Log message:
Update to 9.06 from mef via pkgsrc-wip.

(Upstream) Version 9.06 (2012-08-08)

This is the sixth full release in the stable 9.x series.

Highlights in this release include:

  * PDF/A-2 - pdfwrite now supports the creation of PDF/A-2 files. The 'PDFA'
    command line switch can now take a numeric parameter:
      + 0 = not PDF/A compliant
      + 1 = PDF/A-1b compliant
      + 2 = PDF/A-2b compliant

    Simply specifying "-dPDFA" continues to have the old behaviour of \ 
creating
    PDF/A-1b files. For PDF/A-2 the command line should include "-dPDFA=2".
  * pdfwrite "Server mode" - pdfwrite can now be run in "server \ 
mode" which
    allows the device to be closed without closing the interpreter. This means
    it is no longer necessary to terminate GS before starting a new PDF
    conversion. This leads on to:
  * pdfwrite now supports the "%d" format in the OutputFile switch. If \ 
this is
    set then pdfwrite will output each page of input to an individual file.
  * ps2write - recent exposure to a range of PostScript devices has thrown up
    some interesting deficiencies in those devices. ps2write now emits
    PostScript in slightly different ways in order to produce output on a wider
    variety of devices. In some cases this also results in improved print times
    but it is still important to set the resolution appropriately for the
    output device, especially if the input contains transparency.
  * Ghostscript can now use output intents defined in PDFs by using the
    "-dUsePDFX3Profile" command line option. See ICC Color Parameters for
    details.
  * tiffsep/tiffsep1: support for large numbers of separations improved. The
    previous implementation of those devices utilised a "compressed color
    encoding" to represent the tints for all the plates in one 64 bit value. As
    the number of plates increased, fewer bits were available for the tint for
    each plate, ultimately resulting in an "unencodable pixels" error. \ 
These
    revisions remove the reliance on the compressed color encoding, thus
    ensuring that we have a consistent color bit depth, regardless of the
    number of plates, and ensuring the "unencodable pixels" error will \ 
never
    occur.

    Also as a result of these changes, there are substantial performance
    improvements in jobs with separations and transparency.
  * tiffsep, psdcmyk and psdrgb now support the "downscaler" functionality.
    This brings the "tiffscaled" style functionality to the DeviceN output
    devices, so jobs can be rendered internally in contone and at a high
    resolution, and the output optionally downsampled by a level specified by
    "-dDownScaleFactor=n", and also optionally error diffused to 1bpp \ 
output.
  * Third party libraries: libtiff, libpng, libjpeg , Freetype and zlib have
    all be updated.
  * clist storage, for rendering pages in bands, is now a run-time option:
    -sBandListStorage={file|memory}.
  * Plus the usual round of bug fixes, compatibility changes, and incremental
    improvements.

Log message:
'cups' option needs to ensure BUILD_TARGET 'cups' is added.
configure adjusted to correctly accept CUPSVERSION and, while we're at it,
add missing space between '!' and parens in test for contrib directory.

Log message:
Set USE_GCC_RUNTIME=yes for packages which build shared libraries but do
not use libtool to do so.  This is required to correctly depend upon a
gcc runtime package (e.g. gcc47-libs) when using USE_PKGSRC_GCC_RUNTIME.

Log message:
add utf8 [default] option to ghostscript-gpl

Log message:
Recursive PKGREVISION bump for OpenSSL API version bump.

Log message:
Racy build.

Log message:
Update comment lines to mention the Makefile's referencing.

Log message:
fonts/ghostscript-cidfonts/Makefile now DEPENDS+= print/ghostscript only
and NOT including this file, thanks dholland@.

Log message:
Fix build on OpenBSD 5.3.

* Pass LDFLAGS to gsx build etc.
* Pass -lxcb to LDFLAGS for OpenBSD

Log message:
Try to fix the fallout caused by the fix for PR pkg/47882. Part 3:

Recursively bump package revisions again after the "freetype2" and
"fontconfig" handling was fixed.