Remove unexecuted shell script snippets.

(tron)

Tell locatin of uic and qmake with cmake way too, and avoid to detect
unwanted builtin one.

(obache)

Add comment for first chunk.

(wiz)

Sort contents (no changes to binary package).

(tron)

Updated x11/xtrans to 1.3.4

(wiz)

Update to 1.3.4:
I'm happy to announce xtrans-1.3.4

Hans de Goede (2):
      configure: Also add -D_DEFAULT_SOURCE to .pc cflags to shut up glibc warnings
      xtrans 1.3.4

Mark Kettenis (1):
      Increase UNIX socket buffer size

Thomas Klausner (2):
      Cast ctype(3) function arguments to unsigned char.
      Add missing headers for free() and strlen().

(wiz)

Updated graphics/optipng; graphics/pngcrush; databases/sqlite3; databases/mysql55; databases/mysql56; Added graphics/SDL2_image

(adam)

Changes 5.6.17:

Functionality Added or Changed

Incompatible Change: The AES_ENCRYPT() and AES_DECRYPT() functions now permit control of the block encryption mode and take an optional initialization vector argument:

The new block_encryption_mode system variable controls the mode for block-based encryption algorithms. Its default value is aes-128-ecb, which signifies encryption using a key length of 128 bits and ECB mode.

An optional init_vector argument provides an initialization vector for encryption modes that require it:

AES_ENCRYPT(str,key_str[,init_vector])
AES_DECRYPT(crypt_str,key_str[,init_vector])
A random string of bytes to use for the initialization vector can be produced by calling the new RANDOM_BYTES() function.

For more information, see Encryption and Compression Functions.

These changes make statements that use AES_ENCRYPT() or AES_DECRYPT() unsafe for statement-based replication and they cannot be stored in the query cache. Queries that use RANDOM_BYTES() are unsafe for statement-based replication and cannot be stored in the query cache.

Incompatible Change: The ERROR_FOR_DIVISION_BY_ZERO, NO_ZERO_DATE, and NO_ZERO_IN_DATE SQL modes now are deprecated and setting the sql_mode value to include any of them generates a warning. In MySQL 5.7, these modes do nothing. Instead, their effects are included in the effects of strict SQL mode (STRICT_ALL_TABLES or STRICT_TRANS_TABLES). The motivation for the change in MySQL 5.7 is to reduce the number of SQL modes with an effect dependent on strict mode and make them part of strict mode itself.

To make advance preparation for an upgrade to MySQL 5.7, see SQL Mode Changes in MySQL 5.7. That discussion provides guidelines to assess whether your applications will be affected by the SQL mode changes in MySQL 5.7.

InnoDB: MySQL now supports rebuilding regular and partitioned InnoDB tables using online DDL (ALGORITHM=INPLACE) for the following operations:

OPTIMIZE TABLE

ALTER TABLE ... FORCE

ALTER TABLE ... ENGINE=INNODB (when run on an InnoDB table)

Online DDL support reduces table rebuild time and permits concurrent DML, which helps reduce user application downtime. For additional information, see Overview of Online DDL.

On Solaris, mysql_config --libs now includes -R/path/to/library so that libraries can be found at runtime.

mysql_install_db provides a more informative diagnostic message when required Perl modules are missing.

The IGNORE clause for ALTER TABLE is now deprecated and will be removed in a future version of MySQL. ALTER IGNORE TABLE causes problems for replication, prevents online ALTER TABLE for unique index creation, and causes problems with foreign keys (rows removed in the parent table).

Bugs Fixed

(adam)

Updated security/libguardtime to 0.3.13

(fhajny)

Update libguardtime to 0.3.13.

2014-03-04 release (0.3.13)
* Mostly added convenience functions for wrapper APIs:
  - functions to manage the list of CA roots from client code;
  - functions to convert hash values to and from imprints;
  - a function to extract the data hash from a timestamp.

2014-01-20 release (0.3.12)
* Publications file verification switched away from hardcoded CA root
  certificate:
  - on most platforms OpenSSL trust store is used;
  - on Windows Crypto API and its trust store are used by default,
    with option to switch to OpenSSL and its trust store.
* Now builds with both OpenSSL 0.9.8 and OpenSSL 1.0.
* Lifted a restriction of the underlying WinINet library that limited
  the number of simultaneous HTTP requests to 2 on Windows.
* Various other improvements to the build scripts.
* Replaced test timestamps with more recent ones.

(fhajny)

Changes 1.7.73:
* Restored calls to png_set_crc_action() which had been removed from
  version 1.7.72 for some testing and inadvertently not restored.
* Changed "fix" internal variable name to "salvage" (still set with "-fix")
* Added code to fix/salvage PNG with "bad adaptive filter value" error.
* Avoid calculating CRC during compression trials except for the last trial,
  when the output is actually written.
* Fixed a bug with reducing 16-bit images to 8-bit using "-reduce" option.

(adam)

Added graphics/SDL2_image version 2.0.0

(adam)

This is a simple library to load images of various formats as SDL surfaces.
This library supports BMP, PNM (PPM/PGM/PBM), XPM, LBM, PCX, GIF, JPEG, PNG,
TGA, and TIFF formats.

(adam)

Add binutils-mips-current and gcc-mips-current

(martin)

Updated cross/gcc-mips-current to 20140330

(martin)

Add a gcc-current pkg suitable for NetBSD/playstation2 kernel cross
compilation.

(martin)

Updated cross/binutils-mips-current to 2.23.91

(martin)

Add a binutils-2.23 pkg suitable for NetBSD/playstion2 cross compilation.

(martin)

Updated pkgtools/x11-links to 0.85

(ryoon)

Update to 0.85

* Support NetBSD current (6.99.38 at least)'s new native xorg

(ryoon)

Remove mozilla-mirror.naist.jp from mozilla mirror sites.
It does not exist anymore.
Pointed out by nonaka@, thank you.

(ryoon)

Updated misc/raspberrypi-userland 20140329

(skrll)

Update to 20140329 based on bf068918b5fa6eb53d4b72846b36f2168cfb532e of
https://github.com/jaredmcneill/userland

Various bug and gcc 4.8 fixes

(skrll)

add misc/py-stdnum

(richard)

Add missing header for DragonflyBSD.
PR 48691 by David Shao.

(wiz)

add py-stdnum

(richard)

Add misc/py-stdnum

Parse, validate and reformat standard numbers and codes.

This library offers functions for parsing, validating and reformatting standard numbers and codes in various formats.

(richard)

Updated textproc/mdoclint to 1.29

(wiz)

OpenBSD added CONTEXT section to man9. Update library list. From
jmc@openbsd.
Bump version.

(wiz)

Fix fprintf format string (already fixed the same way in src).

(wiz)

Updated devel/boost-headers to 1.55.0nb1

(wiz)

Fix "implicit conversion changes signedness" warnings in installed
header files.
Reported as https://svn.boost.org/trac/boost/ticket/9725
Bump boost-headers PKGREVISION.

(wiz)

Updated devel/xorg-util-macros to 1.19.0

(wiz)

Update to 1.19.0:

Alan Coopersmith (1):
      XORG_COMPILER_FLAGS: Add -Wlogical-op to default warning set

Arnaud Fontaine (1):
      Add XORG_WITH_M4 macro

Gaetan Nadon (3):
      Bump minimum Autoconf required version to 2.62
      Provide the automake INSTALL file at level 1.11
      Version bump: 1.19.0

Hans de Goede (1):
      XORG_PROG_RAWCPP: Add TRADITIONALCPPFLAGS

(wiz)

Updated x11/xrandr to 1.4.2

(wiz)

Update to 1.4.2:

This minor bugfix release restores the ability to disconnect
providers from each other using "0x0" as the provider XID.  For
example, to disconnect a display offload sink from its source,
use

  xrandr --setprovideroutputsource <sink> 0x0

This release also formats the GUID provided by DisplayPort 1.2
displays in traditional GUID form.

Finally, this release increases the precision of refresh rate
calculations to disambiguate modes with very similar refresh
rates and to improve the accuracy for interlaced modes.

Aaron Plattner (5):
      Split output property printing into a helper function
      Move EDID printing into a helper function
      Special-case printing of the GUID property
      xrandr: document how to disconnect RandR 1.4 providers
      xrandr 1.4.2

Dave Airlie (1):
      xrandr: allow disconnecting of offload and outputs

Ville Syrj辰l辰 (2):
      xrandr: Use more decimal places when printing various rates
      xrandr: Use floating point for VTotal when calculating refresh rate

(wiz)

Updated devel/global to 6.2.11

(wiz)

Update to 6.2.11:

Version 6.2.11 - March 12 2014

[CHANGES]
New facilities:
o global: New -S (--scope) <directory> option. This option prints only tags
  which exist under <directory>. It is similar to the -l option, but different
  from it in that you need not change directory.
o gtags: Now, the skip list allows glob file patterns (*, ?, [...], [!...], [^...]).
o gtags.conf: Added the following glob patterns to the skip list:
  *.orig,*.rej,*.bak,*~,#*#,*.swp,*.tmp,*_flymake.*,*_flymake
o gtags: Now, the -I option supports file list (-f option or gtags.files).
  Instead, idutils 4.5 or later is required.

[FIXED BUGS]
o global: With combination of the --invert-match and --literal, it does not match
  the last line of files. Now it works.
o global: Literal search with the -i option prints an unmatched tag in a certain
  kind of case. Now it works.
o global: 'global -cI' without argument causes segmentation fault.
  Now it works.

(wiz)

Updated audio/libmikmod to 3.3.6

(wiz)

Update to 3.3.6:

libmikmod v3.3.6 was released on 20-Mar-2014.

- UMX loader: Minor corrections to the container parse code with
  relation to upkg file format. (for correctness' sake only...)
- ALSA and ESD drivers: Fixed a dlopen() issue.
- PulseAudio driver: Simplified its command line parser procedure,
  so that it always sets server and sink to some value, NULL or
  otherwise, and fix a very minor memory leak on some occasions.
- Configury, OSS driver: Removed useless check for /dev/sndstat
  presence, added <soundcard.h> check to cover all our bases for
  less common setups.
- Configury, NAS driver: Use the configury detected X header and
  library locations when necessary, not just hardcoded /usr/X11R6/.
- Several configury and makefile clean-ups.
- Added Dreamcast driver from the old 3.2.0b2 dreamcast.es port.

(wiz)

Updated fonts/harfbuzz to 0.9.27

(wiz)

Update to 0.9.27:

Overview of changes leading to 0.9.27
Tuesday, March 18, 2014
=====================================

- Don't use "register" storage class specifier
- Wrap definition of free_langs() with HAVE_ATEXIT
- Add coretext_aat shaper and hb_coretext_face_create() constructor
- If HAVE_ICU_BUILTIN is defined, use hb-icu Unicode callbacks
- Add Myanmar test case from OpenType Myanmar spec
- Only do fallback Hebrew composition if no GPOS 'mark' available
- Allow bootstrapping without gtk-doc
- Use AM_MISSING_PROG for ragel and git
- Typo in ucdn's Makefile.am
- Improve MemoryBarrier() implementation

(wiz)

Updated math/fftw to 3.3.4

(wiz)

Update to 3.3.4:

FFTW 3.3.4

* New functions fftw_alignment_of (to check whether two arrays are
  equally aligned for the purposes of applying a plan) and fftw_sprint_plan
  (to output a description of plan to a string).

* Bugfix in fftw-wisdom-to-conf; thanks to Florian Oppermann for the
  bug report.

* Fixed manual to work with texinfo-5.

* Increased timing interval on x86_64 to reduce timing errors.

* Default to Win32 threads, not pthreads, if both are present.

* Various build-script fixes.

(wiz)

Updated converters/py-zfec to 1.4.24

(wiz)

Update to 1.4.24:
No NEWS found.

(wiz)

Updated devel/py-setuptools to 3.3

(wiz)

Update to 3.3:

---
3.3
---

* Add ``include`` parameter to ``setuptools.find_packages()``.

---
3.2
---

* Pull Request #39: Add support for C++ targets from Cython ``.pyx`` files.
* Issue #162: Update dependency on certifi to 1.0.1.
* Issue #164: Update dependency on wincertstore to 0.2.

(wiz)

Updated time/py-pytz to 2014.2

(wiz)

Update to 2014.2: the usual timezone changes.

(wiz)

Updated textproc/uncrustify to 0.60

(wiz)

Update to 0.60:

Change highlights in uncrustify-0.60 (1 Jan 2013)
-------------------------------------------------------------------------------
Objective C support was greatly enhanced in this release.
Many bugs were squashed and many options were added.

New options:
- align_oc_msg_colon_first
- align_pp_define_together
- indent_ctor_init_leading
- indent_func_def_force_col1
- indent_oc_block
- indent_oc_block_msg
- indent_oc_msg_colon
- ls_code_width
- nl_after_annotation
- nl_before_return
- nl_between_annotation
- nl_brace_struct_var
- nl_oc_msg_args
- nl_oc_msg_leave_one_liner
- nl_remove_extra_newlines
- nl_scope_brace
- nl_typedef_blk_end
- nl_typedef_blk_in
- nl_typedef_blk_start
- nl_unittest_brace
- nl_var_def_blk_end
- nl_var_def_blk_in
- nl_var_def_blk_start
- nl_version_brace
- sp_after_for_colon
- sp_after_oc_dict_colon
- sp_after_oc_msg_receiver
- sp_after_oc_property
- sp_after_throw
- sp_after_tparen_close
- sp_annotation_paren
- sp_before_for_colon
- sp_before_oc_dict_colon
- sp_before_pp_stringify
- sp_before_template_paren
- sp_cpp_lambda_assign
- sp_cpp_lambda_paren
- sp_extern_paren
- sp_inside_sparen_open
- sp_inside_tparen
- sp_permit_cpp11_shift
- sp_ptr_star_paren

Changed default:
- sp_pp_stringify = ignore (was add)

(wiz)

Updated audio/libwildmidi to 0.3.5

(wiz)

Update to 0.3.5:

0.3.5:

While we've not added any additional functionality, we have fixed
many bugs from internal testing (clang static analysis and valgrind)
and also downstream distribution issues. We've also reduced the
memory thrashing dramatically. Going from 16,000 reallocations and
about 400MiB memory usage when playing a 2 minute midi down to
about 600 reallocations and only 8MiB memory usage!

What is new in this release:

    Greatly reduced the heap usage (was a regression introduced in
    0.2.3)
    OpenAL support: Fixed audio output on big-endian systems. Fixed
    audio skips at song start.
    OSS support: No longer uses mmap mode for better compatibility.
    This gains us NetBSD and OpenBSD support.
    Worked around an invalid memory read found by valgrind when
    playing Beethoven's Fur Elise.rmi at 44100 Hz using the old
    MIDIA patch-set from 1994.
    Build fixes for MSVC. Revised visibility attributes usage.

0.3.4:

This release marks the end of further development in the 0.3 branch.
We will do bug-fixes to help stabilize platforms, but no new
functionality will be added after this.

What is new in this release:

    OpenAL support: This gains us OSX and other platforms that
    OpenAL supports for sound output!
    DOS DJGPP support: This goes a long way to helping other DOS
    based applications like UHexen2.
    MinGW support: This gains us win32 and win64 support using this
    toolchain.
    Fedora support: We are now ready to see this get pushed upstream
    to Fedora.
    New portable file and path-name system to handle cross-platform
    support.
    Support for Debian/kFreeBSD, Debian/Hurd and other Debian archs.
    Many bug fixes, code clean-ups and cosmetic fixes.

0.3.3:

This is the first package to be officially uploaded to Debian.

    default to hidden visibility and only export our API functions
    windows lean and mean to help compile times on Windows
    cli and xcode work now on OSX
    better FreeBSD support

Supported platforms are Debian, FreeBSD, Windows and OSX (but only
for WAV output)

0.3.2:

Now with Microsoft Visual Studio support.

(wiz)

Updated devel/py-flakes to 0.8

(wiz)

Update to 0.8:

0.8.0 (2014-03-22):
  - Adapt for the AST in Python 3.4.
  - Fix caret position on SyntaxError.
  - Fix crash on Python 2.x with some doctest SyntaxError.
  - Add tox.ini.
  - The `PYFLAKES_NODOCTEST` environment variable has been replaced with the
    `PYFLAKES_DOCTEST` environment variable (with the opposite meaning).
    Doctest checking is now disabled by default; set the environment variable
    to enable it.
  - Correctly parse incremental `__all__ += [...]`.
  - Catch return with arguments inside a generator (Python <= 3.2).
  - Do not complain about `_` in doctests.
  - Drop deprecated methods `pushFunctionScope` and `pushClassScope`.

(wiz)

Updated devel/py-cffi to 0.8.2

(wiz)

Update to 0.8.2:
Changes not found.

(wiz)

Updated x11/libFS to 1.0.6

(wiz)

Update to 1.0.6:

libFS is the protocol binding library used by clients of X Font Servers (xfs),
such as xfsinfo, fslsfonts, and the X servers themselves.

This minor maintenance release includes fixes for libxtrans 1.3 compatibility,
removal of dead code, and cleanups of issues found by static analyzers.

Adam Jackson (1):
      Fix build with xtrans 1.3

Alan Coopersmith (14):
      Fix typo in readme ("xfslsfonts" should be "fslsfonts")
      Require ANSI C89 pre-processor, drop pre-C89 token pasting support
      Fix a couple comment typos
      Drop Cray (WORD64/MUSTCOPY) support
      Add AC_USE_SYSTEM_EXTENSIONS to expose non-standard extensions
      Reduce scope & remove unneeded assignment of defaultp in FSGetErrorText()
      If EAGAIN == EWOULDBLOCK, only need to check errno for one of them
      Remove unused internal helper _FSGetHostname
      Remove unused internal helper _FSAllocScratch
      Remove unused internal helper _FSEventsQueued
      Remove unused internal helper _FSWireToEvent
      Replace malloc(strlen + 1); strcpy() with strdup()
      Use strlcpy instead of strcpy/strncpy if it is available
      libFS 1.0.6

Jeremy Huddleston Sequoia (2):
      Update error loop to correctly handle error in i==0 case
      FSOpenServer: Fix double-free in error path

(wiz)

Updated devel/physfs to 2.0.3

(wiz)

Update to 2.0.3:

2.0.3:

- Source code now packaged in a .tar.bz2 instead of .tar.gz.
- Fixed "make docs" for out-of-tree builds.
- No longer builds annoying wxWidgets test program by default.
- Fixed logic bug in UTF-8 string processing.
- Fixed infinite loop on some .zip files with symlinks.
- Enabled CD-ROM support on Haiku OS.
- Fixed Win64 support.
- Fixed OpenBSD support.
- Fixed building of readline support in test program.
- Fixed .zip archiver losing files in some cases.
- Corrected copyright dates in license.
- Other minor tweaks and fixes.

2.0.1:

- Fixed building with gcc 4.4.
- Workaround bug with lzma archiver not being able to reopen the same
file twice.
- Fixed reading of /proc/self/exe on Linux when calculating base dir.
Fixes unexpected PHYSFS_init() failures.
- Fixed cross-compiling problems.
- Fixed building with GCC on several platforms that don't support the
-fvisibility=hidden option.
- Added support for building with the Mac OS X 10.6 SDK.
- Fixed latin1-to-utf8 encoding conversion issue.
- Fixed logic error in the OS/2 code.
- Replaced some C++-style comments with C-style equivalents.
- Fixed compiler warning on newer Linux systems.
- Added support for GNU/Hurd.
- Doxygen corrections.
- Setting $HOME to a missing directory on Unix will no longer cause
PHYSFS_init() to fail.
- Improved CD-ROM/DVD-ROM detection for Linux systems.
- Corrected file sharing permissions on Windows and PocketPC.
- Other minor fixes.

2.0:

PhysicsFS 2.0 offers many improvements over the 1.0 branch.

- New CMake-based build system. The autotools mess is gone, as are all
the specialized project files for various toolchains. We now maintain
one text file that describes the project, and use CMake
(http://www.cmake.org/) to generate real project files from there...it
produces standard Makefiles for most Unixes and BeOS, but also project
files for KDevelop, Xcode, Visual Studio 6/7/8, Watcom, Borland, and
other build tools on Windows and Mac OS X. If your platform or build
tool isn't supported, energy is better spent on enhancing CMake than
creating another project file for PhysicsFS. OS/2 still uses a batch
file to build for now, but everything else is either using CMake or will
be dropped.

- New archiver: lzma support (7zip archives), thanks to Dennis Schridde.

- Unicode support! All PhysicsFS APIs that deal with strings now expect
them to be UTF-8 encoded, and will convert behind the scenes as
appropriate, so eventually your UTF-8 encoded Japanese characters will
become 2-byte WCHAR strings when looking for filenames on a Windows NTFS
disc, etc. Windows will try to use the appropriate codepage on
Win95/98/ME, and use the actual Unicode entry points on NT/XP/Vista,
CFStrings on Mac OS X, etc. The platform layers in PhysicsFS for all
supported OSes are now Unicode clean, except OS/2 (to be considered).
There are new PhysicsFS APIs to provide conversion between some common
character encodings.

- Applications may now supply their own allocators for PhysicsFS to use
internally. If you don't want to supply one, PhysicsFS uses a reasonable
default for the platform (such as malloc() on Unix, or CoreFoundation
APIs on Mac OS X).

- New API: PHYSFS_mount(). This supercedes PHYSFS_addToSearchPath().
This lets you put your archives at specific points in the interpolated
file system. If you have an archive mounted to "/some/subdir" then it
treats it as if every file in that archive is under the /some/subdir
directory (so /path/x.txt will be accessible at
/some/subdir/path/x.txt). Developers can still use
PHYSFS_addToSearchPath() if source/binary compatibility with PhysicsFS
1.0.x is important, and even mix and match calls with PHYSFS_mount().

- New API: PHYSFS_isInit(), to determine if the library is ready for use
when you don't have access to the results of a previous PHYSFS_init() call.

- New API: PHYSFS_symbolicLinksPermitted(), to determine this state when
you don't control the calls to PHYSFS_permitSymbolicLinks().

- Symlinks are now supported on Windows Vista and later:
PHYSFS_isSymbolicLink() and PHYSFS_permitSymbolicLinks() work with the
native filesystem as expected in the new Windows version without losing
binary compatibility with older Windows releases.

- Public headers no longer use size_t, so they work without any system
headers pre-included.

- Internal mutexes are now recursive on all platforms, which means it's
now safe to call most PHYSFS_* functions from inside an enumeration
callback (including performing more enumerations from inside an
enumeration!)

- Added unarchiver program as an example application, which actually
does enumerations from inside enumerations.  :)

- Added initial shot at a wxWidgets-based test program, to supersede
test_physfs.c ... still a work in progress.

- Mac OS classic support has been dropped. It could be readded if CMake
is enhanced to support CodeWarrior or MPW, and the code moves from
FSSpec to FSRef functions for Unicode support. Mac OS 8/9 support will
remain in the stable 1.0 branch, and Mac OS X is still, of course, fully
supported everywhere.

- Improvements to support Cygwin, Mingw32, and MSYS.

- Mac OS X now has its own Carbon-based code, split out from unix.c,
which helps with functionality like Unicode and recursive mutexes...the
bits in posix.c are still used on OS X, though.

- OS/2 now builds with Innotek GCC and klibc instead of EMX (although
can probably still work with EMX).

- Most systems can make do with PHYSFS_init(NULL) now (but still should
have argv[0] for cases where they can't!). This includes Linux and
systems that present a Linux-like /proc filesystem with /proc/self/exe ...

- Compiles on BeOS again (was broken in 1.1.0). Haiku is now a supported
target platform as well.

- On GCC 4 and later, will build with -fvisibility=hidden, so the only
symbols exported from the library are the public APIs. This makes the
library smaller and faster when built as a shared library, not to
mention prevents namespace pollution.

- Reduced malloc pressure a little more (see __PHYSFS_smallAlloc() in
physfs_internal.h). More to come.

- Other bug fixes, cleanups, refactoring, and improvements. A LOT of
internal code has changed...you can check the Mercurial repository
history for specific details.

(wiz)

Updated security/py-keyring to 3.7

(wiz)

Update to 3.7:

---
3.7
---

* Gnome keyring no longer relies on the GNOME_KEYRING_CONTROL environment
  variable.
* Issue #140: Restore compatibility for older versions of PyWin32.

---
3.6
---

* `Pull Request #1 (github) <https://github.com/jaraco/keyring/pull/1>`_:
  Add support for packages that wish to bundle keyring by using relative
  imports throughout.

---
3.5
---

* Issue #49: Give the backend priorities a 1.5 multiplier bump when an
  XDG_CURRENT_DESKTOP environment variable matches the keyring's target
  environment.
* Issue #99: Clarified documentation on location of config and data files.
  Prepared the code base to treat the two differently on Unix-based systems.
  For now, the behavior is unchanged.

---
3.4
---

* Extracted FileBacked and Encrypted base classes.
* Add a pyinstaller hook to expose backend modules. Ref #124
* Pull request #41: Use errno module instead of hardcoding error codes.
* SecretService backend: correctly handle cases when user dismissed
  the collection creation or unlock prompt.

---
3.3
---

* Pull request #40: KWallet backend will now honor the ``KDE_FULL_SESSION``
  environment variable as found on openSUSE.

-----
3.2.1
-----

* SecretService backend: use a different function to check that the
  backend is functional. The default collection may not exist, but
  the collection will remain usable in that case.

  Also, make the error message more verbose.

  Resolves https://bugs.launchpad.net/bugs/1242412.

---
3.2
---

* Issue #120: Invoke KeyringBackend.priority during load_keyring to ensure
  that any keyring loaded is actually viable (or raises an informative
  exception).

* File keyring:

  - Issue #123: fix removing items.
  - Correctly escape item name when removing.
  - Use with statement when working with files.

* Add a test for removing one item in group.

* Issue #81: Added experimental support for third-party backends. See
  `keyring.core._load_library_extensions` for information on supplying
  a third-party backend.

---
3.1
---

* All code now runs natively on both Python 2 and Python 3, no 2to3 conversion
  is required.
* Testsuite: clean up, and make more use of unittest2 methods.

-----
3.0.5
-----

* Issue #114: Fix logic in pyfs detection.

-----
3.0.4
-----

* Issue #114: Fix detection of pyfs under Mercurial Demand Import.

-----
3.0.3
-----

* Simplified the implementation of ``keyring.core.load_keyring``. It now uses
  ``__import__`` instead of loading modules explicitly. The ``keyring_path``
  parameter to ``load_keyring`` is now deprecated. Callers should instead
  ensure their module is available on ``sys.path`` before calling
  ``load_keyring``. Keyring still honors ``keyring-path``. This change fixes
  Issue #113 in which the explicit module loading of keyring modules was
  breaking package-relative imports.

-----
3.0.2
-----

* Renamed ``keyring.util.platform`` to ``keyring.util.platform_``. As reported
  in Issue #112 and `mercurial_keyring #31
  <https://bitbucket.org/Mekk/mercurial_keyring/issue/31>`_ and in `Mercurial
  itself <http://bz.selenic.com/show_bug.cgi?id=4029>`_, Mercurial's Demand
  Import does not honor ``absolute_import`` directives, so it's not possible
  to have a module with the same name as another top-level module. A patch is
  in place to fix this issue upstream, but to support older Mercurial
  versions, this patch will remain for some time.

-----
3.0.1
-----

* Ensure that modules are actually imported even in Mercurial's Demand Import
  environment.

---
3.0
---

* Removed support for Python 2.5.
* Removed names in ``keyring.backend`` moved in 1.1 and previously retained
  for compatibilty.

-----
2.1.1
-----

* Restored Python 2.5 compatibility (lost in 2.0).

---
2.1
---

*  Issue #10: Added a 'store' attribute to the OS X Keyring, enabling custom
  instances of the KeyringBackend to use another store, such as the
  'internet' store. For example::

      keys = keyring.backends.OS_X.Keyring()
      keys.store = 'internet'
      keys.set_password(system, user, password)
      keys.get_password(system, user)

  The default for all instances can be set in the class::

      keyring.backends.OS_X.Keyring.store = 'internet'

*  GnomeKeyring: fix availability checks, and make sure the warning
  message from pygobject is not printed.

*  Fixes to GnomeKeyring and SecretService tests.

-----
2.0.3
-----

*  Issue #112: Backend viability/priority checks now are more aggressive about
  module presence checking, requesting ``__name__`` from imported modules to
  force the demand importer to actually attempt the import.

-----
2.0.2
-----

*  Issue #111: Windows backend isn't viable on non-Windows platforms.

-----
2.0.1
-----

*  Issue #110: Fix issues with ``Windows.RegistryKeyring``.

---
2.0
---

*  Issue #80: Prioritized backend support. The primary interface for Keyring
  backend classes has been refactored to now emit a 'priority' based on the
  current environment (operating system, libraries available, etc). These
  priorities provide an indication of the applicability of that backend for
  the current environment. Users are still welcome to specify a particular
  backend in configuration, but the default behavior should now be to select
  the most appropriate backend by default.

-----
1.6.1
-----

* Only include pytest-runner in 'setup requirements' when ptr invocation is
  indicated in the command-line (Issue #105).

---
1.6
---

*  GNOME Keyring backend:

  - Use the same attributes (``username`` / ``service``) as the SecretService
    backend uses, allow searching for old ones for compatibility.
  - Also set ``application`` attribute.
  - Correctly handle all types of errors, not only ``CANCELLED`` and ``NO_MATCH``.
  - Avoid printing warnings to stderr when GnomeKeyring is not available.

* Secret Service backend:

  - Use a better label for passwords, the same as GNOME Keyring backend uses.

---
1.5
---

*  SecretService: allow deleting items created using previous python-keyring
  versions.

  Before the switch to secretstorage, python-keyring didn't set "application"
  attribute. Now in addition to supporting searching for items without that
  attribute, python-keyring also supports deleting them.

*  Use ``secretstorage.get_default_collection`` if it's available.

  On secretstorage 1.0 or later, python-keyring now tries to create the
  default collection if it doesn't exist, instead of just raising the error.

*  Improvements for tests, including fix for Issue #102.

---
1.4
---

* Switch GnomeKeyring backend to use native libgnome-keyring via
  GObject Introspection, not the obsolete python-gnomekeyring module.

---
1.3
---

* Use the `SecretStorage library <https://pypi.python.org/pypi/SecretStorage>`_
  to implement the Secret Service backend (instead of using dbus directly).
  Now the keyring supports prompting for and deleting passwords. Fixes #69,
  #77, and #93.
* Catch `gnomekeyring.IOError` per the issue `reported in Nova client
  <https://bugs.launchpad.net/python-novaclient/+bug/1116302>`_.
* Issue #92 Added support for delete_password on Mac OS X Keychain.

-----
1.2.3
-----

* Fix for Encrypted File backend on Python 3.
* Issue #97 Improved support for PyPy.

-----
1.2.2
-----

* Fixed handling situations when user cancels kwallet dialog or denies access
  for the app.

-----
1.2.1
-----

* Fix for kwallet delete.
* Fix for OS X backend on Python 3.
* Issue #84: Fix for Google backend on Python 3 (use of raw_input not caught
  by 2to3).

---
1.2
---

* Implemented delete_password on most keyrings. Keyring 2.0 will require
  delete_password to implement a Keyring. Fixes #79.

-----
1.1.2
-----

* Issue #78: pyfilesystem backend now works on Windows.

-----
1.1.1
-----

* Fixed MANIFEST.in so .rst files are included.

---
1.1
---

This is the last build that will support installation in a pure-distutils
mode. Subsequent releases will require setuptools/distribute to install.
Python 3 installs have always had this requirement (for 2to3 install support),
but starting with the next minor release (1.2+), setuptools will be required.

Additionally, this release has made some substantial refactoring in an
attempt to modularize the backends. An attempt has been made to maintain 100%
backward-compatibility, although if your library does anything fancy with
module structure or clasess, some tweaking may be necessary. The
backward-compatible references will be removed in 2.0, so the 1.1+ releases
represent a transitional implementation which should work with both legacy
and updated module structure.

* Added a console-script 'keyring' invoking the command-line interface.
* Deprecated _ExtensionKeyring.
* Moved PasswordSetError and InitError to an `errors` module (references kept
  for backward-compatibility).
* Moved concrete backend implementations into their own modules (references
  kept for backward compatibility):

  - OSXKeychain -> backends.OS_X.Keyring
  - GnomeKeyring -> backends.Gnome.Keyring
  - SecretServiceKeyring -> backends.SecretService.Keyring
  - KDEKWallet -> backends.kwallet.Keyring
  - BasicFileKeyring -> backends.file.BaseKeyring
  - CryptedFileKeyring -> backends.file.EncryptedKeyring
  - UncryptedFileKeyring -> backends.file.PlaintextKeyring
  - Win32CryptoKeyring -> backends.Windows.EncryptedKeyring
  - WinVaultKeyring -> backends.Windows.WinVaultKeyring
  - Win32CryptoRegistry -> backends.Windows.RegistryKeyring
  - select_windows_backend -> backends.Windows.select_windows_backend
  - GoogleDocsKeyring -> backends.Google.DocsKeyring
  - Credential -> keyring.credentials.Credential
  - BaseCredential -> keyring.credentials.SimpleCredential
  - EnvironCredential -> keyring.credentials.EnvironCredential
  - GoogleEnvironCredential -> backends.Google.EnvironCredential
  - BaseKeyczarCrypter -> backends.keyczar.BaseCrypter
  - KeyczarCrypter -> backends.keyczar.Crypter
  - EnvironKeyczarCrypter -> backends.keyczar.EnvironCrypter
  - EnvironGoogleDocsKeyring -> backends.Google.KeyczarDocsKeyring
  - BasicPyfilesystemKeyring -> backends.pyfs.BasicKeyring
  - UnencryptedPyfilesystemKeyring -> backends.pyfs.PlaintextKeyring
  - EncryptedPyfilesystemKeyring -> backends.pyfs.EncryptedKeyring
  - EnvironEncryptedPyfilesystemKeyring -> backends.pyfs.KeyczarKeyring
  - MultipartKeyringWrapper -> backends.multi.MultipartKeyringWrapper

* Officially require Python 2.5 or greater (although unofficially, this
  requirement has been in place since 0.10).

---
1.0
---

This backward-incompatible release attempts to remove some cruft from the
codebase that's accumulated over the versions.

* Removed legacy file relocation support. `keyring` no longer supports loading
  configuration or file-based backends from ~. If upgrading from 0.8 or later,
  the files should already have been migrated to their new proper locations.
  If upgrading from 0.7.x or earlier, the files will have to be migrated
  manually.
* Removed CryptedFileKeyring migration support. To maintain an existing
  CryptedFileKeyring, one must first upgrade to 0.9.2 or later and access the
  keyring before upgrading to 1.0 to retain the existing keyring.
* File System backends now create files without group and world permissions.
  Fixes #67.

------
0.10.1
------

* Merged 0.9.3 to include fix for #75.

----
0.10
----

* Add support for using `Keyczar <http://www.keyczar.org/>`_ to encrypt
  keyrings. Keyczar is "an open source cryptographic toolkit designed to make
  it easier and safer for developers to use cryptography in their
  applications."
* Added support for storing keyrings on Google Docs or any other filesystem
  supported by pyfilesystem.
* Fixed issue in Gnome Keyring when unicode is passed as the service name,
  username, or password.
* Tweaked SecretService code to pass unicode to DBus, as unicode is the
  preferred format.
* Issue #71 - Fixed logic in CryptedFileKeyring.
* Unencrypted keyring file will be saved with user read/write (and not group
  or world read/write).

-----
0.9.3
-----

* Ensure migration is run when get_password is called. Fixes #75. Thanks to
  Marc Deslauriers for reporting the bug and supplying the patch.

-----
0.9.2
-----

* Keyring 0.9.1 introduced a whole different storage format for the
  CryptedFileKeyring, but this introduced some potential compatibility issues.
  This release incorporates the security updates but reverts to the INI file
  format for storage, only encrypting the passwords and leaving the service
  and usernames in plaintext. Subsequent releases may incorporate a new
  keyring to implement a whole-file encrypted version. Fixes #64.
* The CryptedFileKeyring now requires simplejson for Python 2.5 clients.

-----
0.9.1
-----

* Fix for issue where SecretServiceBackend.set_password would raise a
  UnicodeError on Python 3 or when a unicode password was provided on Python
  2.
* CryptedFileKeyring now uses PBKDF2 to derive the key from the user's
  password and a random hash. The IV is chosen randomly as well. All the
  stored passwords are encrypted at once. Any keyrings using the old format
  will be automatically converted to the new format (but will no longer be
  compatible with 0.9 and earlier). The user's password is no longer limited
  to 32 characters. PyCrypto 2.5 or greater is now required for this keyring.

---
0.9
---

* Add support for GTK 3 and secret service D-Bus. Fixes #52.
* Issue #60 - Use correct method for decoding.

-----
0.8.1
-----

* Fix regression in keyring lib on Windows XP where the LOCALAPPDATA
  environment variable is not present.

---
0.8
---

* Mac OS X keyring backend now uses subprocess calls to the `security`
  command instead of calling the API, which with the latest updates, no
  longer allows Python to invoke from a virtualenv. Fixes issue #13.
* When using file-based storage, the keyring files are no longer stored
  in the user's home directory, but are instead stored in platform-friendly
  locations (`%localappdata%\Python Keyring` on Windows and according to
  the freedesktop.org Base Dir Specification
  (`$XDG_DATA_HOME/python_keyring` or `$HOME/.local/share/python_keyring`)
  on other operating systems). This fixes #21.

*Backward Compatibility Notice*

Due to the new storage location for file-based keyrings, keyring 0.8
supports backward compatibility by automatically moving the password
files to the updated location. In general, users can upgrade to 0.8 and
continue to operate normally. Any applications that customize the storage
location or make assumptions about the storage location will need to take
this change into consideration. Additionally, after upgrading to 0.8,
it is not possible to downgrade to 0.7 without manually moving
configuration files. In 1.0, the backward compatibilty
will be removed.

-----
0.7.1
-----

* Removed non-ASCII characters from README and CHANGES docs (required by
  distutils if we're to include them in the long_description). Fixes #55.

---
0.7
---

* Python 3 is now supported. All tests now pass under Python 3.2 on
  Windows and Linux (although Linux backend support is limited). Fixes #28.
* Extension modules on Mac and Windows replaced by pure-Python ctypes
  implementations. Thanks to Jerome Laheurte.
* WinVaultKeyring now supports multiple passwords for the same service. Fixes
  #47.
* Most of the tests don't require user interaction anymore.
* Entries stored in Gnome Keyring appears now with a meaningful name if you try
  to browser your keyring (for ex. with Seahorse)
* Tests from Gnome Keyring no longer pollute the user own keyring.
* `keyring.util.escape` now accepts only unicode strings. Don't try to encode
  strings passed to it.

-----
0.6.2
-----

* fix compiling on OSX with XCode 4.0

-----
0.6.1
-----

* Gnome keyring should not be used if there is no DISPLAY or if the dbus is
  not around (https://bugs.launchpad.net/launchpadlib/+bug/752282).

---
0.6
---

* Added `keyring.http` for facilitating HTTP Auth using keyring.

* Add a utility to access the keyring from the command line.

(wiz)

Updated devel/gnome-common to 3.12.0

(wiz)

Update to 3.12.0:

    version 3.12.0

    gnome-autogen.sh: Fix check if stdin is a tty

    Commit 7f977b021462 introduced a check if stdin is ins't a tty, and
    prevent printing terminal codes in this case. But the check was wrong
    and always failed.

    https://bugzilla.gnome.org/show_bug.cgi?id=726892

    gnome-autogen: require automake 1.11.2

    This is the version of automake that introduced ACLOCAL_PATH support, so
    make sure that we have it before attempting to configure the package.

    gnome-common: use ACLOCAL_PATH over ACLOCAL_FLAGS

    jhbuild has set this variable for more than two years.

    https://bugzilla.gnome.org/show_bug.cgi?id=726208

    gnome-autogen.sh: correct IFS value

    Due to a likely copy-paste mistake, the default value of IFS was being
    set to 9 spaces instead of a tab and a space.  Fix that.

    https://bugzilla.gnome.org/show_bug.cgi?id=726208

    gnome-autogen.sh: Check for appdata-xml.m4 with APPDATA_XML

    compiler-flags: Simplify a case statement slightly

    compiler-flags: Replace deprecated AC_HELP_STRING with AS_HELP_STRING

    Read the dirlist file unconditionally

(wiz)

Added devel/py-pretend version 1.0.8

(wiz)

+ py-pretend.

(wiz)

Import py27-pretend-1.0.8 as devel/py-pretend.

Pretend is a library to make stubbing with Python easier.

What is stubbing?

Stubbing is a technique for writing tests. You may hear the term
mixed up with mocks, fakes, or doubles. Basically a stub is an
object that returns pre-canned responses, rather than doing any
computation.

(wiz)

Added textproc/pod2mdoc version 0.0.8

(wiz)

Added textproc/docbook2mdoc version 0.0.5

(wiz)

+ docbook2mdoc, pod2mdoc.

(wiz)

Import docbook2mdoc-0.0.5 as textproc/docbook2mdoc.

The docbook2mdoc utility is a converter from DocBook into mdoc.
Unlike most DocBook utilities, it's a standalone ISC-licensed ISO
C utility that should compile on any modern UNIX system.

(wiz)

Import pod2mdoc-0.0.8 as textproc/pod2mdoc.

The pod2mdoc utility is a converter from POD into mdoc. It's meant
to operate like pod2man; however, it doesn't require a Perl
installation: pod2mdoc is a standalone ISC-licensed ISO C utility
and should compile on any modern UNIX system.

(wiz)

Added fonts/ttf-tlwg version 0.6.0

(wiz)

+ ttf-tlwg

(wiz)

Import ttf-tlwg-0.6.0 as fonts/ttf-tlwg.

Fonts-TLWG (formerly ThaiFonts-Scalable) is a collection of Thai
scalable fonts available in free licenses. Its goal is to provide
fonts that conform to existing standards and recommendations, so
that it can be a reference implementation.

So, the fonts have to be modified from their original versions as
needed, according to following references:

* Recommendation for Thai font design, from the National Font Project
* Adobe Glyph naming convention
* Microsoft Typography
* Common practice for Thai Windows fonts

(wiz)

Added time/py-iso8601 version 0.1.10

(wiz)

+ py-iso8601.

(wiz)

Import py27-iso8601-0.1.10 as time/py-iso8601.

This Python module parses the most common forms of ISO 8601 date
strings (e.g. 2007-01-14T20:34:22+00:00) into datetime objects.

(wiz)

Changes 5.5.37:

Functionality Added or Changed

On Solaris, mysql_config --libs now includes -R/path/to/library so that libraries can be found at runtime.

Bugs Fixed

InnoDB: A regression introduced by Bug 14329288 would result in a performance degradation when a compressed table does not fit into memory.

InnoDB: The maximum value for innodb_thread_sleep_delay is now 1000000 microseconds. The previous maximum value (4294967295 microseconds on 32-bit and 18446744073709551615 microseconds on 64-bit) was unnecessarily large. Because the maximum value of innodb_thread_sleep_delay is limited by the value set for innodb_adaptive_max_sleep_delay (when set to a non-zero value), the maximum value for innodb_thread_sleep_delay is now the same as the maximum value for innodb_adaptive_max_sleep_delay.

InnoDB: In debug builds, creating a unique index on a binary column, with input data containing duplicate keys, would cause an assertion.

InnoDB: InnoDB would fail to start when innodb_data_file_path specified the data file size in kilobytes by appending K to the size value.

InnoDB: An insert buffer merge would cause an assertion error due to incorrectly handled ownership information for externally stored BLOBs.

InnoDB: Assertion failure in thread thread_num in file ibuf0ibuf.cc line 4080
InnoDB: Failing assertion: rec_get_deleted_flag(rec, page_is_comp(page))

InnoDB: Decreasing the auto_increment_increment value would have no affect on the next auto-increment value.

Replication: The server did not handle correctly the insertion of a row larger than 4 GB when using row-based replication.

Replication: When using row-based replication, an additional auto-increment column on the slave version of a table was not updated correctly; a zero was inserted instead.

Replication: Statements involving the Performance Schema tables should not be written to the binary log, because the content of these tables is applicable only to a given MySQL Server instance, and may differ greatly between different servers in a replication topology. The database administrator should be able to configure (INSERT, UPDATE, or DELETE) or flush (TRUNCATE TABLE) performance schema tables on a single server without affecting others. However, when replicating from a MySQL 5.5 master to a MySQL 5.5 or later slave, warnings about unsafe statements updating Performance Schema tables were elevated to errors. For MySQL 5.6 and later slaves, this prevented the simultaneous use of performance_schema and GTIDs (see Replication with Global Transaction Identifiers, in the MySQL 5.6 Manual).

This fix causes all updates on tables in the performance_schema database to be filtered on the master and not replicated, regardless of the type of logging that is in effect. Prior to this fix, statements using were handled by being marked as unsafe for replication, which caused warnings during execution; the statements were nonetheless written to the binary log, regardless of the logging format in effect.

Existing replication behavior for tables in the INFORMATION_SCHEMA database is not changed by this fix.

For more information, see MySQL Performance Schema.

Compilation failed if MySQL was configured with CFLAGS set to include a -Werror option with an argument.

A shared libmysqld embedded server library was not built on Linux.

While printing the server version, the mysql client did not check for buffer overflow in a string variable.

Contraction information in a collation could be mishandled, resulting in incorrect decisions about whether a character is part of a contraction, and miscalculation of contraction weights.

DROP TRIGGER succeeded even with the read_only system variable enabled.

Updating the Performance Schema setup_instruments table on a replication master caused a slave to exit.

Due to a race condition, it was possible for two threads to end up with the same query ID for different queries.

When run by root, mysqld --help --verbose exited with a nonzero error code after displaying the help message.

MySQL client programs from a Community Edition distribution could not connect using SSL to a MySQL server from an Enterprise Edition. This was due to a difference in certificate handling by yaSSL and OpenSSL (used for Community and Enterprise, respectively). OpenSSL expected a blank certificate to be sent when not all of the --ssl-ca, --ssl-cert, and --ssl-key options were specified, and yaSSL did not do so. To resolve this, yaSSL has been modified to send a blank certificate when an option is missing.

A deadlock error occurring during subquery execution could cause an assertion to be raised.

The Performance Schema stage/sql/Waiting to get readlock instrument is no longer used and has been removed.

For system variables that take a string value, SET statements permitted an unquoted value, but values that contained dots were parsed incorrectly and only part of the value was assigned. For example, SET GLOBAL slow_query_log_file = my_slow.log assigned the value my_slow. Now such values must be quoted or an error occurs.

On Windows, mysql_install_db.pl could be run only from within the bin directory under the installation directory.

(adam)

Changes 3.8.4.2:
Fix a potential buffer overread that could result when trying to search a corrupt database file.

(adam)

Changes 0.7.5:
* Upgraded libpng to version 1.6.10-optipng.
* Upgraded zlib to version 1.2.8-optipng.
! Fixed various build issues with libpng-1.5 and libpng-1.6.
* Allowed the handling of huge image files (> millions of pixels
  per row or column) to be independent of the libpng version.
+ Allowed the option -preserve to save the file ownership (UID/GID)
  on Unix.
! Fixed a build issue with Clang.
+ Added various enhancements to the configure script.

(adam)

Note the end of the freeze in preparation for the pkgsrc-2014Q1 branch

(agc)

Start the list of changes on the pkgsrc-2014Q1 branch off with its creation

(agc)