to the catalog on
777 - Imager::Font->new now produces better error messages for the T1
780 - the font has_chars() method now returns perl's true and false
781 values in list context rather than integers, which should be more
783 https://rt.cpan.org/Ticket/Display.html?id=69158
785 - the btm data structure used by the flood_fill code is now
786 initialized more efficiently.
787 https://rt.cpan.org/Ticket/Display.html?id=68994
789 - updated the Thanks list in README
790 https://rt.cpan.org/Ticket/Display.html?id=71607
792 - check there's at least one coefficient for the convolution filter
793 https://rt.cpan.org/Ticket/Display.html?id=68993
795 - make the APIRef synopsis ordering consistent, older versions of
796 perl could order it differently.
797 https://rt.cpan.org/Ticket/Display.html?id=71675
799 - we rely on Config.pm's d_vsnprintf as to whether we use
800 vsnprintf/snprintf, which is defined in the Win32 Config.pm even
801 though it only has _ prefixed versions of these. Define our own
802 prefixed names on Win32.
803 https://rt.cpan.org/Ticket/Display.html?id=71642
805 - fix library detection with MSVC
807 - search a few more library directories, so EU::MM doesn't discard
808 them. Hopefully fixes:
809 https://rt.cpan.org/Ticket/Display.html?id=71643
811 Imager 0.85_01 - 10 Oct 2011
814 - add simple tests for the Imager::Test test_image generators
816 - io_glue I/O buffering re-work:
818 - reorganize io_glue to do it's own buffering by default
820 - the unbuffered functions are available as i_io_raw_read() (or
821 raw_read() from perl) etc but are not recommended for typical
824 - use the new i_io_peekn() when checking for file magic to avoid
825 seek, allowing Imager to detect the file type and read the file
826 from an unseekable stream (for formats that don't use random
829 - added several new I/O layer API functions.
831 - fix the TGA performance problem, most noticably on Win32
832 https://rt.cpan.org/Ticket/Display.html?id=70037
834 - TIFF now uses wrapper functions of the correct types to avoid casts
835 https://rt.cpan.org/Ticket/Display.html?id=69912
837 - the callback IO object did its own buffering, controlled by the
838 maxbuffer parameter supplied to the read() and write() methods.
839 This buffering has been removed, to avoid redundancy with the
840 common io_glue buffering. This also avoids a bug in that code
841 which could rarely pass a zero length to the read callback and
842 then panic about the result.
844 - the callback IO object now tests the result of calling the close
845 callback, which should return true for success.
847 - the PNM reader did its own buffering. This buffering has been
848 removed to avoid redundancy with the common io_glue buffering.
850 - previously the read handlers for fd and callback I/O layers would
851 call the underlying interface (POSIX read or your supplied callback)
852 until it filled the buffer. It now only makes one call.
854 - added public constructors for I/O layer objects (see Imager::IO)
856 - all core file handlers now use the i_io_foo() wrappers to gain
857 access to buffered I/O rather than calling the I/O layer
860 - all core file handlers now check for error on close.
862 - Backward compatibility: if you hava custom file handlers, you can
863 use i_io_write() etc since they're available as macros in older
866 - eliminate the final remnants of io_glue_commit_types().
868 - bump IMAGER_API_VERSION, since the above may break assumptions.
870 - removed the long unused i_gen_reader() and i_gen_writer() utility
873 Imager 0.85 - 29 Aug 2011
876 The main changes in the release versus 0.84 are:
878 - on 64-bit systems, 64-bit types are consistently used for image
879 dimensions and co-ordinated, and for memory block sizes.
881 - handle IFD loops in TIFF files correctly. Previously this would
882 lead to an infinite loop.
886 - fix the link in the getheight() entry in the method index
888 Imager 0.84_02 - 22 Aug 2011
891 Development release, this will become 0.85 if CPAN testers is
896 - the image file limits set by set_file_limits() weren't being
897 checked when reading TIFF files.
898 https://rt.cpan.org/Ticket/Display.html?id=69915
900 - Provide more information about file format module load errors on a
901 failed image file read() or write().
902 https://rt.cpan.org/Ticket/Display.html?id=69194
904 - use TIFFReadDirectory() instead of TIFFSetDirectory() to iterate
905 through TIFF images, since it checks for IFD loops.
906 https://rt.cpan.org/Ticket/Display.html?id=69914
908 - don't leak memory when out of range palette indexes are supplied to
910 https://rt.cpan.org/Ticket/Display.html?id=69242
912 - require a later version of CPAN::Meta to ensure JSON::PP and
913 CPAN::Meta::YAML are available.
914 https://rt.cpan.org/Ticket/Display.html?id=69008
916 - hoist the per-line calculations for the flines implementations, and
917 modernize the tests a bit.
918 https://rt.cpan.org/Ticket/Display.html?id=70126
920 - detect snprintf()/vsnprintf() (cheat by using Config.pm) and use
922 https://rt.cpan.org/Ticket/Display.html?id=69147
924 - if t1lib failed to reinitialize it would be left as marked
926 https://rt.cpan.org/Ticket/Display.html?id=69877
928 - update the bundled (and still modified) Devel::CheckLib
929 https://rt.cpan.org/Ticket/Display.html?id=69170
931 Imager 0.84_01 - 8 Aug 2011
934 Development release as a sanity check for the types re-work.
936 Massive types re-work:
938 - the type used internally for pixel co-ordinates and image sizes is
939 now 64-bit on 64-bit platforms (at least sane ones).
941 - size_t is now used consistently for memory block sizes.
943 - since this changes the binary interface, the Imager API version has
944 been incremented. Any module that uses the API will need to be
945 rebuilt. In most cases that will be enough, but calls to any APIs
946 that take a pointer to image sizes may need source changes.
948 - you should be able to create very large images on 64-bit systems,
949 if you have enough memory. Successfully created a 32768 x 49192 x
950 3 channel image, filled with a tiled image and scaled it. The
951 unscaled image was also successfully saved to a JPEG.
953 - check the image size before attempting to write BMP, GIF, JPEG,
954 PNG, SGI, TGA, TIFF images.
956 - correctly handle reading TGA images over 32767 pixels wide or tall.
960 - the gif_left and gif_top tags are now clamped to non-negative
961 values when writing a GIF image.
963 - removed dead callback read/write code
965 The default maximum memory size when reading files is now 1G.
967 Imager 0.84 - 20 Jun 2011
970 - Imager no longer inherits from Exporter (unless you're running an
973 - Imager can now write progressive JPEGs (it could always read them).
974 https://rt.cpan.org/Ticket/Display.html?id=68691
978 - re-work, document and test Imager's logging facility.
979 https://rt.cpan.org/Ticket/Display.html?id=65227
981 - fix META.yml testing and the generated META.yml
982 https://rt.cpan.org/Ticket/Display.html?id=65235
984 - test and add error reporting to to_*() family methods
986 - add to_rgb_double() method.
987 https://rt.cpan.org/Ticket/Display.html?id=65101
989 - Imager no longer exports anything by default
990 https://rt.cpan.org/Ticket/Display.html?id=65228
992 - convert colors to grayscale if the supplied (or generated) palette
993 contains only grays when performing error diffusion color
995 https://rt.cpan.org/Ticket/Display.html?id=68508
997 - writing a paletted image to GIF wouldn't always use the colors
998 supplied (or generated, eg. via make_colors => "mono"), which was
1000 https://rt.cpan.org/Ticket/Display.html?id=67912
1002 - replace (imager|tony)@imager.perl.org in the doc, since I don't
1003 plan to continue receiving mail at that address.
1004 https://rt.cpan.org/Ticket/Display.html?id=68591
1006 Imager 0.83 - 21 May 2011
1011 - diag() the error message on failure for some TIFF tests that are
1012 failing on a Solaris smoker.
1013 http://www.cpantesters.org/cpan/report/6396db1e-8090-11e0-9682-112b785ebe45
1015 Imager 0.82_01 - 17 May 2011
1018 Dev release, in case the compose tests are too sensitive.
1022 - Imager::Font::T1 incorrectly checked for absolute filename under
1023 Win32. Thanks to kmx for the report and fix.
1024 https://rt.cpan.org/Ticket/Display.html?id=67963
1026 - compose() with the target, source or mask position off the top or
1027 left of the target image didn't clip the source image correctly.
1028 https://rt.cpan.org/Ticket/Display.html?id=67220
1030 - compose() now returns a useful error message on a non-positive
1033 - compose.im now at 100% test coverage. (As opposed to, umm, much,
1036 Imager 0.82 - 14 Mar 2011
1041 - eliminate calls to i_has_format() from the test suite, since it's
1042 no longer a useful way to check for file format support. Eliminate
1043 i_has_format() from the functions exposed via XS.
1044 https://rt.cpan.org/Ticket/Display.html?id=65863
1046 - eliminate calls to note(), which isn't in the (very old) version of
1047 Test::More we have as a pre-requisite. note() is modern enough
1048 that I don't feel a need to require a Test::More upgrade for it.
1049 https://rt.cpan.org/Ticket/Display.html?id=65864
1051 - skip the threads tests on Test::More 2.00_*
1052 https://rt.cpan.org/Ticket/Display.html?id=65812
1054 - add an (unshipped) test to check Imager's internal POD links
1055 https://rt.cpan.org/Ticket/Display.html?id=65749
1057 - improve the library detection summary
1058 https://rt.cpan.org/Ticket/Display.html?id=9675
1060 - increase the version of Imager::Font::Type1 so that upgrades don't
1061 downgrade the version in this file.
1062 https://rt.cpan.org/Ticket/Display.html?id=66250
1064 - if we see an -rpath (or -R) option in $Config{lddlflags} supply
1065 that option for the directories that would normally go in
1066 LD_RUN_PATH. Typically an explicit -rpath overrides LD_RUN_PATH.
1067 https://rt.cpan.org/Ticket/Display.html?id=65955
1069 Imager 0.81 - 14 Feb 2011
1072 - added coverage tests for masked images (maskimg.c @100% test coverage)
1074 - add hsv() method to Imager::Color
1075 Thanks to Leolo (Philip Gwyn)
1076 https://rt.cpan.org/Ticket/Display.html?id=65385
1078 - split libt1 Type 1 font support into a sub-module
1079 https://rt.cpan.org/Ticket/Display.html?id=49616 (partial)
1081 - add a preload() class method for use in forking servers, and to
1082 work around limitations in PAR.
1083 https://rt.cpan.org/Ticket/Display.html?id=65665
1087 - paletted writes to a masked image are now masked correctly.
1088 Revealed by new coverage tests.
1090 - update the filter plugin documentation.
1091 https://rt.cpan.org/Ticket/Display.html?id=56513
1093 - add the matrix() method to Imager::Matrix2d to allow creation of a
1094 matrix with specified co-efficients. You can now multiple an
1095 Imager::Matrix2d object by a 9 element array ref or a number.
1096 https://rt.cpan.org/Ticket/Display.html?id=29938
1098 - really fix loading TTF fonts with FT2 when FT1 isn't available.
1099 Thanks to Leolo (Philip Gwyn)
1100 https://rt.cpan.org/Ticket/Display.html?id=65386
1101 https://rt.cpan.org/Ticket/Display.html?id=62855
1103 - make sure each test script that needs testout/ creates it.
1104 https://rt.cpan.org/Ticket/Display.html?id=65088
1106 - handle a slightly different warning from libtiff 4.x
1107 https://rt.cpan.org/Ticket/Display.html?id=65268
1109 - the sat transform2() op returned an incorrect saturation.
1110 https://rt.cpan.org/Ticket/Display.html?id=65391
1112 Imager 0.80 - 17 Jan 2011
1115 - added coverage tests for Imager::Fountain and Imager::Color::Float
1117 - Imager is now maintained in git
1118 http://git.imager.perl.org/imager.git
1119 git://git.imager.perl.org/imager.git
1123 - images with an translucent alpha channel were not scaled correctly
1124 by the default (qtype=normal) scaling method.
1125 https://rt.cpan.org/Public/Bug/Display.html?id=63922
1127 - Imager::Color::Float now translates "#FFFFFF" to white instead of
1128 just a little darker.
1130 - make the default color map build algorithm "mediancut". This
1131 changes the default color map builder for writing GIFs back to what
1132 it was in 0.77, reverting a performance regression.
1133 https://rt.cpan.org/Ticket/Display.html?id=64785
1135 - handle failure to create a masked image correctly
1137 Imager 0.79 - 10 Dec 2010
1140 - add Imager::Test to the POD coverage tests and document the missing
1143 - the convert() method now optimizes the case where all output
1144 channels are either 0, sourced from a single input channel or 1.
1145 This significantly speeds up presets like "addalpha", "green".
1146 https://rt.cpan.org/Ticket/Display.html?id=51254
1148 - add wiggle.pl sample, as suggested by Dan Oppenheim.
1150 - add the combine() method to combine channels from multiple source
1151 images into a new image
1152 https://rt.cpan.org/Ticket/Display.html?id=11872
1156 - treat the co-efficients for convert() as doubles instead of floats.
1158 - If a higher (earlier) priority font handler failed to load, that
1159 would crash preventing loading of later font handlers.
1160 https://rt.cpan.org/Ticket/Display.html?id=62855
1162 - parse defines from the options returned by pkg-config --cflags
1163 https://rt.cpan.org/Ticket/Display.html?id=63223
1165 - a regen of the MANIFEST revealed that GIF and FT2 tests weren't
1166 included in the tarball. They are now included.
1168 Imager 0.78 - 4 Oct 2010
1173 - don't access deprecated members of the png_structp.
1174 https://rt.cpan.org/Ticket/Display.html?id=60242
1176 - document that using color objects is faster than supplying colors
1178 https://rt.cpan.org/Ticket/Display.html?id=61047
1180 - Imager::Probe now accepts array references for incpath and libpath.
1181 https://rt.cpan.org/Ticket/Display.html?id=60244
1183 Imager 0.77_02 - 27 Sep 2010
1186 - moved Win32, FreeType 2 font support into sub-modules.
1187 https://rt.cpan.org/Ticket/Display.html?id=49616 (partial)
1188 Uses Imager::Probe now.
1189 https://rt.cpan.org/Public/Bug/Display.html?id=61328
1191 - tested successfully with jpeg-8b
1192 https://rt.cpan.org/Ticket/Display.html?id=60221
1196 - from _01: look for missing file support test files in the right
1199 - flood_fill() wouldn't fill the right side of a single scan-line
1201 Thanks to Nicolas Roggli for reporting this.
1203 - flood_fill wouldn't fill to the left edge of the image if the
1204 starting line didn't reach the left edge.
1205 Thanks to Nicolas Roggli for reporting this.
1207 Imager 0.77_01 - 13 Sep 2010
1210 - add each library-directory/pkgconfig/ to the pkg-config search path
1212 Thanks to Justin Davis.
1213 https://rt.cpan.org/Ticket/Display.html?id=60491
1215 - moved GIF, TIFF, JPEG file handling code into sub-modules in
1216 preparation for separate distribution.
1217 https://rt.cpan.org/Ticket/Display.html?id=49616 (partial)
1219 Note: this removed support for libungif from Imager.
1221 - optimize filled box drawing (color, not fill)
1225 - Imager::Probe was calling ExtUtils::Liblist to initialize
1226 LD_RUN_PATH supplying an undefined value rather than the found
1227 directory. Thanks to Justin Davis.
1228 https://rt.cpan.org/Ticket/Display.html?id=60491
1230 - only prepend ./ to font filenames when passing them to T1Lib and
1231 then only when it would use its search mechanisms.
1232 https://rt.cpan.org/Ticket/Display.html?id=60509
1234 - fix the cache check for the X rgb.txt loader. This is typically
1235 used for color translation on Unix-like systems, and the fix
1236 improves performance of supplying colors by name by about 80 times.
1237 Test code that managed 3400 10x10 pixel boxes/second sped up to
1240 - clarify that Imager doesn't write EXIF metadata to images.
1241 https://rt.cpan.org/Ticket/Display.html?id=60637
1243 - Imager::Probe can now search subdirectories under its include path.
1244 Used by the PNG Makefile.PL to find headers and libraries when they
1245 aren't in the base directory, as in cygwin.
1246 https://rt.cpan.org/Ticket/Display.html?id=60635
1248 Imager 0.77 - 11 Aug 2010
1251 I don't want Imager::File::PNG indexed as part of Imager, but forgot
1252 to update the META.yml before updating the version.
1254 - don't index Imager::File::PNG as part of Imager
1256 - add resources to META.yml
1258 Imager 0.76 - not released
1263 - the align_string() method would ignore a string of "0" due to a
1264 mis-use of C< ||= >.
1265 Thanks to Maurice Height for reporting this.
1266 https://rt.cpan.org/Ticket/Display.html?id=60199
1268 Imager 0.75_03 - 09 Aug 2010
1273 - read_types() and write_types() would include png when it wasn't
1274 available due to a problem with the %formats tie
1276 - handle dependent libraries correctly (eg -lpng requiring -lz) in
1277 the code run phase of library probing.
1279 Imager 0.75_02 - 07 Aug 2010
1284 - add file missing from MANIFEST, which was causing TIFF failures.
1286 Imager 0.75_01 - 06 Aug 2010
1289 Test release for the new PNG probe.
1291 - added the ability to read multiple-image PNM files.
1292 Note that unlike the pbm/pgm/ppm specification this accepts mixed
1293 format files and allows white space padding between files.
1294 Thanks to Philip Gwyn (Leolo) for this patch.
1296 - moved the PNG file handling code into a sub-module in preparation
1297 for separate distribution.
1298 https://rt.cpan.org/Ticket/Display.html?id=49616 (partial)
1299 Also helps avoid complications from -I/-L compile/link options from
1304 - Imager->new(data => $data) didn't try to process image file data in
1306 https://rt.cpan.org/Ticket/Display.html?id=58761
1308 - t/t50basicoo.t no longer depends on the other tests to generate its
1310 https://rt.cpan.org/Ticket/Display.html?id=9798
1311 Also, it wasn't testing pnm (pnm vs ppm mix-up)
1313 - update the documentation of hardinvert to match the change in 0.62.
1314 https://rt.cpan.org/Ticket/Display.html?id=59785
1316 - added hardinvertall filter which also inverts the alpha channel
1317 (sorry for the mess)
1319 - when probing for TIFF, set LD_RUN_PATH just as the Makefile does so
1320 the probe can find the library for the test run.
1321 https://rt.cpan.org/Ticket/Display.html?id=57518
1323 Imager 0.75 - 20 Jun 2010
1326 - use PERL_NO_GET_CONTEXT to slightly improve performance on threaded
1327 perls (not measured)
1331 - an opacity fill based on a fountain fill would segfault when
1332 filling an 8-bit/sample image.
1334 - merge thickline branch polygon fix
1335 https://rt.cpan.org/Ticket/Display.html?id=43518
1337 Imager 0.74 - 7 May 2010
1342 - read_multi() didn't handle a missing file format library correctly,
1343 aborting on failing to call i_readgif_multi_wiol() or
1344 i_readtiff_multi_wiol().
1346 - fix spelling errors patched by Debian
1347 http://svn.debian.org/viewsvn/pkg-perl/trunk/libimager-perl/debian/patches/spelling.patch?revision=54839&view=markup
1349 - add an (unshipped) author test to spellcheck Imager's POD.
1351 - update the TIFF file format documentation
1352 https://rt.cpan.org/Ticket/Display.html?id=56510
1354 - lib/Imager/IO.pod was written almost 4 years ago but never shipped.
1356 Imager 0.73 - 15 Mar 2010
1359 - implement outline circles, both anti-aliased and not
1360 https://rt.cpan.org/Ticket/Display.html?id=19755
1362 - a combine => "none" fill to a 1 or 3 channel image would produce
1363 the incorrect colour.
1365 Imager 0.72 - 09 Dec 2009
1368 Bump version for release, since 0.71_03 is stable with CPAN testers.
1370 Imager 0.71_03 - 5 Dec 2009
1373 - further adjust the threads test so it only performs the tests on
1374 perls where it's expected to work, and only if the threads module
1377 Imager 0.71_02 - 1 Dec 2009
1380 - adjust the way we load the threads module for the threads test so
1381 it works with non-threaded perls
1383 Imager 0.71_01 - 30 Nov 2009
1388 - use scanline oriented operations to flip images instead of pixel
1390 https://rt.cpan.org/Ticket/Display.html?id=39278
1392 - use double/sample operations to flip large sample images instead of
1393 8-bit sample operations.
1394 https://rt.cpan.org/Ticket/Display.html?id=39280
1397 https://rt.cpan.org/Ticket/Display.html?id=51874
1399 - prevent double-frees when someone creates Imager objects and then
1400 creates a thread. Note: this just handles some simple cases,
1401 Imager doesn't support perl threads, and isn't likely to.
1402 https://rt.cpan.org/Ticket/Display.html?id=52268
1404 Imager 0.71 - 16 Nov 2009
1407 - add the opacity fill type - an adaptor that modifies the opacity of
1412 - the conv filter now enforces that the sum of the coefficients is
1413 non-zero. Also, rather than skipping pixels off the edge off the
1414 edge of the image, the closest edge pixel is used. Previously
1415 dividing by the zero sum of coefficients could cause invalid
1416 results or runtime exceptions.
1417 Thanks to David Cantrell's Alpha-NetBSD CPAN test box for revealing
1420 Imager 0.70 - 21 Sep 2009
1425 - release image row and comments memory on all error returns in gif
1428 - handle zero length extensions, previously this would cause a null
1430 Thanks to Krzysztof WojtaÅ for the test data and fix for this.
1432 - an integer division meant that preview scaling to below 1 pixel
1433 wide or high (which isn't too useful anyway) was calculating using
1434 NaNs on most platforms, and causing an exception on others.
1435 Thanks to David Cantrell for producing a backtrace of the crash on
1436 his Alpha-NetBSD CPAN test box which made it possible to track this
1439 Imager 0.69 - 08 Sep 2009
1444 - broken test fix - was attempting to call a function skip_all, when
1445 that should be a parameter to plan().
1447 - briefly document apidocs.perl, the tool used to build
1448 Imager::APIRef and make some minor enhancements
1450 - various minor documentation enhancements and fixes.
1452 Imager 0.68 - 07 Sep 2009
1455 - Imager->new(file => $filename) and other similar incantations will
1456 load the given file.
1457 https://rt.cpan.org/Ticket/Display.html?id=48261
1461 - avoid using CHECK as a label in Imager::Test
1462 http://nntp.x.perl.org/group/perl.cpan.testers/5220921
1464 - re-work most image file test files that require a library into
1465 separate library present/not present files to remove stupidly long
1468 - don't treat rubthrough() outside the bounds of the target image as
1470 http://nntp.x.perl.org/group/perl.cpan.testers/5185716
1472 Imager 0.67_01 - 02 Sep 2009
1477 - correct documentation of default of raw image interleave read
1479 https://rt.cpan.org/Ticket/Display.html?id=42074
1481 - add raw_ prefix to raw read parameters, though the original names
1484 - fail the read if an invalid raw_interleave parameter is supplied
1486 - warn if no interleave or raw_interleave parameter is supplied,
1487 since the documented default was wrong, and incompatible with the
1490 - for reading raw images, if raw_storechannels > raw_datachannels,
1491 set the extra channels in the image to 0
1493 - when probing for executables like freetype-config, search for .bat
1494 and .cmd on MSWin32, as well as .exe.
1495 https://rt.cpan.org/Ticket/Display.html?id=49275
1497 - re-work the external libraries section of README:
1498 - list Debian and Redhat package names for each library
1502 - use the new EU::MM META_MERGE facility instead of generating
1503 META.yml from scratch
1504 https://rt.cpan.org/Ticket/Display.html?id=47888
1506 - use Devel::CheckLib (bundled, modified) to check which release of
1507 libtiff is installed and reject 3.9.0
1508 http://bugzilla.maptools.org/show_bug.cgi?id=2088
1509 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=543079
1511 Imager 0.67 - 12 Dec 2008
1516 - fix a packaging error
1518 Imager 0.66 - 12 Dec 2008
1521 - 24-bit color .ICO/.CUR files can now be read.
1525 - an optimization skipping 0 src alpha values could cause the
1526 rubthrough() to read past the end of a buffer.
1527 http://www.nntp.perl.org/group/perl.cpan.testers/2008/05/msg1509184.html
1529 - corrected a reference leak where writing GIFs would leak memory.
1530 This could also happen calling to_paletted().
1531 Also documented the underlying long existing feature where the
1532 colors parameter is filled with the generated color table and added
1534 http://rt.cpan.org/Ticket/Display.html?id=41028
1536 - write out the image size in bytes field of a BMP correctly.
1537 http://rt.cpan.org/Ticket/Display.html?id=41406
1539 - add limited tests for Imager::ExtUtils
1541 - make Imager::ExtUtils->includes use an absolute path, since
1542 a relative path could cause failures using Inline::C.
1543 http://rt.cpan.org/Ticket/Display.html?id=37353
1545 - re-arrange the POD for Imager::Font::BBox:
1546 - mark total_width(), pos_width(), end_offset() obsolete, since
1547 they're mostly for backwards compatibility
1548 - group width methods and height methods
1549 https://rt.cpan.org/Ticket/Display.html?id=39999
1551 Imager 0.65 - 20 May 2008
1556 - In some cases when an error occurs reading those parts of a JPEG
1557 file after the image the scan-line buffer could be freed a second
1558 time. In cases where the the error occured while reading the image
1559 data it's possible that the buffer could have leaked.
1560 Thanks to Gabriel Vasseur for reporting this and help in tracking
1563 - the gif_screen_height tag was overriding the screen width and being
1564 ignored for the screen height when present.
1565 https://rt.cpan.org/Public/Bug/Display.html?id=35568
1567 Imager 0.64 - 23 April 2008
1570 This is a bug fix release. This includes a fix for a possible
1575 - Possible security issue: The floating point sample path for image
1576 based fills had a buffer overflow. This would overwrite the end of
1577 a malloc()ed buffer with double precision floats.
1578 http://rt.cpan.org/Ticket/Display.html?id=35324
1581 - check that the result of fileno($fh) is defined rather than simply
1582 true when read() or write() is supplied with an fh parameter.
1583 http://rt.cpan.org/Ticket/Display.html?id=35139
1585 - i_scale_axis() wasn't checking the result of i_img_new_ch()
1586 resulting in a SIGSEGV when attempting to scale an image to a size
1587 too large to fit in memory. This is a NULL pointer access issue,
1588 not a buffer overflow.
1589 Added a check for the failure.
1590 scale_calculate() (and hence scale()) will now fail if any of the
1591 scale size parameters are a reference.
1592 http://rt.cpan.org/Ticket/Display.html?id=35172
1594 - Regression: filling a greyscale image with a hatch used the wrong
1595 color channels from the supplied fg/bg colors.
1596 https://rt.cpan.org/Ticket/Display.html?id=35278
1598 - fixed a related problem for image fills.
1600 Imager 0.63 - 7 April 2008
1603 This release primarily contains changes to improve ease of use -
1604 rather than you having to convert images to the appropriate number of
1605 channels, Imager handles it internally. How to handle drawing colors
1606 and the default combine mode is a thornier problem left for some other
1609 - the font libraries are now only initialized when needed.
1610 http://rt.cpan.org/Ticket/Display.html?id=28825
1612 - moved the imtoc.perl code into Imager::Preprocess
1614 - paste() and rubthrough() now adapt the source image data to the
1615 destination, so you can now safely paste/rubthrough from greyscale
1616 images to color images or back, or from alpha channel images to
1617 noalpha channels or back.
1618 https://rt.cpan.org/Ticket/Display.html?id=30908
1620 - rubthrough() now falls back to pasting when the source doesn't have
1621 an alpha channel. This effectively treats the source as having a
1622 max alpha channel, the right thing to do.
1623 http://rt.cpan.org/Ticket/Display.html?id=29944
1625 - re-worked most of the area filling code to use a common set of
1626 functions when filling.
1627 Corrected normal combine mode.
1628 Rewrote most of the combine modes to match the way the SVG draft
1629 defines them with respect to a translucent source and destination.
1630 Added tests for translucent source and destination.
1631 Added tests to check 8-bit/sample and double/sample combines work
1633 https://rt.cpan.org/Ticket/Display.html?id=29879
1635 - writing a 2 or 4 channel image to a JPEG file will now write that
1636 image as if composited against a background, black by default,
1637 overridable with the i_background tag/parameter.
1638 https://rt.cpan.org/Ticket/Display.html?id=29876
1640 - writing a 2 or 4 channel image to a PGM/PPM file will now write
1641 that image as if composited against a background, black by default,
1642 overridable with the i_background tag/parameter.
1643 http://rt.cpan.org/Ticket/Display.html?id=30074
1645 - writing a 2 or 4 channel image to a BMP file will now write that
1646 image as if composited against a background, black by default,
1647 overridable with the i_background tag/parameter.
1648 http://rt.cpan.org/Ticket/Display.html?id=30075
1652 - Imager::Matrix2d->translate() now only requires one of the x or y
1654 http://rt.cpan.org/Ticket/Display.html?id=29937
1656 - mixing qtype scaling now sets all channels of a pixel to zero if
1657 the pixel has zero coverage (zero alpha). This should produce more
1658 compressible output files.
1659 http://rt.cpan.org/Ticket/Display.html?id=32324
1661 - removed the pointless #! line from lib/Imager/Font/Wrap.pm
1663 https://bugzilla.redhat.com/show_bug.cgi?id=166254
1664 I'm not changing the #! lines of the sample code, since it's sample
1665 code, not intended for installation.
1666 http://rt.cpan.org/Ticket/Display.html?id=33408
1668 - some TGA images weren't being detected correctly as TGA images
1669 https://rt.cpan.org/Ticket/Display.html?id=32925
1671 - handling of the left-over bit for 16-bit/pixel TGA images has been
1672 changed to match the behaviour of the GIMP. Previously the bit
1673 being set was treated as an opaque pixel, but one user reported a
1674 problem with loading such an image. I haven't been able to find any
1675 tools beyond the GIMP that handle alpha-channel 16-bit TGAs, so
1676 I'll match it's behaviour. See issue 114913 in the GIMP's
1678 http://rt.cpan.org/Ticket/Display.html?id=32926
1680 Imager 0.62 - 10 December 2007
1683 - Makefile.PL now expands ~/path supplied to --incpath or --libpath
1684 to /path under your home directory.
1685 http://rt.cpan.org/Ticket/Display.html?id=29484
1687 - the old dynaload code used Mach API functions to load dynamic
1688 libraries on Mac OS X. These APIs have been deprecated in OS X
1689 10.5 and were causing some build problems.
1690 So henceforth Imager uses the dlopen() family of functions, and you
1691 will need version 10.3 or later of OS X.
1693 - added the det() function to the transform2() engine.
1694 added the sample quad_to_square.pl
1695 Courtesy Richard Fairhurst.
1696 http://rt.cpan.org/Ticket/Display.html?id=31244
1700 - samples/gifscale.pl sourced the base value for gif_top from
1702 Thanks to Eleneldil G. Arilou for pointing this out.
1704 - t/t82inline.t no longer loads B at runtime, to work around a bug
1705 in some 5.005_0[45] installations.
1706 http://rt.cpan.org/Ticket/Display.html?id=30508
1708 - work around Module::Depends::Intrusive bug #21229
1709 http://rt.cpan.org/Ticket/Display.html?id=30520
1711 - the hardinvert filter no-longer inverts the alpha channel.
1712 http://rt.cpan.org/Ticket/Display.html?id=30002
1714 - the hardinvert filter now supports large samples
1716 Imager 0.61_02 - 28 November 2007
1719 - major TIFF support re-work
1720 http://rt.cpan.org/Ticket/Display.html?id=20329
1722 - added a C level image interface for accessing samples from 1-32
1723 bits, exposed this at the perl level in getsamples()
1725 - the conv filter now works at floating point precision for high bit
1728 - added is_bilevel method to test whether an image should be written as
1729 a bilevel image if the image format supports it.
1731 - added -log-stderr as an Imager import list option
1733 - added some important types to Imager::APIRef
1735 - added test_image_double() to Imager::Test
1739 - Imager::Fountain couldn't read GIMP gradient files with 10 or more
1742 - the scale() method with qtype mixing now handles images with an
1743 alpha channel correctly.
1745 - fixed a broken link from the "animated GIF" entry in the concept index.
1746 Thanks to Slaven Rezic.
1747 http://rt.cpan.org/Ticket/Display.html?id=30889
1749 - on some perl's the infix expression parser test would fail due to
1750 actions in the grammar returning false. Made sure all actions return
1752 Thanks to Richard Fairhurst for spending a lot of time in tracking
1754 http://rt.cpan.org/Public/Bug/Display.html?id=29562
1756 Imager 0.61 - 5 November 2007
1759 - added samples/gifscale.pl, which adjusts the screen size/position tags
1760 when scaling an animated gif
1761 http://rt.cpan.org/Ticket/Display.html?id=27591
1765 - correct handling of sz in matrix_transform() - this should allow
1766 perspective type transformations to work now.
1767 http://rt.cpan.org/Ticket/Display.html?id=29936
1769 - prevent a cast to integer warning on x64 builds in datatypes.c
1770 also fixed some other type warnings
1771 https://rt.cpan.org/Ticket/Display.html?id=30204
1773 - some sub-directory tests depended on files produced by the parent
1775 http://rt.cpan.org/Ticket/Display.html?id=30203
1777 - Imager::Font::Wrap doesn't correctly set savepos
1778 thanks to Nikita Dedik and Eleneldil G. Arilou for reporting this.
1779 http://rt.cpan.org/Ticket/Display.html?id=29771
1781 - test 171 in t/t01introvert.t was failing on perls configured to
1783 http://rt.cpan.org/Ticket/Display.html?id=29413
1785 - the code for the transform2() uminus operator was missing a break.
1786 Added tests for better code coverage of the ops.
1787 http://rt.cpan.org/Ticket/Display.html?id=29296
1789 - the SGI RLE compression code could overflow its compression buffer
1790 http://rt.cpan.org/Ticket/Display.html?id=30334
1792 - the 32-bit output function used by the SGI code only handled values
1793 under 0x10000. This was most noticable when writing large RLE images.
1794 http://rt.cpan.org/Ticket/Display.html?id=30335
1796 - validate chan_count for chans == NULL for each of the i_gsamp()
1798 http://rt.cpan.org/Ticket/Display.html?id=28985
1800 - attempt to work around the test failure at
1801 http://www.nntp.perl.org/group/perl.cpan.testers/2007/09/msg650810.html
1802 http://rt.cpan.org/Ticket/Display.html?id=29562
1804 - improve the error messages produced when attempting to read or write
1805 an unknown image file format.
1806 http://rt.cpan.org/Ticket/Display.html?id=30103
1808 - improve the transform2() documentation
1809 http://rt.cpan.org/Ticket/Display.html?id=29267
1811 - correctly generate the author key in META.yml
1812 http://rt.cpan.org/Ticket/Display.html?id=30377
1814 - correctly blend a rotated (or matrix_transformed()) image when
1815 performing interpolation in the presence of an alpha channel.
1816 Also corrected the centring of the rotated image on the output
1819 Imager 0.60 - 30 August 2007
1822 - Finished/rewrote Arnar's old SGI RGB file format support, so Imager
1823 now has full SGI RGB image format, including RLE and 16-bit/sample
1825 https://rt.cpan.org/Ticket/Display.html?id=8666
1827 - logging functions are now available in the API
1829 - applied Gabriel Vasseur's patch
1830 added documentation, further tests, and support for greyscale images
1831 Obviously problems are my fault :)
1832 https://rt.cpan.org/Ticket/Display.html?id=28142
1834 - the mask for ICO/CUR images is now applied as an alpha channel to
1835 the returned image. For the old behaviour, supply ico_masked => 0
1836 to read() or read_multi(). This should be less confusing when
1837 using Imager as a general image processor.
1838 https://rt.cpan.org/Ticket/Display.html?id=29001
1842 - in some cases it's possible for giflib/libungif to return color
1843 indexes outside the range of colors defined by the image's palette.
1844 We now expand the palette to match the indexes used.
1845 Thanks to Gabriel Vasseur for reporting this.
1847 - fixed various memory leaks that could occur when failing to read png,
1848 jpeg, bmp or tga files.
1850 - to avoid confusion, channels not present in the image are returned as
1851 zero by getscanline(). This has no effect on the C level i_glin()
1852 and i_glinf() API functions which continue to not set the unused
1855 - the convert() method now returns an image of the same sample size as
1857 https://rt.cpan.org/Ticket/Display.html?id=28492
1859 - remove repeated text in Imager::Files
1860 http://rt.cpan.org/Ticket/Display.html?id=27589
1862 - be even more explicit that scale() and friends don't modify the source
1863 image, but return a new image.
1864 http://rt.cpan.org/Ticket/Display.html?id=28570
1866 - improve the error message from errstr() when you try to load a font
1867 for which the driver hasn't been built in Imager.
1868 http://rt.cpan.org/Ticket/Display.html?id=27571
1870 - transparency is now enabled by default when writing GIF images
1871 http://rt.cpan.org/Ticket/Display.html?id=27615
1873 - Imager would not load on Windows 98
1874 http://rt.cpan.org/Ticket/Display.html?id=27653
1876 Imager 0.59 - 14 June 2007
1881 - fixes a regression introduced by the fixes for RT 11972
1882 http://rt.cpan.org/Ticket/Display.html?id=27546
1884 - cropping outside the image would return an Imager object with
1885 no low-level image object, instead of returning false.
1886 Fixed by: Philip Gwyn (Leolo)
1887 http://rt.cpan.org/Ticket/Display.html?id=27509
1889 Imager 0.58 - 16 May 2007
1892 No significant changes from 0.57_01.
1894 Imager 0.57_01 - 11 May 2007
1897 - added to_rgb16 to produce a 16-bit/sample version of an image
1899 - improve freetype 1.x text output efficiency
1903 - search another place for rgb.txt, and check all the places
1904 Imager::Color checks when deciding whether to skip testing it
1905 http://rt.cpan.org/Ticket/Display.html?id=26064
1907 - use a convolution kernel size based on the stddev rather than a
1908 fixed size when performing a gaussian blur
1909 http://rt.cpan.org/Ticket/Display.html?id=25645
1911 - document the difference() method's mindist parameter, and debug it.
1913 - put the Imager release number in the Inline::C generated code to
1914 regenerate Inline code when a new release of Imager is installed.
1915 http://rt.cpan.org/Ticket/Display.html?id=26278
1917 - fix rendering on alpha channel images for the FreeType 1.x driver.
1918 http://rt.cpan.org/Ticket/Display.html?id=11972
1920 - fix rendering on alpha channel images for the T1lib driver.
1921 http://rt.cpan.org/Ticket/Display.html?id=11972
1923 - reworked library probing, we can now set more than one probe
1924 function for a library. Disabled the default (non-freetype-config)
1925 library probe and added an extra probe function that searches for
1926 both ft2build.h and whatever it includes, and adds -I as needed.
1927 Hopefully this will fix build problems like
1928 http://www.nntp.perl.org/group/perl.cpan.testers/2007/05/msg472281.html
1929 http://rt.cpan.org/Ticket/Display.html?id=26086
1931 Imager 0.57 - 30 Apr 2007
1934 This is a maintenence release fixing a security issue in Imager.
1936 - CRITICAL: a specially crafted compressed BMP file can cause a buffer
1937 overflow in malloced memory. There will be further discussion of
1938 this issue in the ticket below.
1939 http://rt.cpan.org/Ticket/Display.html?id=26811
1940 CVE-2007-2459 CVE-2007-2413
1941 The descriptions at cve.mitre.org varied in quality, please see the
1942 ticket at rt.cpan.org for a more accurate description of the issue.
1944 Imager 0.56 - 1 Apr 2007
1947 - added support for reading 16-bit/sample PGM/PPM images
1949 - added support for writing 16-bit/sample PGM/PPM images
1951 - improved performance of reading PBM/PGM/PPM images
1953 - added support for writing PBM images if the image is paletted and
1954 contains only black and white
1956 - added a new make_colors value - "mono"
1958 - switched from the svn log Changes to a manual Changes to reduce
1961 - new sample code - samples/flasher.pl
1965 - CRITICAL: the "Imager" typemap entry (not used by Imager itself)
1966 was returning an image object with an extra reference, this
1967 resulted in a memory leak.
1968 http://rt.cpan.org/Ticket/Display.html?id=24992
1970 - fix rendering on alpha channel images for the FreeType 2.x driver
1971 http://rt.cpan.org/Ticket/Display.html?id=11972
1973 - reading bmp files now consitently handles short reads. You can now
1974 supply a parameter to treat a short read as successful and set
1976 http://rt.cpan.org/Ticket/Display.html?id=8426
1978 - previously, reading ASCII PBM files required spaces between samples,
1979 even though the format doesn't require that
1981 - improved documentation of the unsharpmask filter (I hope)
1982 http://rt.cpan.org/Ticket/Display.html?id=25531
1984 - force flushing of the output from i_tt_dump_names() and test output
1985 in t/t35ttfont.t to prevent output from being mixed up.
1986 https://rt.cpan.org/Ticket/Display.html?id=24859
1988 - rewrite a conditional expression as an if() to hopefully work around
1989 a bug in the pre-4.0 GCC Apple shipped with OS X 10.4.
1990 https://rt.cpan.org/Ticket/Display.html?id=25561
1992 - avoid Data::Dumper in regops.perl to support older releases of perl
1993 https://rt.cpan.org/Ticket/Display.html?id=24391
1995 Imager 0.55 - 16 Dec 2006
1998 This is primarily a bug fix release.
2000 Note: Test::More is now a pre-requisite for Imager and is no longer bundled.
2002 There is one new feature:
2004 - the Win32 font driver now supports UTF8 (RT 22166)
2005 http://www.cpanforum.com/threads/3276
2006 http://rt.cpan.org/Ticket/Display.html?id=22166
2008 Several bugs were fixed:
2010 - the string() method would not output the string "0"
2011 http://rt.cpan.org/Public/Bug/Display.html?id=21770
2013 - fills.c was failing to compile on Solaris 10 (compiler unknown)
2014 http://rt.cpan.org/Public/Bug/Display.html?id=21944
2016 - the gif_disposal and gif_user_input tags weren't being read from
2018 http://rt.cpan.org/Public/Bug/Display.html?id=22192
2020 - gif.c was failing to build under MSVC
2021 http://rt.cpan.org/Ticket/Display.html?id=23922
2023 - in some cases strings passed to the string() method were treated as
2024 terminated by NUL (chr 0)
2025 http://rt.cpan.org/Public/Bug/Display.html?id=21770
2027 - on "MSWin32" perl builds we now link to -lzlib instead of -lz since
2028 that's the default build name for zlib on Win32.
2029 http://rt.cpan.org/Ticket/Display.html?id=23064
2031 - search $Config{incpath} for headers too, which we should have been
2034 Win32 font driver fixes:
2036 - the global descent value from bounding box was the wrong sign
2037 http://www.cpanforum.com/threads/3276
2039 - if the first or last glyph overflowed the left or right side of the
2040 advance width they would be clipped
2043 Imager 0.54 - 14 Sep 2006
2046 This is primarily a feature release:
2048 - a new qtype value 'mixing' has been added to the scale()
2049 method. This is faster than 'normal', slower than 'preview'. This
2050 is based on the method used by pnmscale, and seems to produce less
2051 blurry results than normal.
2052 http://rt.cpan.org/Public/Bug/Display.html?id=20677
2054 - the rubthrough() method can now render onto images with an alpha
2056 http://rt.cpan.org/Ticket/Display.html?id=20678
2058 - the read_multi() method now falls back to calling doing a single
2059 image read via the read() method and write_multi() will now fall
2060 back to calling write() if a single image is supplied. This means
2061 you can simply call the read_multi() or write_multi() functions
2062 without having to check if the type is formatted by that method.
2063 http://rt.cpan.org/Ticket/Display.html?id=19457
2064 http://rt.cpan.org/Ticket/Display.html?id=19458
2066 - the GIF loop extension can now be written. If you don't have
2067 libungif/giflib 4.1.4 (or some distribution's bugfixed equivalent) you
2069 http://rt.cpan.org/Ticket/Display.html?id=21185
2071 - getscanline() and setscanline() can now read/write palette index
2072 based data from/to the image for paletted images, by setting type to
2074 http://rt.cpan.org/Ticket/Display.html?id=20338
2076 - we no longer hassle you to disable GIF support
2077 http://rt.cpan.org/Ticket/Display.html?id=20687
2079 - minor documentation fixes
2082 Imager 0.53 - 26 Jul 2006
2085 This is a bugfix release.
2087 Some test code was left in a code path not covered by the test
2088 suite. A test was added to cover this code path and the test code was
2090 http://rt.cpan.org/Public/Bug/Display.html?id=20705
2093 Imager 0.52 - 25 Jul 2006
2096 This is primarily a feature release, but contains a fair few bug
2097 fixes, new features:
2099 - ability to read and write MS Windows ICO and CUR files
2101 - you can now add file format plugins to support new file formats
2103 - add POD coverage tests
2105 - setcolors() and addcolors() now accept color names and so on
2106 instead of requiring Imager::Color objects.
2107 http://rt.cpan.org/Ticket/Display.html?id=20056
2109 - flood_fill() can now fill to a specified border color instead of
2110 just to the area the same color as the seed.
2111 http://rt.cpan.org/Ticket/Display.html?id=19618
2116 - bounding_box for the T1 driver wasn't converting UTF8 to ascii when
2117 calculating the advance width.
2118 http://rt.cpan.org/Public/Bug/Display.html?id=20554
2120 - bounding_box for the T1 driver wasn't including leading and
2121 trailing spaces in the bounding box as the other drivers did, it also
2122 produced strange results for empty strings or strings containing only
2125 - when reading CMYK jpeg images they were being transferred to the
2126 image object as is, producing a four channel image. It only looked ok
2127 due to an old still unfixed Photoshop bug. We now convert from the
2128 inverted CMYK that photoshop (and Corel for example) produce into RGB.
2129 http://rt.cpan.org/Ticket/Display.html?id=20416
2131 - reading a CYMK TIFF would result in a 4 channel image, reading any
2132 image with more than 4 channels (eg. RGB with 2 alpha channels) would
2134 http://rt.cpan.org/Ticket/Display.html?id=20415
2136 - added /usr/local/include to the default include search path, since
2137 we were already searching /usr/local/lib for libraries.
2139 And various minor fixes and documentation updates.
2142 Imager 0.51 - 23 Apr 2006
2145 - fix a validation bug when processing JPEG EXIF data that can cause
2147 http://rt.cpan.org/Public/Bug/Display.html?id=18496
2149 - fix mis-processing of the src_maxx and src_maxy parameters of the
2151 http://rt.cpan.org/Public/Bug/Display.html?id=18712
2153 - fix a problem in Imager's "smart" handling of the color parameter
2155 http://rt.cpan.org/Public/Bug/Display.html?id=18561
2158 Imager 0.50 - 29 Mar 2006
2161 - CRITICAL: fixes a segmentation fault from attempting to write a 2
2162 or 4 channel image to jpeg or a 2 channel image to tga where the
2163 output is an in-memeory buffer.
2164 http://rt.cpan.org/Public/Bug/Display.html?id=18397
2166 - fixes an incorrect pointer parameter in the PNG code
2167 http://rt.cpan.org/Public/Bug/Display.html?id=18051
2169 - skip Inline::C tests when building in a directory with spaces
2170 http://rt.cpan.org/Public/Bug/Display.html?id=18049