Updated graphics/MesaLib to 11.0.3

(tnn)

Changes 2015.10.18:
* get format_id from video file ext
* check for the offline error page
* treat the offline error as an expected ExtractorError
* Look for sm4:video:embed
* Add _extract_url
* Use _extract_url for mtvservices

(adam)

Update to mesa-11.0.3 (2015-10-10)

"In the current release we have a bunch of EGL patches, mangledGL build fixes
and a healthy amount of driver bugfixes - radeonsi, nouveau, i915 and i965.
Last but not least, the KDE/Weston regression introduced with 11.0.2 has also
been resolved."

(tnn)

exit with an error if the JVM crashes and chdir to /tmp before running it,
so we don't get a stale java.core in pkgdb.

(tnn)

OpenBSD and derivatives define MACHINE_ARCH in sys/param.h to "amd64" which
overrides our attempt to set it to "x86_64" and ensure consistency across
platforms.  Work around this by setting it using PKGSRC_MACHINE_ARCH.

Confirmed to fix the issue and not break a variety of other platforms by
Sevan, and approved under duress by Joerg.

(jperkin)

Updated x11/xinput to 1.6.2

(wiz)

Update to 1.6.2:

A couple of minor patches accumulated over the years. One crasher fix, one
exit status correction, two fixes for the test-xi2 printf. And some build
system cleanups.

Gaetan Nadon (1):
      config: replace deprecated use of AC_OUTPUT with AC_CONFIG_FILES

Peter Hutterer (6):
      Drop AM_MAINTAINER_MODE
      Fix crash for enable/disable without a device argument
      Fix valuator printf output in test-xi2
      property: return EXIT_FAILURE if at least one device fails (#90226)
      Print touch event flags in test-xi2
      xinput 1.6.2

(wiz)

Pass -DENABLE_NLS on SunOS to ensure consistency of including libintl.h
as it is pulled in implicitly by locale.h when __EXTENSIONS__ is used.

Fixed build on SmartOS.

(jperkin)

Bump revision for the libxl fix.

(joerg)

Don't depend on the compiler to optimize ffsll away, explicitly use the
builtin. Fix obvious logic error in libxl. Disable the use of global
register variables for clang. Lesser evil, HVM users should use newer
xentools or GCC.

(joerg)

Give _num.c the same optimiser threatment as _io.c for clang.

(joerg)

Don't define nullptr when using a C++11 compiler in general or clang
specifically.

(joerg)

Don't encode the github tag into the distfile name if we are packaging
a github release and the release version happens to look like a tag.
(a date like 20150714 could be misinterpreted as a git "short SHA")

(tnn)

Updated libbson and mongo-c-driver

(fhajny)

Update databases/mongo-c-driver to 1.2.0.

mongo-c-driver 1.2.0
- Much improved reporting of network errors, unavailable servers, and
  authentication failure
- Destroying an exhaust cursor must close its socket
- Various bugs in server reconnection logic
- mongoc_collection_aggregate returned invalid cursor after failure
- Wrong error message after failed network write on Sparc
- Missing JSON test files in release tarball
- Enable runtime asserts in release build.
- mongoc_client_kill_cursor is now deprecated and will be removed
  in version 2.0.

mongo-c-driver 1.1.11
- Undetected network errors when sending messages to the server
- Off-by-one error in mongoc_gridfs_file_seek with mode SEEK_END
- Memory leak parsing a URI that contains an invalid option

(fhajny)

Update devel/libbson 1.1.10 to 1.2.0.

libbson 1.2.0
- Add bson_mem_restore_vtable(), the inverse of bson_mem_set_vtable().
- Enable runtime asserts in release build.
- Fixed compiler warnings and build failures on various platforms.
- Improvements to the formatting and contents of the documentation.

libbson 1.1.11
- Document bson streaming reads with an example, bson-streaming-reader.c.
- Document callback function types bson_reader_destroy_func_t and
  bson_reader_read_func_t.

(fhajny)

+ harfbuzz-1.0.6, libebml-1.3.2, libmatroska-1.4.3, p5-CGI-4.22,
  py-anki2-2.0.33, talloc-2.1.4, wireshark-1.12.8, x264-devel-20151017.

(wiz)

Add textproc/py-xlsxwriter

(fhajny)

Added textproc/py-xlsxwriter version 0.7.7

(fhajny)

Import XlsxWriter as textproc/py-xlsxwriter.

XlsxWriter is a Python module for writing files in the Excel
2007+ XLSX file format.

(fhajny)

Updated devel/py-test to 2.8.2

(wiz)

Update py-test to 2.8.2:

2.8.2
-----

- fix #1085: proper handling of encoding errors when passing encoded byte
  strings to pytest.parametrize in Python 2.
  Thanks Themanwithoutaplan for the report and Bruno Oliveira for the PR.

- fix #1087: handling SystemError when passing empty byte strings to
  pytest.parametrize in Python 3.
  Thanks Paul Kehrer for the report and Bruno Oliveira for the PR.

- fix #995: fixed internal error when filtering tracebacks where one entry
  was generated by an exec() statement.
  Thanks Daniel Hahler, Ashley C Straw, Philippe Gauthier and Pavel Savchenko
  for contributing and Bruno Oliveira for the PR.

2.8.1
-----

- fix #1034: Add missing nodeid on pytest_logwarning call in
  addhook.  Thanks Simon Gomizelj for the PR.

- 'deprecated_call' is now only satisfied with a DeprecationWarning or
  PendingDeprecationWarning. Before 2.8.0, it accepted any warning, and 2.8.0
  made it accept only DeprecationWarning (but not PendingDeprecationWarning).
  Thanks Alex Gaynor for the issue and Eric Hunsberger for the PR.

- fix issue #1073: avoid calling __getattr__ on potential plugin objects.
  This fixes an incompatibility with pytest-django.  Thanks Andreas Pelme,
  Bruno Oliveira and Ronny Pfannschmidt for contributing and Holger Krekel
  for the fix.

- Fix issue #704: handle versionconflict during plugin loading more
  gracefully.  Thanks Bruno Oliveira for the PR.

- Fix issue #1064: ""--junitxml" regression when used with the
  "pytest-xdist" plugin, with test reports being assigned to the wrong tests.
  Thanks Daniel Grunwald for the report and Bruno Oliveira for the PR.

- (experimental) adapt more SEMVER style versioning and change meaning of
  master branch in git repo: "master" branch now keeps the bugfixes, changes
  aimed for micro releases.  "features" branch will only be be released
  with minor or major pytest releases.

- Fix issue #766 by removing documentation references to distutils.
  Thanks Russel Winder.

- Fix issue #1030: now byte-strings are escaped to produce item node ids
  to make them always serializable.
  Thanks Andy Freeland for the report and Bruno Oliveira for the PR.

- Python 2: if unicode parametrized values are convertible to ascii, their
  ascii representation is used for the node id.

- Fix issue #411: Add __eq__ method to assertion comparison example.
  Thanks Ben Webb.

- fix issue 877: properly handle assertion explanations with non-ascii repr
  Thanks Mathieu Agopian for the report and Ronny Pfannschmidt for the PR.

- fix issue 1029: transform errors when writing cache values into pytest-warnings

2.8.0
-----------------------------

- new ``--lf`` and ``-ff`` options to run only the last failing tests or
  "failing tests first" from the last run.  This functionality is provided
  through porting the formerly external pytest-cache plugin into pytest core.
  BACKWARD INCOMPAT: if you used pytest-cache's functionality to persist
  data between test runs be aware that we don't serialize sets anymore.
  Thanks Ronny Pfannschmidt for most of the merging work.

- "-r" option now accepts "a" to include all possible reports, similar
  to passing "fEsxXw" explicitly (isse960).
  Thanks Abhijeet Kasurde for the PR.

- avoid python3.5 deprecation warnings by introducing version
  specific inspection helpers, thanks Michael Droettboom.

- fix issue562: @nose.tools.istest now fully respected.

- fix issue934: when string comparison fails and a diff is too large to display
  without passing -vv, still show a few lines of the diff.
  Thanks Florian Bruhin for the report and Bruno Oliveira for the PR.

- fix issue736: Fix a bug where fixture params would be discarded when combined
  with parametrization markers.
  Thanks to Markus Unterwaditzer for the PR.

- fix issue710: introduce ALLOW_UNICODE doctest option: when enabled, the
  ``u`` prefix is stripped from unicode strings in expected doctest output. This
  allows doctests which use unicode to run in Python 2 and 3 unchanged.
  Thanks Jason R. Coombs for the report and Bruno Oliveira for the PR.

- parametrize now also generates meaningful test IDs for enum, regex and class
  objects (as opposed to class instances).
  Thanks to Florian Bruhin for the PR.

- Add 'warns' to assert that warnings are thrown (like 'raises').
  Thanks to Eric Hunsberger for the PR.

- Fix issue683: Do not apply an already applied mark.  Thanks ojake for the PR.

- Deal with capturing failures better so fewer exceptions get lost to
  /dev/null.  Thanks David Szotten for the PR.

- fix issue730: deprecate and warn about the --genscript option.
  Thanks Ronny Pfannschmidt for the report and Christian Pommranz for the PR.

- fix issue751: multiple parametrize with ids bug if it parametrizes class with
  two or more test methods. Thanks Sergey Chipiga for reporting and Jan
  Bednarik for PR.

- fix issue82: avoid loading conftest files from setup.cfg/pytest.ini/tox.ini
  files and upwards by default (--confcutdir can still be set to override this).
  Thanks Bruno Oliveira for the PR.

- fix issue768: docstrings found in python modules were not setting up session
  fixtures. Thanks Jason R. Coombs for reporting and Bruno Oliveira for the PR.

- added `tmpdir_factory`, a session-scoped fixture that can be used to create
  directories under the base temporary directory. Previously this object was
  installed as a `_tmpdirhandler` attribute of the `config` object, but now it
  is part of the official API and using `config._tmpdirhandler` is
  deprecated.
  Thanks Bruno Oliveira for the PR.

- fix issue808: pytest's internal assertion rewrite hook now implements the
  optional PEP302 get_data API so tests can access data files next to them.
  Thanks xmo-odoo for request and example and Bruno Oliveira for
  the PR.

- rootdir and inifile are now displayed during usage errors to help
  users diagnose problems such as unexpected ini files which add
  unknown options being picked up by pytest. Thanks to Pavel Savchenko for
  bringing the problem to attention in #821 and Bruno Oliveira for the PR.

- Summary bar now is colored yellow for warning
  situations such as: all tests either were skipped or xpass/xfailed,
  or no tests were run at all (this is a partial fix for issue500).

- fix issue812: pytest now exits with status code 5 in situations where no
  tests were run at all, such as the directory given in the command line does
  not contain any tests or as result of a command line option filters
  all out all tests (-k for example).
  Thanks Eric Siegerman (issue812) and Bruno Oliveira for the PR.

- Summary bar now is colored yellow for warning
  situations such as: all tests either were skipped or xpass/xfailed,
  or no tests were run at all (related to issue500).
  Thanks Eric Siegerman.

- New `testpaths` ini option: list of directories to search for tests
  when executing pytest from the root directory. This can be used
  to speed up test collection when a project has well specified directories
  for tests, being usually more practical than configuring norecursedirs for
  all directories that do not contain tests.
  Thanks to Adrian for idea (#694) and Bruno Oliveira for the PR.

- fix issue713: JUnit XML reports for doctest failures.
  Thanks Punyashloka Biswal.

- fix issue970: internal pytest warnings now appear as "pytest-warnings" in
  the terminal instead of "warnings", so it is clear for users that those
  warnings are from pytest and not from the builtin "warnings" module.
  Thanks Bruno Oliveira.

- Include setup and teardown in junitxml test durations.
  Thanks Janne Vanhala.

- fix issue735: assertion failures on debug versions of Python 3.4+

- new option ``--import-mode`` to allow to change test module importing
  behaviour to append to sys.path instead of prepending.  This better allows
  to run test modules against installated versions of a package even if the
  package under test has the same import root.  In this example::

        testing/__init__.py
        testing/test_pkg_under_test.py
        pkg_under_test/

  the tests will run against the installed version
  of pkg_under_test when ``--import-mode=append`` is used whereas
  by default they would always pick up the local version.  Thanks Holger Krekel.

- pytester: add method ``TmpTestdir.delete_loaded_modules()``, and call it
  from ``inline_run()`` to allow temporary modules to be reloaded.
  Thanks Eduardo Schettino.

- internally refactor pluginmanager API and code so that there
  is a clear distinction between a pytest-agnostic rather simple
  pluginmanager and the PytestPluginManager which adds a lot of
  behaviour, among it handling of the local conftest files.
  In terms of documented methods this is a backward compatible
  change but it might still break 3rd party plugins which relied on
  details like especially the pluginmanager.add_shutdown() API.
  Thanks Holger Krekel.

- pluginmanagement: introduce ``pytest.hookimpl`` and
  ``pytest.hookspec`` decorators for setting impl/spec
  specific parameters.  This substitutes the previous
  now deprecated use of ``pytest.mark`` which is meant to
  contain markers for test functions only.

- write/refine docs for "writing plugins" which now have their
  own page and are separate from the "using/installing plugins`` page.

- fix issue732: properly unregister plugins from any hook calling
  sites allowing to have temporary plugins during test execution.

- deprecate and warn about ``__multicall__`` argument in hook
  implementations.  Use the ``hookwrapper`` mechanism instead already
  introduced with pytest-2.7.

- speed up pytest's own test suite considerably by using inprocess
  tests by default (testrun can be modified with --runpytest=subprocess
  to create subprocesses in many places instead).  The main
  APIs to run pytest in a test is "runpytest()" or "runpytest_subprocess"
  and "runpytest_inprocess" if you need a particular way of running
  the test.  In all cases you get back a RunResult but the inprocess
  one will also have a "reprec" attribute with the recorded events/reports.

- fix monkeypatch.setattr("x.y", raising=False) to actually not raise
  if "y" is not a pre-existing attribute. Thanks Florian Bruhin.

- fix issue741: make running output from testdir.run copy/pasteable
  Thanks Bruno Oliveira.

- add a new ``--noconftest`` argument which ignores all ``conftest.py`` files.

- add ``file`` and ``line`` attributes to JUnit-XML output.

- fix issue890: changed extension of all documentation files from ``txt`` to
  ``rst``. Thanks to Abhijeet for the PR.

- fix issue714: add ability to apply indirect=True parameter on particular argnames.
  Thanks Elizaveta239.

- fix issue890: changed extension of all documentation files from ``txt`` to
  ``rst``. Thanks to Abhijeet for the PR.

- fix issue957: "# doctest: SKIP" option will now register doctests as SKIPPED
  rather than PASSED.
  Thanks Thomas Grainger for the report and Bruno Oliveira for the PR.

- issue951: add new record_xml_property fixture, that supports logging
  additional information on xml output. Thanks David Diaz for the PR.

- issue949: paths after normal options (for example `-s`, `-v`, etc) are now
  properly used to discover `rootdir` and `ini` files.
  Thanks Peter Lauri for the report and Bruno Oliveira for the PR.

2.7.3 (compared to 2.7.2)
-----------------------------

- Allow 'dev', 'rc', or other non-integer version strings in `importorskip`.
  Thanks to Eric Hunsberger for the PR.

- fix issue856: consider --color parameter in all outputs (for example
  --fixtures). Thanks Barney Gale for the report and Bruno Oliveira for the PR.

- fix issue855: passing str objects as `plugins` argument to pytest.main
  is now interpreted as a module name to be imported and registered as a
  plugin, instead of silently having no effect.
  Thanks xmo-odoo for the report and Bruno Oliveira for the PR.

- fix issue744: fix for ast.Call changes in Python 3.5+.  Thanks
  Guido van Rossum, Matthias Bussonnier, Stefan Zimmermann and
  Thomas Kluyver.

- fix issue842: applying markers in classes no longer propagate this markers
  to superclasses which also have markers.
  Thanks xmo-odoo for the report and Bruno Oliveira for the PR.

- preserve warning functions after call to pytest.deprecated_call. Thanks
  Pieter Mulder for PR.

- fix issue854: autouse yield_fixtures defined as class members of
  unittest.TestCase subclasses now work as expected.
  Thannks xmo-odoo for the report and Bruno Oliveira for the PR.

- fix issue833: --fixtures now shows all fixtures of collected test files, instead of just the
  fixtures declared on the first one.
  Thanks Florian Bruhin for reporting and Bruno Oliveira for the PR.

- fix issue863: skipped tests now report the correct reason when a skip/xfail
  condition is met when using multiple markers.
  Thanks Raphael Pierzina for reporting and Bruno Oliveira for the PR.

- optimized tmpdir fixture initialization, which should make test sessions
  faster (specially when using pytest-xdist). The only visible effect
  is that now pytest uses a subdirectory in the $TEMP directory for all
  directories created by this fixture (defaults to $TEMP/pytest-$USER).
  Thanks Bruno Oliveira for the PR.

(wiz)

Updated pkgtools/pkglint to 4.157

(wiz)

Bump version for recent fixes.

(wiz)

Updated devel/py-cython to 0.23.4

(wiz)

Update to 0.23.4:

0.23.4 (2015-10-10)
===================

Bugs fixed
----------

* Memory leak when calling Python functions in PyPy.

* Compilation problem with MSVC in C99-ish mode.

* Warning about unused values in a helper macro.

0.23.3 (2015-09-29)
===================

Bugs fixed
----------

* Invalid C code for some builtin methods.  This fixes ticket 856 again.

* Incorrect C code in helper functions for PyLong conversion and string
  decoding.  This fixes ticket 863, ticket 864 and ticket 865.
  Original patch by Nikolaus Rath.

* Large folded or inserted integer constants could use too small C
  integer types and thus trigger a value wrap-around.

Other changes
-------------

* The coroutine and generator types of Cython now also register directly
  with the ``Coroutine`` and ``Generator`` ABCs in the ``backports_abc``
  module if it can be imported.  This fixes ticket 870.

(wiz)

Updated security/py-cryptography to 1.0.2

(wiz)

Update py-cryptography to 1.0.2:

1.0.2 - 2015-09-27
~~~~~~~~~~~~~~~~~~
* **SECURITY ISSUE**: The OpenSSL backend prior to 1.0.2 made extensive use
  of assertions to check response codes where our tests could not trigger a
  failure.  However, when Python is run with ``-O`` these asserts are optimized
  away.  If a user ran Python with this flag and got an invalid response code
  this could result in undefined behavior or worse. Accordingly, all response
  checks from the OpenSSL backend have been converted from ``assert``
  to a true function call. Credit **Emilia K辰sper (Google Security Team)**
  for the report.

(wiz)

Updated editors/joe to 4.1

(adam)

Changes 4.1:
* New or improved syntax files for the following languages:
        * Groovy, R, Clojure, Rust, Coffeescript, Java, Scala, Swift, D,
          AVR, Ruby, Perl
* New translations
        * Chinese (zh_TW)
* UTF-16 support
        * JOE can now edit UTF-16BE and UTF-16LE files.  It does this
          by converting them to UTF-8 during load and back to UTF-16 during
          save.
        * Within JOE, native byte order is called UTF-16 and reversed order
          is called UTF-16R.
        * If you change the encoding (with ^T E) between UTF-8, UTF-16 and UTF-16R,
          JOE will convert the file to the desired encoding on save.
* New regular expression engine
* Unicode improvements
* Code clean up
* Bugs fixed
* Minor enhancements
* jmacs fixes:
* ESC g (grep/find) and ESC c (compile) improvements

(adam)

Add optional (default-off) curses support to cunit.

>From Alexander Vasarab on tech-pkg.

(wiz)

+ hdf5-1.8.15patch1.

(wiz)

Updated games/lgogdownloader to 2.26

(wiz)

Update lgogdownloader to 2.26:

LGOGDownloader 2.26
- Check that stdin is a tty before reading (patch by: Stephen Kitt)
- Added version string and number to cached game details
- Fixed priority handling for DLCs
- Combined priority handling with --platform and --language options
* Use "+" as separator to combine values and "," to set priority
* Syntax for --platform and --language options was changed: see --help or man page for details
- Added support for new language: Slovak
- Use include/exclude options to choose what files to download
* --cover, --no-cover, --no-installers, --no-extras, --no-patches, --no-language-packs and --no-dlc options have been removed
* Default: --include "all" --exclude "covers"
* See --help for available options

(wiz)

Updated net/rabbitmq to 3.5.6

(fhajny)

Update net/rabbitmq to 3.5.6

- Erlang 18.1 compatibility.
- Prevent EACCESS errors on Windows when queue journal is cleared.
- When multiple authorization backends are used, user tags from all
  of them should be preserved.
- Force a (per-queue, not global) GC when a queue pages messages
  to disk.
- MQTT Plugin: Queues used by QoS 1 subscriptions are no longer
  deleted when the only subscriber disconnects.
- STOMP Plugin: Trailing new line character now can be optional.

(fhajny)

Updated textproc/sphinxsearch to 2.2.10

(fhajny)

Update textproc/sphinxsearch to 2.2.10.

New features and updates

- The list of SphinxQL reserved keywords is now much, much smaller than
  it used to be, and also consistent across different places in the code
  and in the docs.
- IN() function now accepts string attributes (#2233). Also, filtering
  string attributes by an empty value in WHERE clause now works.
- WHERE on float attributes got improved a bit: we added floatcol<>123.0
  syntax, and fixed an issue with using an integer reference value asi
  in floatcol=123 or floatcol!=123, see bug #2121.
- searchd --replay-flags switch (useful for binlog replay troubleshooting)
  now accepts a new option 'ignore-open-errors' that lets you simply
  ignore missing binlog files without deleting binlog meta info.
- Indexer now exits with code 2 in case of rotation failure.

Notable bug fixes
- unified min_prefix_len, min_infix_len behavior between RT and plain indexes
- expressions dependent on aggregation results (eg. as in SELECT MAX(id) m1,
  m1+10 m2) were not computed properly in RT indexes
- LENGTH() was 2x off for 64-bit MVA attributes
- OPTIMIZE could occasionally break big RT indexess
- multi-wordforms with clashing prefixes were processed in a wrong order
- disabled and later re-enabled indexes were not picked up again by searchd
  on SIGHUP

Full release notes:

  http://sphinxsearch.com/docs/latest/rel2210.html

(fhajny)

pullups 4839 4840 4841

(spz)

set HOMEPAGE to the hg revision log URL to make it easier to figure out if
the package needs an update.

(tnn)

Pullup ticket #4841 - requested by tnn
multimedia/adobe-flash-plugin11: security update

Revisions pulled up:
- multimedia/adobe-flash-plugin11/Makefile                      1.50,1.52
- multimedia/adobe-flash-plugin11/distinfo                      1.47,1.49

-------------------------------------------------------------------
  Module Name: pkgsrc
  Committed By: tsutsui
  Date: Sun Oct 18 14:15:23 UTC 2015

  Modified Files:
  pkgsrc/multimedia/adobe-flash-plugin11: Makefile distinfo

  Log Message:
  Update adobe-flash-plugin11 to 11.2.202.540.

  Upstream announcement:

    https://helpx.adobe.com/security/products/flash-player/apsb15-27.html

  Adobe Security Bulletin

  Security updates available for Adobe Flash Player

  Release date: October 16, 2015

  Vulnerability identifier: APSB15-27

  CVE number: CVE-2015-7645, CVE-2015-7647, CVE-2015-7648

  Platform: All Platforms

  To generate a diff of this commit:
  cvs rdiff -u -r1.51 -r1.52 pkgsrc/multimedia/adobe-flash-plugin11/Makefile
  cvs rdiff -u -r1.48 -r1.49 pkgsrc/multimedia/adobe-flash-plugin11/distinfo

(spz)

Updated databases/py-elasticsearch to 2.0.0

(fhajny)

Update databases/py-elasticsearch to 2.0.0.

2.0.0 (2015-10-14)
- Elasticsearch 2.0 compatibility release

1.8.0 (2015-10-14)
- removed thrift and memcached connections, if you wish to continue
  using those, extract the classes and use them separately.
- added a new, parallel version of the bulk helper using thread pools

(fhajny)

Pullup ticket #4840 - requested by mrg
lang/perl5: fix for sparc

Revisions pulled up:
- lang/perl5/hacks.mk                                          1.18

-------------------------------------------------------------------
  Module Name:    pkgsrc
  Committed By:  mrg
  Date:          Thu Oct 15 00:15:52 UTC 2015

  Modified Files:
          pkgsrc/lang/perl5: hacks.mk

  Log Message:
  enable the GCC 4.5 op.c hack for sparc.

  To generate a diff of this commit:
  cvs rdiff -u -r1.17 -r1.18 pkgsrc/lang/perl5/hacks.mk

(spz)

Pullup ticket #4839 - requested by gdt
sysutils/dbus: build fix

Revisions pulled up:
- sysutils/dbus/distinfo                                        1.67
- sysutils/dbus/patches/patch-tools_dbus-print-message.c        1.1

-------------------------------------------------------------------
  Module Name: pkgsrc
  Committed By: joerg
  Date: Thu Oct  8 14:32:09 UTC 2015

  Modified Files:
  pkgsrc/sysutils/dbus: distinfo
  Added Files:
  pkgsrc/sysutils/dbus/patches: patch-tools_dbus-print-message.c

  Log Message:
  Fix build on NetBSD 5.x and 6.x.

  To generate a diff of this commit:
  cvs rdiff -u -r1.66 -r1.67 pkgsrc/sysutils/dbus/distinfo
  cvs rdiff -u -r0 -r1.1 \
      pkgsrc/sysutils/dbus/patches/patch-tools_dbus-print-message.c

(spz)

Updated databases/elasticsearch to 1.7.3

(fhajny)

Update databases/elasticsearch to 1.7.3.
Fix the plugin script.

elasticsearch 1.7.3
===================

Enhancements
  Tribe Node:
    TransportNodesAction shouldn't hold on to cluster state

Bug fixes
  Internal:
    - An inactive shard is activated by triggered synced flush
    - Pending operations in the translog prevent shard from being
      marked as inactive
  Mapping:
    - Ensure more specific analyzer is used independent of the mapping order
  REST:
    - Expose nodes operation timeout in REST API
  Snapshot/Restore:
    - Snapshot restore operations throttle more than specified
  Tribe Node:
    - Increment tribe node version on updates

Upgrades
  Network:
    - Upgrade Netty to 3.10.5.final

elasticsearch 1.7.2
===================

Enhancements
  Allocation:
    - Avoid extra reroutes of delayed shards in RoutingService
  Packaging:
    - Shade joda-convert
  Query DSL:
    - Add a check if FieldMappers exists

Bug fixes
  Allocation:
    - Take relocating shard into consideration during awareness allocation
    - Take initializing shards into consideration during awareness allocation
    - Fix messaging about delayed allocation
  Internal:
    - Fix concurrency issue in PrioritizedEsThreadPoolExecutor.
  Nested Docs:
    - Nested query should only use bitset cache for parent filter
  Packaging:
    - Fix upgrade RPM script
  Plugins:
    - Plugin script: Fix ES_HOME with spaces
  Query DSL:
    - Do not track named queries that are null
    - multi_match query applies boosts too many times.
  Recovery:
    - Rethrow exception during recovery finalization even if source is
      not broken
  Snapshot/Restore:
    - Snapshot restore request should accept indices options
  Suggesters:
    - Prevent DirectCandidateGenerator to reuse an unclosed analyzer

(fhajny)

Updated audio/mpg123 to 1.22.4; audio/wavpack to 4.75.2; audio/py-audiotools to 3.1; graphics/py-Pillow to 3.0.0; graphics/hugin to 2015.0.0

(adam)

Changes 2015.0.0:
Hugin 2015.0 has a number of new features. Most of the changes are under the hood to improve stability, allow easier maintenance and easier addition of new features.
The makefile based stitching engine has been replaced with direct calls to the underlying programs. This should reduce the overhead of calling the same program too often and so speed up the whole process. Hopefully this also allows the usage of more unusual characters in the filename. Instead of pto2mk and make we now have hugin_executor which can also be used for running the assistant from the command line.
Hugin now has it's own blender, verdandi, based on a watershed algorithm, which has been included in 'nona'. verdandi can be chosen as the blender by choosing "builtin" in the stitcher tab, and can also be set in the Preferences. verdandi can also be called as a command line tool.
The lensfun library has been removed as it did not fulfill our expectations. It has been replaced with our own camera and lens database which uses a data mining approach and operates automatically without user intervention. Geometric distortion and vignetting data have to be loaded manually.
The fast preview window has a new tool to add or remove control points to selected areas in the output projection.
Automatic exposure stack detection is now applied when loading images, and an option has been added to unlink image position when adding stacks.
The fine-tune and auto-estimate functions in the control point tab have been made projection aware. It should now work also with images with different fov or different projections. The fine-tune feature has had a significant speed up when Hugin is compiled with the libfftw3 library (optional).
PTBatcherGUI has more choice for the end of the process: depending on the operating system the PTBatcherGUI can be closed, the computer can be shut down or send to the hibernate mode.
PTBatcherGUI now shows the thumbnails when searching for images in directories.
Many of the underlying tools in hugin are now able to use available cpu cores.

(adam)

Changes 3.0.0:
* Saving Multipage Images
* Tiff ImageFileDirectory Rewrite
* Deprecated Methods
* LibJpeg and Zlib are Required by Default

(adam)

Changes 3.1:
Bug fixes and improvements.

(adam)

Changes 4.75.2:
* fixed: corrupt mono or multichannel files created with assembly code (rare)
* fixed: building on Clang systems like Darwin and FreeBSD (req. Clang 3.5+)
* fixed: explicitly sign-extend audio data (< 4-byte) to avoid corrupt files
* fixed: rare decoding errors due to integer overflow (ARM assembly code)
* added: assembly optimizations for "extra" mode on mono or multichannel

(adam)

Changes 1.22.4:
- Fix stupid regression introduced with a bugfix in 1.14.1 to make
  free format streams work again. It took 3 years for someone to notice.
  I added a regression test for that now!

(adam)

Updated multimedia/adobe-flash-plugin11 to 11.2.202.540

(tsutsui)

Update adobe-flash-plugin11 to 11.2.202.540.

Upstream announcement:

https://helpx.adobe.com/security/products/flash-player/apsb15-27.html

Adobe Security Bulletin

Security updates available for Adobe Flash Player

Release date: October 16, 2015

Vulnerability identifier: APSB15-27

CVE number: CVE-2015-7645, CVE-2015-7647, CVE-2015-7648

Platform: All Platforms

(tsutsui)

pbulk-0.59: Belatedly bump PBULK_CONFIG_VERSION to ensure that
sync_packages etc are present.

(joerg)

add NVIDIA firmware license

(jmcneill)

Added sysutils/tegra-firmware version 20151018

(jmcneill)

+ tegra-firmware

(jmcneill)

API bump for devel/deforaos-libsystem

(khorben)

Added textproc/multimarkdown version 4.7.1

(ryoon)

Add multimarkdown

(ryoon)

Import multimarkdown-4.7.1 as textproc/multimarkdown.

MultiMarkdown, or MMD, is a tool to help turn minimally marked-up
plain text into well formatted documents, including HTML, PDF (by
way of LaTeX), OPML, or OpenDocument (specifically, Flat OpenDocument
or '.fodt', which can in turn be converted into RTF, Microsoft
Word, or virtually any other word-processing format).

MMD is a superset of the Markdown syntax, originally created by
John Gruber. It adds multiple syntax features (tables, footnotes,
and citations, to name a few), in addition to the various output
formats listed above (Markdown only creates HTML). Additionally,
it builds in "smart" typography for various languages (proper left-
and right-sided quotes, for example).

(ryoon)

Updated net/rabbitmq-c to 0.7.1

(fhajny)

Update net/rabbitmq-c to 0.7.1.

- Win32: Use WSAEWOULDBLOCK instead of EWOULDBLOCK on Win32 (win32
  clients would fail to connect)
- Lib: if channel_max is 0 use server's channel_max
- Lib: fix build on OpenBSD

(fhajny)

Updated databases/redis to 3.0.5

(fhajny)

Update databases/redis to 3.0.5.

--[ Redis 3.0.5 ] Release date: 15 Oct 2015

Upgrade urgency: MODERATE, the most important thing is a fix in the replication
                code that may make the slave hanging forever if the master
                remains with an open socket even if it is no longer able to
                reply.

* [FIX] MOVE now moves the TTL as well. A bug lasting forever... finally
        fixed thanks to Andy Grunwald that reported it.
        (reported by Andy Grunwald, fixed by Salvatore Sanfilippo)
* [FIX] Fix a false positive in HSTRLEN test.
* [FIX] Fix a bug in redis-cli --pipe mode that was not able to read back
        replies from the server incrementally. Now a mass import will use
        a lot less memory, and you can use --pipe to do incremental streaming.
        (reported by Twitter user @fsaintjacques, fixed by Salvatore
        Sanfilippo)
* [FIX] Slave detection of master timeout. (fixed by Kevin McGehee, refactoring
        and regression test by Salvatore Sanfilippo)

* [NEW] Cluster: redis-trib fix can fix an additional case for opens lots.
        (Salvatore Sanfilippo)
* [NEW] Cluster: redis-trib import support for --copy and --replace options
        (David Thomson)

--[ Redis 3.0.4 ] Release date: 8 Sep 2015

Upgrade urgency: HIGH for Redis and Sentinel. However note that in order to
                fix certain replication bugs, the replication internals were
                modified in a very heavy way. So while this release is
                conceptually saner, it may contain regressions. For this
                reason, before the release, QA activities were performed by
                me (antirez) and Redis Labs and no evident bug was found.

* [FIX] A number of bugs related to replication PSYNC and the (yet experimental)
        diskless replication feature were fixed. The bugs could lead to
        inconsistency between masters and slaves. (Salvatore Sanfilippo, Oran
        Agra fixed the issue found by Yuval Inbar)
* [FIX] A replication bug in the context of PSYNC partial resynchonization was
        found and fixed. This bug happens even when diskless replication is off
        in the case different slaves connect at different times while the master
        is creating an RDB file, and later a partial resynchronization is
        attempted by a slave that connected not as the first one. (Salvatore
        Sanfilippo, Oran Agra)
* [FIX] Chained replication and PSYNC interactions leading to potential stale
        chained slaves data set, see issue #2694. (Salvatore Sanfilippo fixed
        an issue reported by "GeorgeBJ" user at Github)
* [FIX] redis-cli --scan iteration fixed when returned cursor overflows
        32 bit signed integer. (Ofir Luzon, Yuval Inbar)
* [FIX] Sentinel: fixed a bug during the master switch process, where for a
        failed conditional check, the new configuration is rewritten, during
        a small window of time, in a corrupted way where the master is
        also reported to be one of the slaves. This bug is rare to trigger
        but apparently it happens in the wild, and the effect is to see
        a replication loop where the master will try to replicate with itself.
        The bug was found by Jan-Erik Rediger using a static analyzer and
        fixed by Salvatore Sanfilippo.
* [FIX] Sentinel lack of arity checks for certain commands.
        (Rogerio Goncalves, Salvatore Sanfilippo)

* [NEW] Replication internals rewritten in order to be more resistant to bugs.
        The replication handshake in the slave side was rewritten as a non
        blocking state machine. (Salvatore Sanfilippo, Oran Agra)
* [NEW] New "replication capabilities" feature introduced in order to signal
        from the master to the slave what are the features supported, so that
        the master can choose the kind of replication to start (diskless or
        not) when master and slave are of different versions. (Oran Agra,
        Salvatore Sanfilippo)
* [NEW] Log clients details when SLAVEOF command is received. (Salvatore
        Sanfilippo with inputs from Nick Craver and Marc Gravell).

(fhajny)

Updated inputmethod/kinput2 to 3.1nb4

(tsutsui)

Apply kinput2-v3.1-canna37.diff, which enables Canna 3.7 new APIs.

Bump PKGREVISION.

Tested on NetBSD/i386 7.0 and NetBSD/luna68k 7.0.

(tsutsui)

Updated databases/mongodb to 3.0.7

(ryoon)

Update to 3.0.7

Changelog:
3.0.7 – Oct 13, 2015

    Improvements to WiredTiger memory handling and performance: SERVER-20159, SERVER-20204, SERVER-20091, and SERVER-20176.
    Fixed issue where replica set nodes will not step down: SERVER-19464.
    Additional privileges for built-in roles: SERVER-19131, SERVER-15893, and SERVER-13647.
    All issues closed in 3.0.7

(ryoon)

Updated inputmethod/canna to 3.7pl3
Updated inputmethod/canna-canuum to 3.7pl3
Updated inputmethod/canna-dict to 3.7pl3
Updated inputmethod/canna-lib to 3.7pl3
Updated inputmethod/canna-server to 3.7pl3

(tsutsui)

Update canna packages to 3.7pl3.

pkgsrc changes:
- reorganize patch filenames for maintainability
  - patch-ac is also merged into existing patch-dic_phono_Imakefile
- fix a missed 64 bit time_t issue in libRK
- add a hook in cmd.tmpl to disable installation of cmd/forcpp and
  cmd/kpdic in canna-lib
- tweak SUBDIRS patches in Makefiles to avoid unnecessary builds
  for each packages
- pull various updates to canuum from the original FreeWnn's uum,
  including posix_openpt(3) support, and disable the default but
  obsolete libspt support
- reorganize curses(3) ops in canuum per curses(3) documents
- fix misc compiler warnings
- take maintainership

Upstream changes:

All changes are noted in ChangeLog and CHANGES.jp in distfile.
I won't bother to quote ~1700 line ChangeLogs here,
because this latest 3.7pl3 was released on 2004/05/20 and
the previous 3.6pl4 was 2003/09/22.
Highlight points are:
  - binary dictionary format is changed
  - shlib versions are bumped to 1.2 from 1.1
  - new APIs for wchar
  - experimental 64 bit architecture support
  - IPv6 support

Both server and clients (canuum and mule-1.1) are tested on
NetBSD/i386 7.0, NetBSD/amd64 7.0, and NetBSD/luna68k 7.0.
Also thanks to Nobuyoshi SATO, for his initial patch posted
to tech-pkg-ja@.

(tsutsui)

Updated www/drupal7 to 7.40

(wen)

Update to 7.40

Upstream changes:
Drupal 7.40, 2015-10-14
-----------------------
- Made Drupal's code for parsing .info files run much faster and use much less
  memory.
- Prevented drupal_http_request() from returning an error when it receives a
  201 through 206 HTTP status code.
- Added support for autoloading traits via the registry on sites running PHP
  5.4 or higher.
- Allowed the user-picture.tpl.php theme template to have HTML classes besides
  the default "user-picture" class printed in it (markup change).
- Fixed the URL text filter to convert e-mail addresses with plus signs into
  mailto: links.
- Added alternate text to file icons displayed by the File module, to improve
  accessibility (string change, and minor API addition to theme_file_icon()).
- Changed one-time login link failure messages to be displayed as errors or
  warnings as appropriate, rather than as regular status messages (minor UI
  change and data structure change).
- Changed the default settings.php configuration to exclude private files from
  the "404_fast_paths" behavior.
- Changed the page that displays filter tips for a particular text format, for
  example filter/tips/full_html, to return "page not found" or "access denied"
  if the format does not exist or the user does not have access to it. This
  change adds a new menu item to the Filter module's hook_menu() entry (minor
  data structure change).
- Added a new hook, hook_block_cid_parts_alter(), to allow modules to alter the
  cache keys used for caching a particular block.
- Made drupal_set_message() display and return messages when "0" is passed in
  as the message to set.
- Fixed non-functional "Files displayed by default" setting on file fields.
- The "worker callback" provided in hook_cron_queue_info() and the "finished"
  callback specified during batch processing can now be any PHP callable
  instead of just functions.
- Prevented drupal_set_time_limit() from decreasing the time limit in the case
  where the PHP maximum execution time is already unlimited.
- Changed the default thousand marker for numeric fields from a space ("1 000")
  to nothing ("1000") (minor UI change: https://www.drupal.org/node/1388376).
- Prevented malformed theme .info files (without a "name" key) from causing
  exceptions during menu rebuilds. If an .info file without a "name" key is
  found in a module or theme directory, Drupal will now use the module or
  theme's machine name as the display name instead.
- Made the format column in the {date_format_locale} database table
  case-sensitive, to match the equivalent column in the {date_formats} table.
- Fixed a bug in the Statistics module that caused JavaScript files attached to
  a node while it is being viewed to be omitted from the page.
- Added an optional 'project:' prefix that can be added to dependencies in a
  module's .info file to indicate which project the dependency resides in (API
  addition: https://www.drupal.org/node/2299747).
- Fixed various bugs that occurred after hooks were invoked early in the Drupal
  bootstrap and that caused module_implements() and drupal_alter() to cache an
  incomplete set of hook implementations for later use.
- Set the X-Content-Type-Options header to "nosniff" when possible, to prevent
  certain web browsers from picking an unsafe MIME type.
- Prevented the database API from executing multiple queries at once on MySQL,
  if the site's PHP version is new enough to do so. This is a secondary defense
  against SQL injection (API change: https://www.drupal.org/node/2463973).
- Fixed a bug in the Drupal 6 to Drupal 7 upgrade path which caused the upgrade
  to fail when there were multiple file records pointing to the same file.
- Numerous small bug fixes.
- Numerous API documentation improvements.
- Additional automated test coverage.

(wen)

Updated www/mediawiki to 1.25.3

(wen)

Updated textproc/p5-Pod-Markdown to 3.003

(wen)

Update to 1.25.3

Upstream changes:
== Security fixes ==

* Wikipedia user RobinHood70 reported two issues in the chunked upload API.
The API failed to correctly stop adding new chunks to the upload when the
reported size was exceeded (T91203), allowing a malicious users to upload add an
infinite number of chunks for a single file upload. Additionally, a
malicious user could upload chunks of 1 byte for very large files, potentially
creating a very large number of files on the server's filesystem (T91205).
<https://phabricator.wikimedia.org/T91203>
<https://phabricator.wikimedia.org/T91205>

* Internal review discovered that it is not possible to throttle file
uploads.
<https://phabricator.wikimedia.org/T91850>

* Internal review discovered a missing authorization check when removing
suppression from a revision. This allowed users with the 'viewsuppressed'
user right but not the appropriate 'suppressrevision' user right to unsuppress
revisions.
<https://phabricator.wikimedia.org/T95589>

* Richard Stanway from teamliquid.net reported that thumbnails of PNG files
generated with ImageMagick contained the local file path in the image
metadata.
<https://phabricator.wikimedia.org/T108616>

== Bug Fixes in 1.25.3 ==

* Fix having multiple callbacks for a single hook.
<https://phabricator.wikimedia.org/T98975>
* maintenance/refreshLinks.php did not always remove all links pointing to
  nonexistent pages.
<https://phabricator.wikimedia.org/T107632>
* $wgEmergencyContact and $wgPasswordSender now use their default value if
set to an empty string.
<https://phabricator.wikimedia.org/T104142>
* Provide fallbacks for use of mb_convert_encoding() in HtmlFormatter. It
was causing an error when accessing the api help page if the mbstring PHP
  extension was not installed.
<https://phabricator.wikimedia.org/T62174>
* Confirmation emails would sometimes contain invalid codes.
<https://phabricator.wikimedia.org/T105896>
* Fixed edit stash inclusion queries.
<https://phabricator.wikimedia.org/T105597>

(wen)

Whitespace.

(dholland)

Take a stab at enabling threads on openbsd (most of the logic seems to
already be in place, except in the configure script) now that openbsd
has threads.

(dholland)

Fix build with HAVE_ICONV not set, needed on OpenBSD.

(dholland)

Since the configure script objects to OpenBSD's grep and this depends
on coreutils anyway, have it always just use ggrep.

(dholland)

Improve the plop message generated when the OS isn't recognized.

(dholland)

Treat openbsd and bitrig as netbsd.

(dholland)

Updated devel/deforaos-libsystem to 0.2.3

(khorben)

Package DeforaOS libSystem 0.2.3

This release brings:
- fixes for memory leaks in the Config and Plugin classes
- fix for memory corruption in the Event class
- improvements to configctl(1)
- additional test for the Event class
- portability fixes (build...)

(khorben)

Turns out help2man is needed, so it is not in BUILD_DEPENDS

(adam)

Updated sysutils/duplicity to 0.6.26

(adam)

Changes 0.6.26:
Enhancements:
* Merged in lp:~m4ktub/duplicity/0.6-reliability
  - Per fix proposed in Bug 1395341.
* Modded .bzrignore to ignore *.egg test dependencies, normalized, sorted.
* Backport fix for bug 1278529
  - Use try/except around get_bucket() rather than bare lookup()
* Backport _librsyncmodule.c compilation bug

(adam)

Updated textproc/p5-Pod-Markdown to 3.003

(wen)

Update to 3.003

Upstream changes:
3.003    2015-09-28T00:29:45Z

  - Bump Pod::Simple prereq to fix bugs that were causing test failures.
  - Use Pod::Simple's nbsp_for_S(1) by default instead of handling S<> sequences internally.
    This means the feature can now also be disabled if desired.
  - Portability improvements for non-ascii platforms.

3.002    2015-08-21T03:24:24Z

  - Add attribution to changelog.

3.001    2015-08-21T03:17:57Z

  - Increase Pod::Simple requirement to 3.26 for detected_encoding().
    Thanks to Lee J for reporting! (gh-15).

3.000    2015-08-16T21:55:14Z

  [Bug Fixes]
  - Literal ampersands and left angle brackets are encoded as html entities
    when necessary to avoid interpretation as html.
    http://stackoverflow.com/questions/28496298/escape-angle-brackets-using-podmarkdown

  [API Changes]
  - Make accessors read/write (instead of read-only)
    for consistency with Pod::Simple classes.
  - Passing unknown arguments to the constructor now produces a warning
    and in the future may throw an error.
  - For consistency with the other html-entity changes, NBSP characters (U+00A0)
    are now used literally for S<> sequences.
    This also fixes a bug with code spans nested inside of S<> sequences.

  [New Attributes]
  - Add `html_encode_chars` attribute to allow customizing what characters
    should be html entity encoded.
  - Add `match_encoding` attribute to use the same encoding as the input pod
    when writing to the output handle.
  - Add `output_encoding` attribute to specify the desired encoding
    to apply to the output handle.

  [pod2markdown script]
  - Add command line options corresponding to new module attributes.
  - The script now defaults to UTF-8 encoding if no encoding options are specified.
    Previous versions did not produce consistent output and would sometimes emit warnings.
    Closes gh-14 and rt-101536.

  [Documentation]
  - Remove documentation for deprecated API
    (parse_from_file(), parse_from_filehandle(), and as_markdown())
    to further discourage use.

  [New Packages]
  - Added Pod::Perldoc::ToMarkdown to enable `perldoc -o Markdown ...`
    Pod::Markdown maintains a backward compatible interface that is incompatible
    with perldoc's `-o` so this module has been added to enable the functionality.
    Thanks to Alberto Sim繭es for investigating (gh-12) and providing some code to fix it (gh-13).

(wen)

Updated www/awstats to 7.4; devel/py-pbr to 1.8.1; devel/py-lockfile to 0.11.0

(adam)

Changes 0.11.0:
* Add deprecated warnings to index.rst
* Fix PIDLockFile.acquire() may loop indefinitely
* Fix failure - from lockfile import *
* lockfile.acquire doesn't accept a timeout of 0
* Update README format for our release script
* Begin moving some of the common code to a shared base
* Add pbr to dependency list
* The version of sphinx being brought in is broken
* Fix Git URLs

(adam)