Updated multimedia/gstreamer1 to 1.4.3

(wiz)

Update gstreamer to 1.4.3:

Note that this announcement includes everything from 1.4.2 too, which was
never officially released as some critical bugs were found.

Bug reports fixed in this release:

GStreamer core:
      * 734412 : multiqueue: The buffering logic can lead to a pipeline stuck in PAUSED forever
      * 735574 : buffer: do not touch memory tag flag when copying buffer flags
      * 736295 : multiqueue: posts buffering message holding lock
      * 736424 : query: add annotations to gst_query_set_nth_allocation_pool
      * 736680 : basesrc: possible pool and allocator leak in prepare_allocation()
      * 736736 : query: add annotations to gst_query_add_allocation_pool
      * 736813 : typefindelement leaks sticky events upon flush_stop
      * 737102 : queue: Do not hold GST_QUEUE_LOCK while posting ERROR messages
      * 737133 : Missing gstconfig.h include

GStreamer Plugins Base:
      * 732908 : audioresample: skips samples unless input buffers have correct size
      * 727255 : playbin: filter out buffering messages when switching URI and the previous URI is still playing
      * 729811 : output-selector: test example in gstreamer plugin base is giving " Internal data flow error "
      * 735569 : rtspconnection: Crash due to no protection of watchs readsrc
      * 735748 : playbin: can't play an avi file in totem with audio-filter=scaletempo
      * 735800 : textoverlay: Two textoverlay in sequence fail to negotiate (regression)
      * 735844 : basetextoverlay/pango: overlay negotiation fails when it should not
      * 735952 : videorate: GstStructure refcount critical message
      * 736071 : audiobasesink: Don't hold object lock while calling into other objects like the clock
      * 736118 : videofilter: The buffer is not writable in transform_frame_ip
      * 736739 : audiocdsrc: do not leak uid after parsing TOC select event
      * 736779 : typefind: h265 IRAP picture always true
      * 736788 : audiodecoder: leaks events
      * 736796 : videoencoder: do not leak events when flushing them
      * 736861 : playbin: Reference count bug
      * 736679 : videodecoder: do not leak pool and allocator in error case
      * 736969 : queue2: dead lock when buffering
      * 709868 : Keep still meaningfull pending events on FLUSH_STOP

GStreamer Plugins Good:
      * 719359 : vp8dec: Doesn't handle changes in resolution
      * 733607 : v4l2transform: Rank should have been NONE
      * 734266 : vp8dec: fails when input format changes
      * 735520 : aacparse: skip valid ADTS/LOAS frames
      * 735804 : smpte: Creates incomplete raw video caps
      * 735833 : matroskademux: parse error at end of file
      * 735859 : videomixer: Dynamically changing the FPS leads to an incorrect buffer time
      * 736192 : avidemux: some AVI files crash (regression)
      * 736266 : wavparse: error in reading adtl chunk
      * 736384 : v4l2sink: pool not unreffed after usage
      * 736670 : v4l2allocator: O_CLOEXEC needs _GNU_SOURCE
      * 736805 : multipartdemux leaks new stream events
      * 736807 : rtpbin: pad leaked in error case
      * 735660 : v4l2: fix new v4l2 code not working with certain devices (regression)
      * 736944 : videoscale: vs_image_scale_4tap_Y offset should use stride to calculate buffer offset
      * 737219 : flacparse:  When generating headers, leave total_samples at 0 if upstream duration query returns GST_CLOCK_TIME_NONE.

GStreamer Plugins Bad:
      * 735861 : dataurisrc: make src thread safe
      * 736090 : aiffparse: duplicate else-if condition
      * 736390 : tsdemux: plug for a memory leak
      * 736426 : mpegpsmux: memory leak with h264/avc stream
      * 736474 : vc1parse: malformed sequence layer header and STRUCT_C
      * 736490 : tsdemux: fix overflow of packet_length field of PESHeader
      * 736729 : glmixer: do not leak pool in error cases
      * 736730 : gltestsrc: do not leak pool in error cases
      * 736731 : openni2src: do not leak pool
      * 736732 : glfilter: do not leak pool in error cases
      * 736733 : vdpdecoder: do not leak pool
      * 736735 : waylandsink: do not leak buffer pool in error case
      * 736750 : vc1parse: fix sequence-layer/frame-layer endianness
      * 736871 : codecparsers_vc1: sequence-layer parser is broken due to endianness issue.
      * 736919 : hlsdemux: attempt to unlock an already unlocked mutex in gst_hls_demux_change_playlist
      * 736951 : vc1parse: initialize sent_codec_tag before using it

GStreamer Plugins Ugly:
      * 736060 : asfdemux: add GUID for ASF_Metadata_Library_Object

GStreamer libav Plugins:
      * 734661 : avviddec: After draining frames, flush the libav decoder
      * 736515 : avviddec: keep draining buffers from libav until libav says so
      * 737144 : avauddec: keep draining buffers from libav until libav says so

GStreamer RTSP Server:
      * 735570 : Race condition between close() and handle_tunnel() causing crash
      * 736017 : Sequence number is not monotonic after PAUSE command

(wiz)

Mark SunOS as NOT_FOR_BULK_PLATFORM, currently the build spins indefinitely
on CPU.

(jperkin)

Updated databases/redis to 2.8.17

(fhajny)

Update redis to 2.8.17.

--[ Redis 2.8.17 ] Release date: 19 Sep 2014

# UPGRADE URGENCY: HIGH for Redis Sentinel.
                  LOW for Redis Server (unmodified compared to 2.8.16).

* [FIX] Resolved a memory leak in the hiredis library causing a memory leak
        in Redis Sentinel when a monitored instance or another Sentinel is
        unavailable. Every reconnection attempt will leak a small amount of
        memory, but in the long run the process can reach a considerable size.

--[ Redis 2.8.16 ] Release date: 16 Sep 2014

# UPGRADE URGENCY: HIGH for Redis if you are using 2.8.15 + AOF.
                  LOW for Sentinel.

* [FIX] The ability to load truncated AOF files introduced with Redis 2.8.15
        contains a bug fixed in this release: after loading the file was not
        truncated to the last valid command, so the new commands are appended
        after a non well formed command. This means that:

        1) The first AOF rewrite triggered by the server will automatically
          fix the problem.
        2) However, if the server is restarted before the rewrite, Redis may
          not be able to load the file and you need to manually fix it.

        In order to fix a corrupted file you should start the redis-check-aof
        utility WITHOUT the --fix option, just to check the offset where the
        corruption is found. Around the offset reported by the check utility
        you'll find, inside your AOF file, a command which is not complete
        according to the Redis protocol. Just remove this incomplete command
        leafing the file unaltered before and after the offending command,
        and restart the server.

        IMPORTANT #1: Redis 2.8.15 is the only stable version of Redis with
        this bug so probably no actual real-world problem happened since the
        problem is automatically fixed at the first automatic AOF rewrite.

        IMPORTANT #2: Before upgrading to Redis 2.8.16, if you are using Redis
        2.8.15 with AOF enabled, make sure to trigger a manual AOF rewrite
        using the BGREWRITEAOF command.

* [FIX] SAVE is no longer propagated to AOF / slaves.

--[ Redis 2.8.15 ] Release date: 12 Sep 2014

# UPGRADE URGENCY: LOW for Redis, HIGH for Sentinel.

* [FIX] Sentinel critical bug fixed: the absolute majority was computed in a
        wrong way because of a programming error. Now the implementation does
        what the specification says and the majority to authorize a failover
        (that should not be confused with the ODOWN quorum) is the majority of
        *all* the Sentinels ever seen for a given master, regardless of their
        current state.
* [FIX] GETRANGE test no longer fails for 32 bit builds (Matt Stancliff).
* [FIX] Limit SCAN latency when the hash table is in an odd state (very few
        populted buckets because rehashing is in progress). (Xiaost and
        Salvatore Sanfilippo)

* [NEW] Redis is now able to load truncated AOF files without requiring a
        redis-check-aof utility run. The default now is to load truncated
        (but apparently not corrupted) AOFs, you can change this in redis.conf.
        (Salvatore Sanfilippo).
* [NEW] Sentinel: ability to announce itself with an arbitrary IP/port to work
        in the context of natted networks. However this is probably still
        not enough since there is no equivalent mechanism for slaves listed
        in the master INFO output. (Dara Kong and Salvatore Sanfilippo)

--[ Redis 2.8.14 ] Release date: 1 Sep 2014

# UPGRADE URGENCY: HIGH for Lua scripting users, the server could crash because
                  of a bug introduced in Redis 2.8.10, otherwise LOW.
                  LOW for Redis Sentinel.

* [FIX] Don't prevent use of shared integers if maxmemory policy is non-LRU.
        (Salvatore Sanfilippo)
* [FIX] Fail SYNC if background save child aborted due to a signal.
        (Yossi Gottlieb)
* [FIX] Different small redis-cli fixes. (Dov Murik, Charsyam, cubicdaiya,
        Kashif Rasul, Jan-Erik Rediger, Matt Stancliff)
* [FIX] AIX compilation fixes. (Siah Lyimo)
* [FIX] A number of other smaller issues.
* [FIX] Improved SIGINT handling (Matt Stancliff, Salvatore Sanfilippo)
* [FIX] Use unsigned types in SDS header to raise limit to 4GB.
        (Matt Stancliff, Salvatore Sanfilippo)
* [FIX] Handle signed/unsigned comparisons with more care around the code.
        (Salvatore Sanfilippo)
* [FIX] Colorized test output fixed to don't change the background color.
        (Mariano P辿rez Rodr鱈guez)
* [FIX] More Sentinel IPv6 fixes. (Eiichi Sato)
* [FIX] Deny CLIENT command in scripts. (Matt Stancliff)
* [FIX] Allow datasets with more than 2 billion of keys, initial work.
* [FIX] Fix a Lua scripting crash by storing the length of the static
        argv when first allocated. (Paddy Byers)

* [NEW] Pub/Sub PING. (Salvatore Sanfilippo)
* [NEW] Much faster ZUNIONSTORE. (Kyle Hubert, Salvatore Sanfilippo)
* [NEW] Faster ll2string() implementation. (Salvatore Sanfilippo)
* [NEW] **WARNING, minor API change**: PUBSUB NUMSUB: return type modified
        to integer. (Matt Stancliff)
* [NEW] redis-benchmark support for AUTH. (CharSyam)

(fhajny)

Add missing pkg-config

(fhajny)

Add libgit2 and ruby-rugged under devel

(fhajny)

Added devel/ruby-rugged version 0.21.0

(fhajny)

Include fontconfig builtin3.mk.

(darcy)

Import devel/ruby-rugged.

Rugged is a library for accessing libgit2 in Ruby. It gives you the speed
and portability of libgit2 with the beauty of the Ruby language.

(fhajny)

Added devel/libgit2 version 0.21.1

(fhajny)

Import devel/libgit2 based on wip/libgit2.

libgit2 is a portable, pure C implementation of the Git core methods provided as
a re-entrant linkable library with a solid API, allowing you to write native
speed custom Git applications in any language which supports C bindings.

(fhajny)

Updated archivers/xz to 5.0.7; lang/perl5 to 5.20.1; devel/cmake to 3.0.2; devel/apr-util to 1.5.4; databases/openldap to 2.4.40; databases/mysql56 to 5.6.21; databases/mysql55 to 5.5.40; print/ghostscript-apgl to 9.15; devel/git to 2.1.2

(adam)

Updated print/cups-filters to 1.0.59

(wiz)

Update to 1.0.59:

CHANGES IN V1.0.59

- cupsfilters.drv: Added PPD file for a Generic IPP Everywhere
  Printer, generating PWG Raster output.
- gstoraster, pdftoraster, imagetoraster: Allow PWG Raster
  output with print queues using a PPD file, using the new
  "PWGRaster" PPD attribute.
- pdftoraster: Removed "cm_disabled" flag in selectConvFunc()
- libcupsfilters: Allowed color management to continue while
  invalid input
- rastertopdf: Streamlined PDF conversion code
- rastertopdf: Invert all CUPS_CSPACE_K documents by default
- foomatic-rip: Clean trailing white space from PPD file lines to
  avoid a segfault caused by it (Bug #1227).

(wiz)

Updated time/py-vdirsyncer to 0.3.0

(wiz)

Update to 0.3.0:

Version 0.3.0
=============

*released on 20 September 2014*

- Add ``verify_fingerprint`` parameter to
  :py:class:`vdirsyncer.storage.HttpStorage`,
  :py:class:`vdirsyncer.storage.CaldavStorage` and
  :py:class:`vdirsyncer.storage.CarddavStorage`,
  see issue `#99`_ and pull request `#106`_.

- Add ``passwordeval`` parameter to :ref:`general_config`, see issue `#108`_
  and pull request `#117`_.

- Emit warnings (instead of exceptions) about certain invalid responses from
  the server, see issue `#113`_.  This is apparently required for compatibility
  with Davmail.

.. _`#99`: https://github.com/untitaker/vdirsyncer/issues/99
.. _`#106`: https://github.com/untitaker/vdirsyncer/pull/106
.. _`#108`: https://github.com/untitaker/vdirsyncer/issues/108
.. _`#113`: https://github.com/untitaker/vdirsyncer/issues/113
.. _`#117`: https://github.com/untitaker/vdirsyncer/pull/117

(wiz)

Updated net/py-twisted to 14.0.2

(wiz)

Update to 14.0.2. To quote upstream:
"No significant changes have been made for this release."
To my understanding, a test has been fixed.

(wiz)

Updated devel/py-test to 2.6.3

(wiz)

Update to 2.6.3:

2.6.3
-----------

- fix issue575: xunit-xml was reporting collection errors as failures
  instead of errors, thanks Oleg Sinyavskiy.

- fix issue582: fix setuptools example, thanks Laszlo Papp and Ronny
  Pfannschmidt.

- Fix infinite recursion bug when pickling capture.EncodedFile, thanks
  Uwe Schmitt.

- fix issue589: fix bad interaction with numpy and others when showing
  exceptions.  Check for precise "maximum recursion depth exceed" exception
  instead of presuming any RuntimeError is that one (implemented in py
  dep).  Thanks Charles Cloud for analysing the issue.

- fix conftest related fixture visibility issue: when running with a
  CWD outside a test package pytest would get fixture discovery wrong.
  Thanks to Wolfgang Schnerring for figuring out a reproducable example.

- Introduce pytest_enter_pdb hook (needed e.g. by pytest_timeout to cancel the
  timeout when interactively entering pdb).  Thanks Wolfgang Schnerring.

- check xfail/skip also with non-python function test items. Thanks
  Floris Bruynooghe.

(wiz)

Updated databases/py-sqlparse to 0.1.12

(wiz)

Update to 0.1.12:

Release 0.1.12 (Sep 20, 2014)
-----------------------------

Bug Fixes
* Fix handling of NULL keywords in aliased identifiers.
* Fix SerializerUnicode to split unquoted newlines (issue131, by Michael Schuller).
* Fix handling of modulo operators without spaces (by gavinwahl).

Enhancements
* Improve parsing of identifier lists containing placeholders.
* Speed up query parsing of unquoted lines (by Michael Schuller).

(wiz)

Updated devel/py-setuptools to 6.0.2

(wiz)

Update to 6.0.2:

-----
6.0.2
-----

* Issue #262: Fixed regression in pip install due to egg-info directories
  being omitted. Re-opens Issue #118.

-----
6.0.1
-----

* Issue #259: Fixed regression with namespace package handling on ``single
  version, externally managed`` installs.

---
6.0
---

* Issue #100: When building a distribution, Setuptools will no longer match
  default files using platform-dependent case sensitivity, but rather will
  only match the files if their case matches exactly. As a result, on Windows
  and other case-insensitive file systems, files with names such as
  'readme.txt' or 'README.TXT' will be omitted from the distribution and a
  warning will be issued indicating that 'README.txt' was not found. Other
  filenames affected are:

    - README.rst
    - README
    - setup.cfg
    - setup.py (or the script name)
    - test/test*.py

  Any users producing distributions with filenames that match those above
  case-insensitively, but not case-sensitively, should rename those files in
  their repository for better portability.
* Pull Request #72: When using ``single_version_externally_managed``, the
  exclusion list now includes Python 3.2 ``__pycache__`` entries.
* Pull Request #76 and Pull Request #78: lines in top_level.txt are now
  ordered deterministically.
* Issue #118: The egg-info directory is now no longer included in the list
  of outputs.
* Issue #258: Setuptools now patches distutils msvc9compiler to
  recognize the specially-packaged compiler package for easy extension module
  support on Python 2.6, 2.7, and 3.2.

---
5.8
---

* Issue #237: ``pkg_resources`` now uses explicit detection of Python 2 vs.
  Python 3, supporting environments where builtins have been patched to make
  Python 3 look more like Python 2.

(wiz)

Updated multimedia/libdvdnav to 5.0.1

(wiz)

Update to 5.0.1. Set LICENSE.

libdvdnav (5.0.1)
* removed assert on unknown VM commands, fixing playback of some DVDs
* fixed a double free in dvdnav_free_dup
* fixed an integer overflow, a data race condition and return inconsistency

libdvdnav (5.0.0)
* Android support
* fixed numerous crashes, assertions and corruptions
* rewrite of the build-system, including silent rules
* clean code, simplify and remove warnings
* fix compilation in C++ applications
* removal of remap .MAP files unused feature
* remove the dvdnavmini library
* fix compilation on OS/2 and Hurd

libdvdnav (4.2.1)
* fixed failed assertion on menu-less DVDs
* fixed playback issues on multi-angle DVDs
* fixed crash when playing a DVD from different region than currently set
  in the DVD drive
* fixed segfaults when reading certain DVDs (including "Jack Reacher",
  "Ghost Protocol", "Inside Man", "The Dictator" and "Tangled")

(wiz)

Add dvdcss.

(wiz)

Some cleanup.

(wiz)

Updated multimedia/libdvdread to 5.0.0

(wiz)

Update to 5.0.0:

libdvdread (5.0.0)
* added support for Android
* fix build support for OS/2
* fix crashes in IFO parsing (Shark Week The Great Bites Collection)

libdvdread (4.9.9)
  * major cleanup
  * rewritten build system, including removal of configure2 and dvdread-config
  * fix numerous crashes
  * added QNX support
  * updated to support libdvdcss 1.3.0
  * improved dlopen support for Win32
  * simplify static linking of libdvdcss

libdvdread (4.2.1)
  * fix huge memory consumption when reading DVDs: "Up" and "Dark Knight"
  * numerous crashes fixed, notably on ARCCOS protected DVDs,
    like Transformers 3, Brave, The Express, Sherlock Holmes
    and Cars 2,
  * fix packed structures with mingw gcc >= 4.7

(wiz)

Fix ABI break (2 enums change value) when I added NetBSD credential support

(prlw1)

Add security/php-ssh2

(fhajny)

Added security/php-ssh2 version 0.12

(fhajny)

Import security/php-ssh2 (based on wip/php-ssh2).

Provides bindings to the functions of libssh2 which implements
the SSH2 protocol.

(fhajny)

Updated converters/help2man to 1.46.3

(wiz)

Update to 1.46.3:

help2man (1.46.3) unstable; urgency=medium

  * Update Serbian translation (thanks to Мирослав Николић).
  * Add Spanish translation (thanks to Antonio Ceballos).

-- Brendan O'Dea <bod@debian.org>  Sun, 21 Sep 2014 19:01:16 +1000

(wiz)

Changes 2.1.2:
* "git push" over HTTP transport had an artificial limit on number of
  refs that can be pushed imposed by the command line length.
* When receiving an invalid pack stream that records the same object
  twice, multiple threads got confused due to a race.
* An attempt to remove the entire tree in the "git fast-import" input
  stream caused it to misbehave.
* Reachability check (used in "git prune" and friends) did not add a
  detached HEAD as a starting point to traverse objects still in use.
* "git config --add section.var val" used to lose existing
  section.var whose value was an empty string.
* "git fsck" failed to report that it found corrupt objects via its
  exit status in some cases.

(adam)

Updated security/libgpg-error to 1.16

(wiz)

Update to 1.16:

Noteworthy changes in version 1.16 (2014-09-18) [C12/A12/R2]
-----------------------------------------------

* Support building for iOS.

* Fixed a prototype mismatch.

* Fix es_fclose for streams opened with "samethread".

(wiz)

Updated sysutils/py-psutil to 2.1.2

(wiz)

Update to 2.1.2:

2.1.2 - 2014-09-21
==================

**Enhancements**

- #407: project moved from Google Code to Github; code moved from Mercurial
  to Git.
- #492: use tox to run tests on multiple python versions.  (patch by msabramo)
- #505: [Windows] distribution as wheel packages.
- #511: new examples/ps.py sample code.

**Bug fixes**

- #340: [Windows] Process.get_open_files() no longer hangs.  (patch by
  Jeff Tang)
- #501: [Windows] disk_io_counters() may return negative values.
- #503: [Linux] in rare conditions Process exe(), open_files() and
  connections() methods can raise OSError(ESRCH) instead of NoSuchProcess.
- #504: [Linux] can't build RPM packages via setup.py
- #506: [Linux] python 2.4 support was broken.
- #522: [Linux] Process.cpu_affinity() might return EINVAL.  (patch by David
  Daeschler)
- #529: [Windows] Process.exe() may raise unhandled WindowsError exception
  for PIDs 0 and 4.  (patch by Jeff Tang)
- #530: [Linux] psutil.disk_io_counters() may crash on old Linux distros
  (< 2.6.5)  (patch by Yaolong Huang)
- #533: [Linux] Process.memory_maps() may raise TypeError on old Linux distros.

(wiz)

Updated audio/musicpd to 0.18.16

(wiz)

Update to 0.18.16:

ver 0.18.16 (2014/09/26)
* fix DSD breakage due to typo in configure.ac

ver 0.18.15 (2014/09/26)
* command
  - list: reset used size after the list has been processed
* fix MixRamp
* work around build failure on NetBSD

(wiz)

Updated databases/py-ldap to 2.4.17

(wiz)

Update to 2.4.17:

Released 2.4.17 2014-09-27

Changes since 2.4.16:

Lib/
* New hook syncrepl_refreshdone() in ldap.syncrepl.SyncReplConsumer
  (thanks to Petr Spacek and Chris Mikkelson)

Modules/
* Added support for getting file descriptor of connection
  with ldap.OPT_DESC

(wiz)

Updated devel/pcre to 8.36

(wiz)

Update to 8.36:

Version 8.36 26-September-2014
------------------------------

1.  Got rid of some compiler warnings in the C++ modules that were shown up by
    -Wmissing-field-initializers and -Wunused-parameter.

2.  The tests for quantifiers being too big (greater than 65535) were being
    applied after reading the number, and stupidly assuming that integer
    overflow would give a negative number. The tests are now applied as the
    numbers are read.

3.  Tidy code in pcre_exec.c where two branches that used to be different are
    now the same.

4.  The JIT compiler did not generate match limit checks for certain
    bracketed expressions with quantifiers. This may lead to exponential
    backtracking, instead of returning with PCRE_ERROR_MATCHLIMIT. This
    issue should be resolved now.

5.  Fixed an issue, which occures when nested alternatives are optimized
    with table jumps.

6.  Inserted two casts and changed some ints to size_t in the light of some
    reported 64-bit compiler warnings (Bugzilla 1477).

7.  Fixed a bug concerned with zero-minimum possessive groups that could match
    an empty string, which sometimes were behaving incorrectly in the
    interpreter (though correctly in the JIT matcher). This pcretest input is
    an example:

      '\A(?:[^"]++|"(?:[^"]*+|"")*+")++'
      NON QUOTED "QUOT""ED" AFTER "NOT MATCHED

    the interpreter was reporting a match of 'NON QUOTED ' only, whereas the
    JIT matcher and Perl both matched 'NON QUOTED "QUOT""ED" AFTER '. The test
    for an empty string was breaking the inner loop and carrying on at a lower
    level, when possessive repeated groups should always return to a higher
    level as they have no backtrack points in them. The empty string test now
    occurs at the outer level.

8.  Fixed a bug that was incorrectly auto-possessifying \w+ in the pattern
    ^\w+(?>\s*)(?<=\w) which caused it not to match "test test".

9.  Give a compile-time error for \o{} (as Perl does) and for \x{} (which Perl
    doesn't).

10. Change 8.34/15 introduced a bug that caused the amount of memory needed
    to hold a pattern to be incorrectly computed (too small) when there were
    named back references to duplicated names. This could cause "internal
    error: code overflow" or "double free or corruption" or other memory
    handling errors.

11. When named subpatterns had the same prefixes, back references could be
    confused. For example, in this pattern:

      /(?P<Name>a)?(?P<Name2>b)?(?(<Name>)c|d)*l/

    the reference to 'Name' was incorrectly treated as a reference to a
    duplicate name.

12. A pattern such as /^s?c/mi8 where the optional character has more than
    one "other case" was incorrectly compiled such that it would only try to
    match starting at "c".

13. When a pattern starting with \s was studied, VT was not included in the
    list of possible starting characters; this should have been part of the
    8.34/18 patch.

14. If a character class started [\Qx]... where x is any character, the class
    was incorrectly terminated at the ].

15. If a pattern that started with a caseless match for a character with more
    than one "other case" was studied, PCRE did not set up the starting code
    unit bit map for the list of possible characters. Now it does. This is an
    optimization improvement, not a bug fix.

16. The Unicode data tables have been updated to Unicode 7.0.0.

17. Fixed a number of memory leaks in pcregrep.

18. Avoid a compiler warning (from some compilers) for a function call with
    a cast that removes "const" from an lvalue by using an intermediate
    variable (to which the compiler does not object).

19. Incorrect code was compiled if a group that contained an internal recursive
    back reference was optional (had quantifier with a minimum of zero). This
    example compiled incorrect code: /(((a\2)|(a*)\g<-1>))*/ and other examples
    caused segmentation faults because of stack overflows at compile time.

20. A pattern such as /((?(R)a|(?1)))+/, which contains a recursion within a
    group that is quantified with an indefinite repeat, caused a compile-time
    loop which used up all the system stack and provoked a segmentation fault.
    This was not the same bug as 19 above.

21. Add PCRECPP_EXP_DECL declaration to operator<< in pcre_stringpiece.h.
    Patch by Mike Frysinger.

(wiz)

Updated net/py-foolscap to 0.7.0

(wiz)

Update to 0.7.0:

* Release 0.7.0 (23-Sep-2014)

** Security Fixes

The "flappserver" feature was found to have a vulnerability in the
service-lookup code which, when combined with an attacker who has the ability
to write files to a location where the flappserver process could read them,
would allow that attacker to obtain control of the flappserver process.

Users who run flappservers should upgrade to 0.7.0, where this was fixed as
part of #226.

Each flappserver runs from a "base directory", and uses multiple files within
the basedir to track the services that have been configured. The format of
these files has changed. The flappserver tool in 0.7.0 remains capable of
reading the old format (safely), but will upgrade the basedir to the new
format when you use "flappserver add" to add a new service. Brand new
servers, created with "flappserver create", will use the new format.

The flappserver tool in 0.6.5 (or earlier) cannot handle this new format, and
will believe that no services have been configured. Therefore downgrading to
an older version of Foolscap will require manual reconstruction of the
configured services.

** Major Changes

UnauthenticatedTub has been deprecated, and will be removed in the next
release (0.8.0). This seldom-used feature provides Foolscap's RPC semantics
without any of the security, and was included to enable the use of Foolscap
without depending upon the (challenging-to-install) PyOpenSSL library.
However, in practice, the lack of a solid dependency on PyOpenSSL has made
installation more difficult for applications that *do* want the security, and
UnauthenticatedTub is a footgun waiting to go off. Foolscap's code and
packaging will be simpler without it. (#67)

** Minor Changes

The "git-foolscap" tools, which make it possible to publish and clone Git
repositories over a Foolscap (flappserver) connection, have been moved from
their hiding place in doc/examples/ into their own project, hosted at
https://github.com/warner/git-foolscap . They will also be published on PyPI,
to enable "pip install git-foolscap".

The documentation was converted from Lore to ReStructuredText (.rst). Thanks
to Koblaid for the patient work. (#148)

The connection-hint parser in 0.7.0 has been changed to handle all TCP forms
of Twisted's "Client Endpoint Descriptor" syntax, including the short
"tcp:127.0.0.1:9999" variant. A future version should handle arbitrary
endpoint descriptors (including Tor and i2p, see #203), but this small step
should improve forward compatibility. (#216, #217)

(wiz)

Updated misc/py-anki2 to 2.0.29

(wiz)

Update to 2.0.29:

Changes in 2.0.29

    Fix "app is damaged" message on OSX Yosemite

    A new "added" graph that shows how many new cards have been added
    to the collection.

    When exporting notes as text, say the number of notes that were
    exported, not the number of cards.

    Allow tags to be separated with an ideographic space character
    (eg. Japanese full width space)

    Allow closing the edit current dialog with Ctrl-Enter.

    Show more info when syncing media, so you can see it窶冱 progressing
    and get an idea of how long is left.

    All the web components in Anki default to accepting focus and
    allowing copying now, which should fix issues with some screens
    not allowing copy to the clipboard, such as the card info screen.

    Fix a bug where <br> tags were being inserted in place of newlines
    when importing with html enabled.

    Fix a bug when trying to recover from a corrupt profile.

    Fix addons being downloaded with double newlines on Windows.

    If an existing anki instance is hung, start a new instance when
    user clicks on anki icon.

    A few translation fixes thanks to Derbeth.

    Disable a better error message when some ssl errors occur.

    Make the tab width in the template editor reasonable.

    Strip any formatting when pasting into the template editor, as
    rich text has no meaning there.

    If the computer has no desktop folder, place collection packages
    in the user窶冱 home directory.

    Fix a bug that was allowing nesting other decks under filtered
    decks.

    Fix a bug that was causing templates to get out of sync with field
    names on

    Don窶冲 jump around in the browser窶冱 card list when deleting cards.

    On Linux systems that don窶冲 have a documents folder defined,
    default to ~/Documents/Anki (in English) for the profile folder.

    We no longer encode and decode %20 and so on when fields are
    edited. This should fix the bug where users were having text with
    percent signs in it being turned into other characters.

    Tags that differ only in case are no longer allowed, as tags were
    sometimes case-sensitive and sometimes not. The case that you use
    the first time you enter will be used for all future cards added
    with that tag. If you want to change the capitalization of that
    tag, you can temporarily remove the tag from all cards and then
    add it again with the new case.

    Ignore >100MB files when syncing instead of having the server
    throw an error.

(wiz)

Note update of the "spamassassin" package to version 3.4.0nb2.

(tron)

Make the DNSRBL based plug-ins work with version 0.76 or newer of the
"p5-Net-DNS" package. Patch taken from the SVN repository.

(tron)

Updated x11/xterm to 311

(wiz)

Update to 311:

Patch #311 - 2014/09/18

* mention  xclip  in manpage as an alternative workaround for copying
  clipboard  data, noting that selectToClipboard (from patch #209) is
  the recommended approach (Debian #639094).
* correct comparison in do_select_regex() of working position against
  starting column.
* correct    initialization  for  regular-expression  feature  of
  exec-selectable and insert-selectable actions (Debian #758633).
* modify  logic  for  exec-formatted  and insert-formatted actions to
  ensure  that  the  formatting  occurs  just  after the selection is
  received (Debian #758633).
* account  for  state  of  reverse-video  in  special  case of cursor
  coloring (Debian #759734).
* fix  a  sign-extension  problem  in  ReGIS  support,  as  well  as
  correcting  a case where ignored-characters were not really ignored
  (patch by Ross Combs).
* Enable the "Escape Sequence" menu entry when an OSC 50 is received.
  Also  disable  it  if the escape sequence specifies no font (Debian
  #760208).
* improve fix for the fontsel menu entry from patch #304: because the
  recovery  used  the  "current  font",  it  would  fail if one first
  selected a valid font, then an invalid font (Debian #760207).
* correct  an  off-by-one  in  limit-check for ScrnLimitChar function
  (report by Egmont Koblinger).

(wiz)

wxmaxima-14.09.0.

(wiz)

Remove wxMaxima; wxmaxima replaces it.

(wiz)

Import wxmaxima-14.09.0 as math/wxmaxima. The casing of the package name
is finally consistent.

This is a graphical UI for the computer algebra system maxima.

Changes compared to math/wxMaxima:

14.09.0:
Bugfixes and compatibility with wxGTK-3.x.

(wiz)

Updated www/p5-CGI to 4.04

(wiz)

Update to 4.04, add a build (test?) dependency.

4.04 2014-09-04

    [ RELEASE NOTES ]
    - this release removes some long deprecated modules/functions and
      includes refactoring to the temporary file handling in CGI.pm. if
      you are doing anything out of the ordinary with regards to temp
      files you should test your code before deploying this update as
      temp files may no longer be stored in previously used locations

    [ REMOVED / DEPRECATIONS ]
    - startform and endform methods removed (previously deprecated, you
      should be using the start_form and end_form methods)
    - both CGI::Apache and CGI::Switch have been removed as these modules
      1) have been deprecated for *years*, and 2) do nothing whatsoever

    [ SPEC / BUG FIXES ]
    - handle multiple values in X-Forwarded-Host header, we follow the
      logic in most other frameworks and take the last value from the list
      (RT #54487)
    - refactor CGITempFile::find_tempdir to use File::Spec->tmpdir
      (related: RT #71799)
    - fix warnings when QUERY_STRING has empty key=value pairs (RT #54511)
    - pad custom 500 status response messages to > 512 for MSIE (RT #81946)
    - make Vars tied hash delete method return the value deleted from the hash
      making it act like perl's delete (RT #51020)

    [ TESTING ]
    - add .travis.yml (https://travis-ci.org)
    - test case for RT #53966 - disallow filenames with ~ char
    - test case for RT #55166 - calling Vars to get the filename does not return
      a filehandle, so this cannot be used in the call to uploadinfo, also
      update documentation for the uploadInfo to show that ->Vars should not be
      used to get the filename for this method
    - fix t/url.t to pass on Win32 platforms that have the SCRIPT_NAME env
      variable set (RT #89992)
    - add procedural call tests for upload and uploadInfo to confirm these work
      as should (RT #91136)

    [ DOCUMENTATION ]
    - tweak perldoc for -utf8 option (RT #54341, thanks to Helmut Richter)
    - explain the HTML generation functions should no longer be used and that
      they may be deprecated in a future release

4.03 2014-07-02

    [ REMOVED / DEPRECATIONS ]
    - the -multiple option to popup_menu is now IGNORED as this did not
      function correctly. If you require a menu with multiple selections
      use the scrolling_list method. (RT #30057)

    [ SPEC / BUG FIXES ]
    - support redirects in mod_perl2, or fall back to using env variable
      for up to 5 redirects, when getting the query string (RT #36312)
    - CGI::Cookie now correctly supports the -max-age argument, previously
      if this was passed the value of the -expires argument would be used
      meaning there was no way to supply *only* this argument (RT #50576)
    - make :all actually import all methods, except for :cgi-lib, and add
      :ssl to the :standard import (RT #70337)

    [ DOCUMENTATION ]
    - clarify documentation regarding query_string method (RT #48370)
    - links fixed in some perldoc (Thanks to Michiel Beijen)

    [ TESTING ]
    - add t/changes.t for testing this Changes file
    - test case for RT #31107 confirming multipart parsing is to spec
    - improve t/rt-52469.t by adding a timeout check

4.02 2014-06-09

    [ NEW FEATURES ]
    - CGI::Carp learns noTimestamp / $CGI::Carp::NO_TIMESTAMP to prevent
      timestamp in messages (RT #82364, EDAVIS@cpan.org)
    - multipart_init and multipart_start learn -charset option (RT #22737)

    [ SPEC / BUG FIXES ]
    - Support multiple cookies when passing an ARRAY ref with -set-cookie
      (RT #15065, JWILLIAMS@cpan.org)

    [ DOCUMENTATION ]
    - Made licencing information consistent and remove duplicate comments
      about licence details, corrected location to report bugs (RT #38285)

(wiz)

Updated www/p5-Template-Toolkit to 2.26

(wiz)

Update to 2.26:

#-----------------------------------------------------------------------
# Version 2.26 - 17th September 2014
#------------------------------------------------------------------------

* Andy Wardley added outline directives. See Template::Manual::Syntax and
  Template::Manual::Config for details of the OUTLINE_TAG option and new
  'outline' TAG_STYLE.  See t/outline.t for examples.

* Andy Wardley improved the handling of keywords when the ANYCASE option
  is in use.  See t/anycase.t for examples.

* Chromatic fixed UTF-8 encoding in URLs in URL plugin.
  https://github.com/abw/Template2/pull/31

* Brian Fraser added support for platforms without LC_ALL/setlocale.
  https://github.com/abw/Template2/pull/34

* Amiri Barksdale fixed RT46691 to plug filter memory leaks
  https://github.com/abw/Template2/pull/36

* John Lightsey fixed RT59208 to improve SET
  https://github.com/abw/Template2/pull/38
  UPDATE: This had to be reverted as it cause a subtle breakage elsewhere

(wiz)

Updated devel/p5-Path-Class to 0.35

(wiz)

Update to 0.35:

0.35  Sun Sep 14 21:29:07 CDT 2014

- fce4b8e - Fixed a t/03-filesystem.t test error on Windows <Andrew Gregory>

(wiz)

Updated devel/p5-Exporter-Tiny to 0.040

(wiz)

Update to 0.040:

0.040 2014-09-17

[ Packaging ]
- Repackage as a stable release.

0.039_01 2014-07-20

[ Documentation ]
- Document warning and error messages produced by Exporter::Tiny.

[ Other ]
- Exporter::Tiny would previously cause B.pm to be loaded into memory any
  time it exported anything. It no longer does.
- No longer die when redefining locally defined subs.
- Warn when redefining any subs.

(wiz)

Updated devel/p5-Variable-Magic to 0.54

(wiz)

Update to 0.54:

0.54    2014-09-22 17:30 UTC
        + Add : The new constant VMG_COMPAT_CODE_COPY_CLONE evaluates to true
                if your perl calls 'copy' magic when a magical code prototype
                is cloned, which is currently the case for perl 5.17.0 and
                above.
        + Fix : [RT #90205] : copy magic on subs puts raw CV in $_[3]
                $_[3] will now contain a reference to the cloned code when
                'copy' magic is called for a coderef.
                Thanks Lukas Mai for reporting.
        + Fix : t/35-stash.t has been taught about perl 5.21.4.
        + Fix : Tests using run_perl() in t/17-ctl.t will no longer fail on
                Android.

(wiz)

Updated time/p5-DateTime-TimeZone to 1.75

(wiz)

Update to 1.75:

1.75    2014-09-30

- This release is based on version 2014h of the Olson database. This release
  only includes historical changes.

(wiz)

Updated devel/p5-Package-Stash to 0.37

(wiz)

Update to 0.37. Fix dependency.

0.37  2014-09-21
      - fix spurious warning in taint mode (Niko Tyni, #12)

(wiz)

Updated textproc/p5-YAML to 1.12

(wiz)

Update to 1.12:

1.12 Mon Sep 22 08:24:43 PDT 2014
- Fix https://rt.cpan.org/Ticket/Display.html?id=97870

(wiz)

Updated archivers/p5-Archive-Tar to 2.02

(wiz)

Update to 2.02:

2.02  14/09/2014
- Tests should now be able to be run in parallel

(wiz)

Updated devel/p5-IO-Compress to 2.066

(wiz)

Update to 2.066:

  2.066 21 Sept 2014

      * IO::Uncompress::Gzip
        Documentation of ExtraFlags stated the XFL values for BEST_COMPRESSION
        and BEST_SPEED use the values 2 & 4 respectively. They should
        be 4 & 2. Code for setting XFL was correct.

      * RT #95494: IO::Uncompress::Gunzip: Can no longer gunzip to in-memory
        file handle

(wiz)

Updated mail/p5-MIME-Types to 2.09

(wiz)

Update to 2.09:

version 2.09: Sun Sep 14 18:12:08 CEST 2014

Improvements:
- documentation use consistent parameter casing
- do not use '//' to support 5.8.*
  rt.cpan.org#98664 [Paul]

(wiz)

Updated devel/p5-Compress-Raw-Bzip2 to 2.066

(wiz)

Update to 2.066:
-------

  2.066 21 Sept 2014

      * No Changes

(wiz)

Updated devel/p5-Compress-Raw-Zlib to 2.066

(wiz)

Update to 2.066:

  2.066 21 Sept 2014

      * Another COW violation
        [#98069]

      * misleading nesting/indentation (found by Coverity)
        [#95405]

(wiz)

Updated misc/p5-Business-ISBN to 2.09

(wiz)

Update to 2.09:

2.09 - Sat Sep 20 18:15:37 2014
* Break circular dependency on Test::ISBN

2.08 - Fri Sep 19 12:53:21 2014
* Add increment and decrement to create new article numbers
(Markus Spann)

(wiz)

Updated misc/p5-Business-ISBN-Data to 20140910.002

(wiz)

Update to 20140910.002. Set LICENSE.

20140910.02 - Mon Sep 22 14:31:21 2014
* Look in the current directory for RangeMessage.xml if it's
not in other locations. This can help with various Perl app
packagers. (But, also try ISBN_RANGE_MESSAGE env var)

20140910.001 - Thu Sep 18 06:46:01 2014
* Update to the latest data (2014-09-10)

(wiz)

Updated net/p5-Net-DNS to 0.80

(wiz)

Update to 0.80:

**** 0.80 Sep 22, 2014

Removal of Win32::IPHelper support with cygwin

Resolvers on Cygwin can get their DNS configuration from the
registry directly via the /proc filesystem.  Getting rid of
the other method reduces dependencies and makes installations
less error prone.

Rework rt.cpan.org #96119

"Too late to run INIT block" warning for require Net::DNS

(wiz)

Updated databases/p5-DBD-postgresql to 3.4.2

(wiz)

Update to 3.4.2:

Version 3.4.2  Released September 25, 2015 (git commit 61440e1f4ccb6c293c5838676da1942e0df67271)

  - Fix bug where single-quoted type arguments to the table_info()
    method were causing a SQL error.
    [Greg Sabino Mullane] (CPAN bug #99144)

(wiz)

Updated finance/gnucash to 2.6.3nb10

(wiz)

According to upstream, guile-www is not used. Remove dependency
and patch.
See https://bugzilla.gnome.org/show_bug.cgi?id=726164

Bump PKGREVISION.

(wiz)

Mention meld-devel removal.

(wiz)

Remove meld-devel, meld is now newer again.

(wiz)

Updated devel/meld to 3.12.0

(wiz)