the problem leading to PR pkg/44181 (PLIST problem) is understood
and reported upstream -- adjust comments

(drochner)

Revert last change

(adam)

Fix setting dovecot_fdlimit

(adam)

Note dependencies of kajongg

(markd)

Updated mail/dovecot2-pigeonhole to 0.2.2.

(ghen)

Update Dovecot Pigeonhole plugin to 0.2.2 (goes with Dovecot 2.0.8).

* LDA Sieve plugin: started using Dovecot LDA reject API for the
  reject extension. This means that the LDA reject_reason and
  reject_subject settings now also work for Pigeonhole's LDA Sieve
  plugin.
* Did some work on the new sieve-filter tool. It is mostly
  functional, but it is not finished yet.
* Dovecot change: services' default vsz_limits weren't being enforced
  correctly in earlier v2.0 releases. Now that they are enforced, you
  might notice that the default limits are too low and you need to
  increase them. This problem will show up in logs as "out of memory"
  errors. See default_vsz_limit and service { vsz_limit } settings.
- Imap4flags: fixed segfault bug occurring in multiscript context.
- Added version checking to the ManageSieve settings plugin. This
  plugin was forgotten when the LDA plugin was updated with this
  change in the previous release.
- LDA Sieve plugin: fixed memory leak at deinitialization.

(ghen)

sync w/ base pkg

(drochner)

update odt2txt to 0.4

The changes include:
Mention other file formats than ODT in the manual page
Remove check for mime type
Fix typo in manual page
Make iconv usage configurable
Build on HP-UX 11

This is for PR #43599

(reed)

Update from maintainer.

Update package database/ipa_sdb to 1.0.2 version.

Changes in the package:

    *  MASTER_SITES was changed.

    *  DESCR was updated.

    *  PLIST was updated.

Changes in the ipa_sdb-1.0.2:

    *  autoconf 2.59 -> 2.68, automake 1.9.6 -> 1.11.1,
        libtool 1.5.20 -> 2.2.10.

    *  Code optimization and documentation corrections.

    *  configure.ac was optimized.

    *  Now lint can check source code of IPA_SDB.

    *  If there is close_fd=yes for rule and db_set_limit_state() fails,
        then the ipa_db_sdb module did not close database file descriptor
        for a limit in some cases.

    *  Added checks for uint8_t, uint16_t, uint32_t and uint64_t
        to configure.  Check that sizeof(long) >= sizeof(int).

    *  Now autoheader is used for generating src/config.h.in.

    *  Now src/ipa_sdb.h is not generated, instead new sdb_config.in.h
        file is used.

    *  New options for configure: --enable-gcc-more-warns,
        --disable-gcc-warns, --with-db-pthread, --with-st-pthread.

    *  Options --enable-FEATURE and --disable-FEATURE disabled FEATURE
        in the configure script.

    *  Use readdir_r() function in thread-safe version of the ipa_st_sdb
        module.

    *  Opened directories never were closed in the ipa_st_sdb module.

(reed)

Updated sysutils/fabric to 0.9.3

(gls)

Update fabric to 0.9.3

Pkgsrc changes:
- add depends to py-crypto

Upstream changes:

0.9.3
=====

Feature additions

* #255: Added stderr and succeeded attributes to local.
* #254: Backported the .stderr and .succeeded attributes on run/sudo return
values, from the Git master/pre-1.0 branch. Please see those functions? API
docs for details.

Bugfixes

* #228: We discovered that the pip + PyCrypto installation problem was limited
to Python 2.5 only, and have updated our setup.py accordingly.
* #230: Arbitrary or remainder commands (fab <opts> -- <run command here>) will
no longer blow up when invoked with no fabfile present. Thanks to IRC user
orkaa for the report.
* #242: Empty string values in task CLI args now parse correctly. Thanks to
Aaron Levy for the catch + patch.

Documentation updates

* #239: Fixed typo in execution usage docs. Thanks to Pradeep Gowda and Turicas
for the catch.

0.9.2
=====

Feature additions

* The reboot operation has been added, providing a way for Fabric to issue a
reboot command and then reconnect after the system has restarted.
* python setup.py test now runs Fabric?s test suite (provided you have all the
prerequisites from the requirements.txt installed). Thanks to Eric Holscher for
the patch.
* Added functionality for loading fabfiles which are Python packages
(directories) instead of just modules (single files.) See Fabfile discovery.
* Added output lines informing the user of which tasks are being executed (e.g.
[myserver] Executing task 'foo'.)
* Added support for lazy (callable) role definition values in env.roledefs.
* Added contrib.django module with basic Django integration.
* env.local_user was added, providing easy and permanent access to the local
system username, even if an alternate remote username has been specified.
* #29: Added support for arbitrary command-line-driven anonymous tasks via
fab [options] -- [shell command]. See Arbitrary remote shell commands.
* #52: Full tracebacks during aborts are now displayed if the user has opted to
see debug-level output.
* #101: Added colors module with basic color output support. (#101 is still
open: we plan to leverage the new module in Fabric?s own output in the future.)
* #137: Commas used to separate per-task arguments may now be escaped with a
backslash. Thanks to Erich Heine for the patch.
* #144: hosts (and roles) will now expand a single, iterable argument instead of
requiring one to use e.g. @hosts(*iterable).
* #151: Added a puts utility function, which allows greater control over
fabfile-generated (as opposed to Fabric-generated) output. Also added fastprint,
an alias to puts allowing for convenient unbuffered, non-newline-terminated
printing.
* #208: Users rolling their own shell completion or who otherwise find
themselves performing text manipulation on the output of --list may now use
--shortlist to get a plain, newline-separated list of task names.

Bugfixes
* The interactive ?what host to connect to?? prompt now correctly updates the
appropriate environment variables (hostname, username, port) based on user input.
* Fixed a bug where Fabric?s own internal fabfile would pre-empt the user?s
fabfile due to a PYTHONPATH order issue. User fabfiles are now always loaded at
the front of the PYTHONPATH during import.
* Disabled some DeprecationWarnings thrown by Paramiko when that library is
imported into Fabric under Python 2.6.
* #44, #63: Modified rsync_project to honor the SSH port and identity file
settings. Thanks to Mitch Matuson and Morgan Goose.
* #123: Removed Cygwin from the ?are we on Windows? test; now, only Python
installs whose sys.platform says 'win32' will use Windows-only code paths
(e.g. importing of pywin32).

Documentation updates
* Added a few new items to the FAQ.
* #173: Simple but rather embarrassing typo fix in README. Thanks to Ted Nyman
for the catch.
* #194: Added a note to the install docs about a possible edge case some
Windows 64-bit Python users may encounter.
* #216: Overhauled the process backgrounding FAQ to include additional
techniques and be more holistic.

Packaging updates
* #86, #158: Removed the bundled Paramiko 1.7.4 and updated the setup.py to
require Paramiko >=1.7.6. This lets us skip the known-buggy Paramiko 1.7.5 while
getting some much needed bugfixes in Paramiko 1.7.6.

(gls)

Updated security/py-paramiko to 1.7.6nb2

(gls)

Direct MASTER_SITES to pypi.org, as the original MASTER_SITE doesn't host the
distfile anymore. Add DIST_SUBDIR to force re-download of the tarball, which
content has been modified.

While here, switch DEPENDS to py-crypto instead of py-amkCrypto.

Bump PKG_REVISION.

(gls)

Updated misc/stellarium to 0.10.6

(adam)

Changes 0.10.6:
* New feature for installing landscapes from ZIP archives.
* New plugin: Solar System editor.
* New plugin: Time Zone manual override.
* New translatable strings: stars and nebulae.
* Selected plugins loaded by default.
* Fixed on change of sky culture.
* Fixed clearing of single constellation selection.
* Fixed constellation art engine, removing distortion of images.
* Changed GUI styles (all dialogs now use "Title Bar", the two different tabbed
  widget implementations are styled consistently).
* Changed Solar System engine, allowing "Comets" and "Minor planets" to be
  handled as classes of objects separate from "Planets".
* Translatable plugin names, descriptions and some other strings.
* Oculars plugin: now uses a configuration file instead of SQL.
* Oculars plugin: added manual override of keyboard shortcuts.
* Oculars plugin: CCD sensors formula fixed.
* Satellites plugin: added new orbit prediction engine.
* Satellites plugin: added trajectory visualization.
* Satellites plugin: fixed problems leaving satellite TLEs out of date.
* Satellites plugin: can now update TLEs from a local file.
* Satellites plugin: satellite visibility can now be saved from satellites
  dialog.
* Telescope control plugin: fixed bad serial port names on Windows.
* Telescope control plugin: fixed telescopes disappearance on startup.
* Telescope control plugin: added manual equinox/epoch override.
* Various rendering problems resolved.
* SQLite dependency dropped. The Qt SQLite driver no longer needs to be
  packaged with Stellarium.

(adam)

Upgrade lcov to 1.9. Branch coverage analysis added and defaults to on.

>From the CHANGES:

Version 1.9
===========

genhtml:
- Improved wording for branch representation tooltip text
- Fixed vertical alignment of HTML branch representation

geninfo:
- Improved warning message about --initial not generating branch coverage data
- Debugging messages are now printed to STDERR instead of STDOUT
- Fixed problem with some .gcno files. Reported by gui@futarque.com.
  (file.gcno: reached unexpected end of file)
- Fixed problem with relative build paths. Reported by zhanbiao2000@gmail.com.
  (cannot find an entry for ^#src#test.c.gcov in .gcno file, skipping file!)
- Fixed problem where coverage data is missing for some files. Reported by
  weston_schmidt@open-roadster.com
- Fixed problem where exclusion markers are ignored when gathering
  initial coverage data. Reported by ahmed_osman@mentor.com.
- Fixed large execution counts showing as negative numbers in HTML output.
  Reported by kkyriako@yahoo.com.
- Fixed problem that incorrectly associated branches outside of a block with
  branches inside the first block

lcov:
- Fixed problem that made lcov ignore --kernel-directory parameters when
  specifying --initial. Reported by hjia@redhat.com.
- Added --list-full-path option to prevent lcov from truncating paths in list
  output
- Added lcov_list_width and lcov_list_truncate_max directives to the
  lcov configuration file to allow for list output customization
- Improved list output

COPYING:
- Added license text to better comply with GPL recommendations

Version 1.8
===========

gendesc:
- Fixed problem with single word descriptions

genhtml:
- Added support for branch coverage measurements
- Added --demangle-cpp option to convert C++ function names to human readable
  format. Based on a patch by slava.semushin@gmail.com.
- Improved color legend: legend display takes up less space in HTML output
- Improved coverage rate limits: all coverage types use the same limits
  unless specified otherwise
- Fixed CRLF line breaks in source code when generating html output. Based
  on patch by michael.knigge@set-software.de.
- Fixed warning when $HOME is not set
- Fixed problem with --baseline-file option. Reported by sixarm@gmail.com.
  (Undefined subroutine &main::add_fnccounts called at genhtml line 4560.)
- Fixed problem with --baseline-file option and files without function
  coverage data (Can't use an undefined value as a HASH reference at genhtml
  line 4441.)
- Fixed short-name option ambiguities
- Fixed --highlight option not showing line data from converted test data
- Fixed warnings about undefined value used. Reported by nikita@zhuk.fi.
- Fixed error when processing tracefiles without function data. Reported
  by richard.corden@gmail.com (Can't use an undefined value as a HASH
  reference at genhtml line 1506.)

geninfo:
- Added support for branch coverage measurements
- Added support for exclusion markers: Users can exclude lines of code from
  coverage reports by adding keywords to the source code.
- Added --derive-func-data option
- Added --debug option to better debug problems with graph files
- Fixed CRLF line breaks in source code when generating tracefiles. Based on
  patch by michael.knigge@set-software.de.
- Fixed problems with unnamed source files
- Fixed warning when $HOME is not set. Reported by acalando@free.fr.
- Fixed errors when processing unnamed source files
- Fixed help text typo
- Fixed errors when processing incomplete function names in .bb files
- Fixed filename prefix detection
- Fixed problem with matching filename
- Fixed problem when LANG is set to non-english locale. Reported by
  benoit_belbezet@yahoo.fr.
- Fixed short-name option ambiguities

genpng:
- Fixed runtime-warning

lcov:
- Added support for branch coverage measurements
- Added support for the linux-2.6.31 upstream gcov kernel support
- Added --from-package and --to-package options
- Added --derive-func-data option
- Added overall coverage result output for more operations
- Improved output of lcov --list
- Improved gcov-kernel handling
- Fixed minor problem with --diff
- Fixed double-counting of function data
- Fixed warning when $HOME is not set. Reported by acalando@free.fr.
- Fixed error when combining tracefiles without function data. Reported by
  richard.corden@gmail.com. (Can't use an undefined value as a HASH reference
  at lcov line 1341.)
- Fixed help text typo
- Fixed filename prefix detection
- Fixed lcov ignoring information about converted test data

README:
- Added note to mention required -lgcov switch during linking

(reed)

Added net/samba35 version 3.5.6

(adam)

Added net/samba35 version 3.5.6

(adam)

Samba provides file and print services for Microsoft Windows clients.
These services may be hosted off any TCP/IP-enabled platform.  The
Samba project includes not only an impressive feature set in file and
print serving capabilities, but has been extended to include client
functionality, utilities to ease migration to Samba, tools to aid
interoperability with Microsoft Windows, and administration tools.

Status:

Vendor Tag: TNF
Release Tags: pkgsrc-base

(adam)

Add upstream bug report URL.

(wiz)

Seems like Darwin also supports ACLs

(adam)

Updated www/varnish to 2.1.4

(wiz)

Update to 2.1.4, provided by Matthew Sporleder in PR 44165.
Includes new startup scripts.

Summary of changes from 2.1.3 to 2.1.4

    * A bug in the binary heap layout caused inflated object counts, this has been fixed.
    * Much more comprehensive documentation.
    * A DNS director that uses DNS lookups for choosing which backend to route requests to has been added.
    * The client director now uses the variable client.identity for choosing which backend to send a given request to.
    * String representation of now, making it easier to add Expires headers
    * Portability fixes for Solaris.
    * Various bug fixes.

Summary of changes from 2.1.2 to 2.1.3

    * The scalability of critbit, the default hashing method, has been improved.
    * A bug in varnishd would in some cases confuse varnishncsa leading to lost or wrong log lines.
    * Some bugs in the handling of Range requests has been fixed. This only matters if you enable Range support.
    * Add 束log損 command to VCL which will log to the Varnish log.

Summary of changes from 2.1.1 to 2.1.2

    * When adding Range support in 2.1.1 an error was introduced. Garbage was appended to some objects. This affected some load balancers - but clients seemed not to be affected.

Summary of changes from 2.1.0 to 2.1.1

    * Experimental support for the Range header, must be enabled with the http_range_support parameter.
    * A bug in workspace rollback prevented ESI from working correctly in all situations. This is now fixed.
    * A race condition and a deadlock in the critbit hasher have been fixed.
    * HEAD requests are no longer converted to GET requests for pass and pipe.
    * Support for completely obliterating objects including all variants (束nuke損). See http://varnish-cache.org/changeset/4668 for details

Summary of changes from 2.0.6 to 2.1.0

Varnish 2.1.0 has just been released. It contains lots of changes relative to 2.0.6, the most important ones being:

    * Experimental support for persistent cache
    * The regular expression engine is now PCRE
    * Saint mode, where we can serve a cached object if the backend is down or otherwise faulty.
    * A more scalable hashing method called critbit
    * Increased scalability, removing a limitation of maximum 64k connections
    * obj_workspace is removed, this is now scaled automatically
    * Hashing and client IP based directors

Note: The release has some syntaxchanges in VCL, please read the full changelog

(wiz)

Change MAINTAINER on bzr packages.

With epg@'s concurrence, reset MAINTAINER on *bzr* to Jan Danielsson.

(gdt)

Fix build on DragonFly, from Rumko in PR 44193.

(wiz)

Update KDE

(markd)

Find recent libwpd, but remove koffice-libwpd option from the default
build as currently doesn't build against it.  Bump PKGREVISION.

(markd)

Fix build with recent subversion.

(markd)

Don't try to build and install doxygen generated docs if doxygen is found

(markd)

Update to KDE SC 4.5.4

4.5.4 brings bugfixes and translation updates

(markd)

ftp.conserver.com re-directs to a machine that does not run an ftp
server, so fetch the sources via http.

Sort out pkg version, while we are here.

(hauke)

fix PLIST for last

(abs)

Remove comment-less patch which isn't needed to build on NetBSD-current.

(wiz)

Updated graphics/kipi-plugins to 1.6.0nb1

(wiz)

Now that opencv is new enough, enable redeyes plugin.
Bump PKGREVISION.

(wiz)

Updated inputmethod/ibus-table-chinese to 1.3.0.20101206

(obache)

Update ibus-table-chinses to 1.3.0.20101206

- New tables which was not include in original:
  + array30, array30-big, wubi-haifeng
- Table removed from original version because of license issues:
  + zhengma, ziranma
- Add package review ID
- Add build tag for el6

(obache)

+ ImageMagick-6.6.6.3, JBidwatcher-2.1.3, balance-3.54, calibre-0.7.32,
  clive-2.2.18, eric4-4.4.10, feh-1.10.1, gmime24-2.4.21,
  gnome-commander-1.2.8.9, gssdp-0.9.1, gupnp-0.15.0, libpcap-1.1.1,
  mu-0.9, p5-PerlMagick-6.6.6.3, py-dbus-0.83.2, rasqal-0.9.21,
  scmgit-base-1.7.3.3, scmgit-docs-1.7.3.3, shared-mime-info-0.90,
  snort-2.9.0.2, tcpdump-4.1.1, vigra-1.7.1, warzone2100-2.3.6,
  wine-1.2.2, x264-devel-20101204, xf86-video-savage-2.3.2.

(wiz)

Pullup tickets #3293, #3294 and #3295.

(tron)

Pullup ticket #3295 - requested by bouyer
mail/clamav: bug fix update

Revisions pulled up:
- mail/clamav/Makefile 1.106-1.108
- mail/clamav/distinfo 1.68-1.69
- mail/clamav/patches/patch-ab 1.14
- mail/clamav/patches/patch-ac deleted
- mail/clamav/patches/patch-ah deleted
---
Module Name:    pkgsrc
Committed By:  adam
Date:          Tue Oct 26 06:38:37 UTC 2010

Modified Files:
        pkgsrc/mail/clamav: Makefile distinfo

Log Message:
Changes 0.96.4:
* This is a bugfix release recommended for all users.
---
Module Name:    pkgsrc
Committed By:  bouyer
Date:          Thu Oct 28 20:15:38 UTC 2010

Modified Files:
        pkgsrc/mail/clamav: Makefile distinfo
Removed Files:
        pkgsrc/mail/clamav/patches: patch-ah

Log Message:
Revert previous, this patch is already in 0.96.4.
---
Module Name:    pkgsrc
Committed By:  adam
Date:          Wed Dec  1 07:44:10 UTC 2010

Modified Files:
        pkgsrc/mail/clamav: Makefile distinfo
        pkgsrc/mail/clamav/patches: patch-ab
Removed Files:
        pkgsrc/mail/clamav/patches: patch-ac

Log Message:
Changes 0.96.5:
* libclamav/pdf.c: fix crashes
* libclamav/pe_icons.c: off by one while
* libclamav: fix detection of embedded executables
* libclamav/matcher-ac.c: fix offset handling for sigs with {x-y} wildcards
* freshclam/manager.c: fix error path infinite loop
* clamd/clamd.c: fix RLIMIT_DATA setting on BSD
* freshclam: improve mirror management
* libclamav: fix possible use of uninitialized values
* libclamav: Set the unreliability flag on (un)packed files
* libclamav/c++: Update embedded copy of LLVM to version 2.8
* freshclam: make query format backward compatible
* freshclam: get detection stats directly from clamd
* libclamav/cache.c,c++/bytecode2llvm.cpp}: make cl_load thread safe
* freshclam: load database in subprocess
* clamd: add new commands DETSTATS and DETSTATSCLEAR
* libclamav/7z.c: fix file descriptor leak
* clamd, libclamavll: add ability to logg messages from libclamav
* libclamav/builtin_bytecodes.h: Don't disable JIT on pentium4

(tron)

Updated graphics/kdegraphics4 to 4.5.3nb2

(wiz)

Add ../../sysutils/desktop-file-utils/desktopdb.mk and bump PKGREVISION.

(wiz)

Remove trailing whitespace.

(wiz)

Add patch to fix build with exiv2-0.21. From upstream SVN.

(wiz)

Detect builtin MIT Kerberos on Mac OS X 10.6.x

(adam)

Updated comms/conserver8 to 8.1.18

(hauke)

Updated comms/conserver8 to 8.1.18

(hauke)

Updating conserver 8 to v8.18

version 8.1.18 (Nov 11, 2010):
        - install man pages read-only and improved the contributed redhat init
          script - patches by Eric Biederman <ebiederm@aristanetworks.com>
        - spec file improvements in contrib/redhat-rpm - patch by Jodok Ole
          Muellers <jodok.muellers@aschendorff.de>
        - GSS-API patch for client code - patch by Andras Horvath
          <Andras.Horvath@cern.ch>

version 8.1.17 (Sep 29, 2009):
        - fix for interface detection when HAVE_SA_LEN is defined - first
          detected on NetBSD 5.0 and patched by Chris Ross
          <cross+conserver@distal.com>
        - first person to connect to a console wanting read/write now gets it
          once the active user drops read/write - suggested by Thomas Gardner
          <tmg@pobox.com>
        - fix typo when setting nonblocking socket for client connections,
          fixing stall issues - patch by Eric Biederman
          <ebiederm@aristanetworks.com>
        - GSS-API patch (--with-gssapi) to help with Kerberos tokens - patch by
          Nate Straz <nstraz@redhat.com>
        - authenticate username without @REALM when using GSS-API
          (--with-striprealm) - based on patch by Andras Horvath
          <Andras.Horvath@cern.ch>
        - various contrib/redhat-rpm fixes - patch by Fabien Wernli
          <wernli@in2p3.fr>
        - fix handling of read(stdin) returning -1 in console client - patch by
          Ed Swierk <eswierk@arastra.com>

patch-ac has been included upstream.

(hauke)

Pullup ticket #3294 - requested by bouyer
net/ocsinventory-agent: compatibility update

Revisions pulled up:
- net/ocsinventory-agent/Makefile 1.3
- net/ocsinventory-agent/distinfo 1.2
- net/ocsinventory-agent/patches/patch-ab 1.2
- net/ocsinventory-agent/patches/patch-ac 1.2
- net/ocsinventory-agent/patches/patch-ap 1.1
---
Module Name: pkgsrc
Committed By: bouyer
Date: Wed Dec  1 21:00:51 UTC 2010

Modified Files:
pkgsrc/net/ocsinventory-agent: Makefile distinfo
pkgsrc/net/ocsinventory-agent/patches: patch-ab patch-ac
Added Files:
pkgsrc/net/ocsinventory-agent/patches: patch-ap

Log Message:
Update to 1.1.2.1. Main motivation is to avoid warnings with perl 5.12.
Other changes are mainly improvements in backend to support more OS,
more OS variants or versions.

(tron)

Pullup ticket #3293 - requested by bouyer
mail/imp: dependence fix

Revisions pulled up:
- mail/imp/Makefile 1.56
---
Module Name: pkgsrc
Committed By: bouyer
Date: Sun Nov 14 21:17:47 UTC 2010

Modified Files:
pkgsrc/mail/imp: Makefile

Log Message:
php5-pear-Mail_mimeDecode is required to properly display certain
mail (especially ones with a smime.p7s signature).
Bump PKGREVISION

(tron)

Bump revision for default options changes on DragonFly.

(asau)

Default to use kqueue on DragonFly.
Requested by Rumko in PR pkg/44194

(asau)

Note update of www/wordpress to 3.0.2

(morr)

Mention security/py-crypto update to 2.3

(gls)

Update py-crypto to version 2.3

pkgsrc changes:
- add a patch to fix an hardcoded interpreter path
- add a patch to actually find gmp library.
- add CONFLICTS with py-amkCrypto, both have files in common.

upstream changes:

2.3
===
* Fix NameError when attempting to use deprecated getRandomNumber()
function.
* _slowmath: Compute RSA u parameter when it's not given to
RSA.construct. This makes _slowmath behave the same as _fastmath in
this regard.
* Make RSA.generate raise a more user-friendly exception message when
the user tries to generate a bogus-length key.

2.2
===
* Deprecated Crypto.Util.number.getRandomNumber(), which had confusing
semantics.  It's been replaced by getRandomNBitInteger and
getRandomInteger.  (Thanks: Lorenz Quack)
* Better isPrime() and getPrime() implementations that do a real
Rabin-Miller probabilistic primality test (not the phony test we did
before with fixed bases).  (Thanks: Lorenz Quack)
* getStrongPrime() implementation for generating RSA primes.
(Thanks: Lorenz Quack)
* Support for importing and exporting RSA keys in DER and PEM format
(Thanks: Legrandin)
* Fix PyCrypto when floor division (python -Qnew) is enabled.
* When building using gcc, use -std=c99 for compilation.  This should
fix building on FreeBSD and NetBSD.

(gls)

Security update. Changes:

* Fix moderate security issue where a malicious Author-level user could gain further access to the site.

* Remove pingback/trackback blogroll whitelisting feature as it can easily be abused.
* Fix canonical redirection for permalinks containing %category% with nested categories and paging.
* Fix occasional irrelevant error messages on plugin activation.
* Minor XSS fixes in request_filesystem_credentials() and when deleting a plugin.
* Clarify the license in the readme
* Multisite: Fix the delete_user meta capability
* Multisite: Force current_user_can_for_blog() to run map_meta_cap() even for super admins
* Multisite: Fix ms-files.php content type headers when requesting a URL with a query string
* Multisite: Fix the usage of the SUBDOMAIN_INSTALL constant for upgraded WordPress MU installs

While here, set license.

(morr)

Make mplayer build on NetBSD/sh3el.

Build tested on NetBSD/landisk 5.1 and player tested on NetBSD/dreamcast 5.1.
(though some more packages need hacks/fixes for XFree86 and SH3)

(tsutsui)

Add patches for GLib>2.25, taken from upstream.

(obache)

Updated devel/cpuflags to 1.42

(abs)

Updated devel/cpuflags to 1.42

Add initial powerpc support for Darwin

(abs)

Updated graphics/opencv to 2.2.0

(wiz)

Update to 2.2.0.

2.2

  (December, 2010)

  General Modifications and Improvements
    * The library has been reorganized. Instead of cxcore, cv, cvaux,
      highgui and ml we now have several smaller modules:
          * opencv_core - core functionality (basic structures, arithmetics
            and linear algebra, dft, XML and YAML I/O ...).
          * opencv_imgproc - image processing (filter, GaussianBlur, erode,
            dilate, resize, remap, cvtColor, calcHist etc.)
          * opencv_highgui - GUI and image & video I/O
          * opencv_ml - statistical machine learning models (SVM, Decision
            Trees, Boosting etc.)
          * opencv_features2d - 2D feature detectors and descriptors (SURF,
            FAST etc.,
              * including the new feature detectors-descriptor-matcher
                framework)
          * opencv_video - motion analysis and object tracking (optical flow,
            motion templates, background subtraction)
          * opencv_objdetect - object detection in images (Haar & LBP face
            detectors, HOG people detector etc.)
          * opencv_calib3d - camera calibration, stereo correspondence and
            elements of 3D data processing
          * opencv_flann - the Fast Library for Approximate Nearest Neighbors
            (FLANN 1.5) and the OpenCV wrappers
          * opencv_contrib - contributed code that is not mature enough
          * opencv_legacy - obsolete code, preserved for backward
            compatibility
          * opencv_gpu - acceleration of some OpenCV functionality using CUDA
            (relatively unstable, yet very actively developed part of OpenCV)
    * If you detected OpenCV and configured your make scripts using CMake or
      pkg-config tool, your code will likely build fine without any changes.
      Otherwise, you will need to modify linker parameters (change the
      library names) and update the include paths.
    * It is still possible to use #include <cv.h> etc. but the recommended
      notation is:
          * #include "opencv2/imgproc/imgproc.hpp"
          * ..
    * Please, check the new C and C++ samples
      (https://code.ros.org/svn/opencv/trunk/opencv/samples), which now
      include the new-style headers.
    * The new-style wrappers now cover much more of OpenCV 2.x API. The
      documentation and samples are to be added later. You will need numpy
      in order to use the extra added functionality.
          * SWIG-based Python wrappers are not included anymore.
    * OpenCV can now be built for Android (GSoC 2010 project), thanks to
      Ethan Rublee; and there are some samples too. Please, check
      http://opencv.willowgarage.com/wiki/Android
    * The completely new opencv_gpu acceleration module has been created
      with support by NVidia. See below for details.

  New Functionality, Features
    * core:
          * The new cv::Matx<T, m, n> type for fixed-type fixed-size matrices
            has been added. Vec<T, n> is now derived from Matx<T, n, 1>. The
            class can be used for very small matrices, where cv::Mat use
            implies too much overhead. The operators to convert Matx to Mat
            and backwards are available.
          * cv::Mat and cv::MatND are made the same type: typedef cv::Mat
            cv::MatND. Note that many functions do not check the matrix
            dimensionality yet, so be careful when processing 3-, 4- ...
            dimensional matrices using OpenCV.
          * Experimental support for Eigen 2.x/3.x is added (WITH_EIGEN2
            option in CMake). Again, there are convertors from Eigen2
            matrices to cv::Mat and backwards. See
            modules/core/include/opencv2/core/eigen.hpp.
          * cv::Mat can now be print with "<<" operator. See
            opencv/samples/cpp/cout_mat.cpp.
          * cv::exp and cv::log are now much faster thanks to SSE2
            optimization.
    * imgproc:
          * color conversion functions have been rewritten;
              * RGB->Lab & RGB->Luv performance has been noticeably
                improved. Now the functions assume sRGB input color space
                (e.g. gamma=2.2). If you want the original linear RGB->L**
                conversion (i.e. with gamma=1), use CV_LBGR2LAB etc.
              * VNG algorithm for Bayer->RGB conversion has been added. It's
                much slower than the simple interpolation algorithm, but
                returns significantly more detailed images
              * The new flavors of RGB->HSV/HLS conversion functions have
                been added for 8-bit images. They use the whole 0..255 range
                for the H channel instead of 0..179. The conversion codes
                are CV_RGB2HSV_FULL etc.

          * special variant of initUndistortRectifyMap for wide-angle cameras
            has been added: initWideAngleProjMap()
    * features2d:
          * the unified framework for keypoint extraction, computing the
            descriptors and matching them has been introduced. The previously
            available and some new detectors and descriptors, like SURF,
            Fast, StarDetector etc. have been wrapped to be used through the
            framework. The key advantage of the new framework (besides the
            uniform API for different detectors and descriptors) is that it
            also provides high-level tools for image matching and textured
            object detection. Please, see documentation
            http://opencv.willowgarage.com/documentation/cpp/features2d_common_interfaces_of_feature_detectors.html
              * and the C++ samples:
                    * descriptor_extractor_matcher.cpp - finding object in a
                      scene using keypoints and their descriptors.
                    * generic_descriptor_matcher.cpp - variation of the above
                      sample where the descriptors do not have to be computed
                      explicitly.
                    * bagofwords_classification.cpp - example of extending
                      the framework and using it to process data from the VOC
                      databases:
                        * http://pascallin.ecs.soton.ac.uk/challenges/VOC/

          * the newest super-fast keypoint descriptor BRIEF by Michael
            Calonder has been integrated by Ethan Rublee. See the sample
            opencv/samples/cpp/video_homography.cpp
          * SURF keypoint detector has been parallelized using TBB (the patch
            is by imahon and yvo2m)
    * objdetect:
          * LatentSVM object detector, implementing P. Felzenszwalb
            algorithm, has been contributed by Nizhniy Novgorod State
            University (NNSU) team. See
              * opencv/samples/c/latentsvmdetect.cpp
    * calib3d:
          * The new rational distortion model:
              * x' = x*(1 + k1*r2 + k2*r4 + k3*r6)/(1 + k4*r2 + k5*r4 +
                k6*r6) + <tangential_distortion for x>,

                y' = y*(1 + k1*r2 + k2*r4 + k3*r6)/(1 + k4*r2 + k5*r4 +
                k6*r6) + <tangential_distortion for y>

              * has been introduced. It is useful for calibration of cameras
                with wide-angle lenses. Because of the increased number of
                parameters to optimize you need to supply more data to
                robustly estimate all of them. Or, simply initialize the
                distortion vectors with zeros and pass

                CV_CALIB_RATIONAL_MODEL to enable the new model +
                CV_CALIB_FIX_K3 + CV_CALIB_FIX_K4 + CV_CALIB_FIX_K5 or other
                such combinations to selectively enable or disable certain
                coefficients.

          * rectification of trinocular camera setup, where all 3 heads are
            on the same line, is added. see samples/cpp/3calibration.cpp
    * ml:
          * Gradient boosting trees model has been contributed by NNSU team.
    * highgui:
          * Experimental Qt backend for OpenCV has been added as a result of
            GSoC 2010 project, completed by Yannick Verdie. The backend has a
            few extra features, not present in the other backends, like text
            rendering using TTF fonts, separate "control panel" with sliders,
            push-buttons, checkboxes and radio buttons, interactive zooming,
            panning of the images displayed in highgui windows, "save as"
            etc. Please, check the youtube videos where Yannick demonstrates
            the new features: http://www.youtube.com/user/MrFrenchCookie#p/u
              * The new API is described here:
                http://opencv.willowgarage.com/documentation/cpp/highgui_qt_new_functions.html
                To make use of the new API, you need to have Qt SDK (or
                libqt4 with development packages) installed on your machine,
                and build OpenCV with Qt support (pass -DWITH_QT=ON to
                CMake; watch the output, make sure Qt is used as GUI
                backend)

          * 16-bit and LZW-compressed TIFFs are now supported.
          * You can now set the mode for IEEE1394 cameras on Linux.
    * contrib:
          * Chamfer matching algorithm has been contributed by Marius Muja,
            Antonella Cascitelli, Marco Di Stefano and Stefano Fabri. See
            samples/cpp/chamfer.cpp
    * gpu:
          * This is completely new part of OpenCV, created with the support
            by NVidia. Note that the package is at alpha, probably early beta
            state, so use it with care and check OpenCV SVN for updates.

            In order to use it, you need to have the latest NVidia CUDA SDK
            installed, and build OpenCV with CUDA support (-DWITH_CUDA=ON
            CMake flag). All the functionality is put to cv::gpu namespace.
            The full list of functions and classes can be found at
            opencv/modules/gpu/include/opencv2/gpu/gpu.hpp, and here are some
            major components of the API:
              * image arithmetics, filtering operations, morphology,
                geometrical transformations, histograms
              * 3 stereo correspondence algorithms: Block Matching, Belief
                Propagation and Constant-Space Belief Propagation.
              * HOG-based object detector. It runs more than order of
                magnitude faster than the CPU version!
                    * See opencv/samples/cpp/

    * python bindings:
          * A lot more of OpenCV 2.x functionality is now covered by Python
            bindings.

  Documentation, Samples
    * Links to wiki pages (mostly empty) have been added to each function
      description, see http://opencv.willowgarage.com
    * All the samples have been documented; most samples have been converted
      to C++ to use the new OpenCV API.

  Bug Fixes
    * Over 300 issues have been resolved. Most of the issues (closed and
      still open) are listed at https://code.ros.org/trac/opencv/report/6.

(wiz)

Update devel/apache-ant to 1.8.1
Fixes PR pkg/43752

(tonio)

Update devel/apache-ant to 1.8.1
Fixes PR pkg/43752

Changes from Ant 1.8.0 TO Ant 1.8.1:
- Changes that could break older environments:
* ant-trax.jar is no longer produced since TrAX is included in JDK 1.4+.
* Ant no longer ships with Apache Xerces-J or the XML APIs but relies
  on the Java runtime to provide a parser and matching API versions.
* The stylebook ant task and the ant-stylebook.jar are removed.

- Fixed bugs:
* Tasks that iterate over task or type definitions, references or
  targets now iterate over copies instead of the live maps to avoid
  ConcurrentModificationExceptions if another thread changes the
  maps.
* The filesmatch condition threw a NullPointerException when
  comparing text files and the second file contained fewer lines than
  the first one.
* Regression: The <ear> task would allow multiple
  META-INF/application.xml files to be added.
* VectorSet#remove(Object) would fail if the size of the vector
  equaled its capacity.
* Regression : ant -diagnostics was returning with exit code 1
* Fix for exec task sometimes inserts extraneous newlines
* SymlinkTest#testSymbolicLinkUtilsMethods failing on MacOS
* If <concat>'s first resourcecollection child is a <resources>,
  any subsequently added child resourcecollection joins the first.
* <get> with an invalid URL could trigger an NPE in some JVMs.
* Broken Pipe issue under Ubuntu Linux
* Properties wrongly read from file or not update during read
* AntClassLoader in Ant 1.8.0 has been considerably slower than in
  1.7.1
* ANT_CMD_LINE_ARGS are rippling through lower level Ant usage
* email : IO error sending mail with plain mimetype
* the complete-ant-cmd.pl script failed to create a proper cache of
  target if "ant -p" failed.
* <rmic>'s sourcebase attribute was broken.
* <copy>'s failonerror didn't work as expected when copying a single
  element resource collection to a file.
* <get> no longer followed redirects if the redirect URL was relative
  and not an absolute URL.
* fixed a performance degradation in the code that expands property
  references.
* <jar filesetmanifest="merge"> was broken on Windows.
* <symlink> delete failed if the link attribute was a relative path
  to a link inside the current directory without a leading ".".
* <telnet> and <rexec> failed to find the expected strings when
  waiting for responses and thus always failed.

- Other changes:
* Project provides new get methods that return copies instead of the
  live maps of task and type definitions, references and targets.
* Ant is now more lenient with ZIP extra fields and will be able to
  read archives that it failed to read in earlier versions.
* The <zip> family of tasks has been sped up for bigger archives.
* Add removeKeepExtension option to NetRexxC task.
* Add prefix attribute to loadproperties task.
* Add resource attribute to length task.
* PropertyResource will effectively proxy another Resource if ${name}
  evaluates to a Resource object.
* Added forcestring attribute to equals condition to force evaluation
  of Object args as strings; previously only API-level usage of the
  equals condition allowed Object args, but Ant 1.8.x+ property
  evaluation may yield values of any type.
* BuildFileTest.assertPropertyUnset() fails with a slightly more
  meaningful error message
* <junit> will now throw an exception if a test name is empty.  This
  used to manifest itself in unrelated errors like
* A change that made <exec> more reliable on Windows (Bugzilla Report
  5003) strongly impacts the performance for commands that execute
  quickly, like attrib.  Basically no single execution of a command
  could take less than a second on Windows.
  A few timeouts have been tweaked to allow these commands to finish
  more quickly but still they will take longer than they did with Ant
  1.7.1.
* Added SimpleBigProjectLogger, intermediate between NoBannerLogger and
  BigProjectLogger.
* <mappedresources> supports new attributes enablemultiplemappings
  and cache.
* Added the augment task to manipulate existing references via Ant's basic
  introspection mechanisms.

(tonio)

Updated graphics/opencv to 2.1.0

(wiz)

Update to 2.1. Changelog of most insteresting changes:

2.1

  (April, 2010)

  General Modifications
      - The whole OpenCV is now using exceptions instead of the old
      libc-style mechanism.
          * That is, instead of checking error code with cvGetErrStatus()
            (which currently always returns 0) you can now just call OpenCV
            functions inside C++ try-catch statements, cv::Exception is now
            derived from std::exception.
      - All the parallel loops in OpenCV have been converted from OpenMP
          * to Intel TBB (http://www.threadingbuildingblocks.org/). Thus
            parallel version of OpenCV can now be built using MSVC 2008
            Express Edition or using earlier than 4.2 versions of GCC.

      - SWIG-based Python wrappers are still included,
          * but they are not built by default and it's generally preferable
            to use the new wrappers. The python samples have been rewritten
            by James Bowman to use the new-style Python wrappers, which have
            been also created by James.
      - OpenCV can now be built and run in 64-bit mode on MacOSX 10.6 and
      Windows (see HighGUI and known problems below).
          * On Windows both MSVC 2008 and mingw64 are known to work.
      - In theory OpenCV is now able to determine the host CPU on-fly and
      make use of SSE/SSE2/... instructions,
          * if they are available. That is, it should be more safe to use
            WITH_SSE* flags in CMake. However, if you want maximum
            portability, it's recommended to turn on just WITH_SSE and
            WITH_SSE2 and leave other SSE* turned off, as we found that using
            WITH_SSE3, WITH_SSSE3 and WITH_SSE4_1 can yield the code
            incompatible with Intel's pre-Penryn or AMD chips.
      - Experimental "static" OpenCV configuration in CMake was contributed
      by Jose Luis Blanco.
          * Pass "BUILD_SHARED_LIBS=OFF" to CMake to build OpenCV statically.

  New Functionality, Features
    * - cxcore, cv, cvaux:
          * Grabcut (http://en.wikipedia.org/wiki/GrabCut) image segmentation
            algorithm has been implemented.
              * See opencv/samples/c/grabcut.cpp
          * new improved version of one-way descriptor is added. See
            opencv/samples/c/one_way_sample.cpp
          * modified version of H. Hirschmuller semi-global stereo matching
            algorithm that we call SGBM
              * (semi-global block matching) has been created. It is much
                faster than Kolmogorov's graph cuts-based algorithm and yet
                it's usually better than the block matching StereoBM
                algorithm. See opencv/samples/c/stereo_matching.cpp.
          * existing StereoBM stereo correspondence algorithm by K. Konolige
            was noticeably improved:
              * added the optional left-right consistency check and speckle
                filtering, improved performance (by ~20%).
          * User can now control the image areas visible after the stereo
            rectification
              * (see the extended stereoRectify/cvStereoRectify), and also
                limit the region where the disparity is computed (see
                CvStereoBMState::roi1, roi2; getValidDisparityROI).
          * Mixture-of-Gaussian based background subtraction algorithm has
            been rewritten for better performance
              * and better accuracy. Alternative C++ interface
                BackgroundSubtractor has been provided, along with the
                possibility to use the trained background model to segment
                the foreground without updating the model. See
                opencv/samples/c/bgfg_segm.cpp.

      - highgui:
          * MacOSX: OpenCV now includes Cocoa and QTKit backends, in addition
            to Carbon and Quicktime.
              * Therefore you can build OpenCV as 64-bit library. Thanks to
                Andre Cohen and Nicolas Butko, which components Note however
                that the backend are now in the alpha state, they can crash
                or leak memory, so for anything more serious than quick
                experiments you may prefer to use Carbon and Quicktime. To
                do that, pass USE_CARBON=ON and USE_QUICKTIME=ON to CMake
                and build OpenCV in 32-bit mode (i.e. select i386
                architecture in Xcode).
          * Windows. OpenCV can now be built in 64-bit mode with MSVC 2008
            and also mingw64.
          * Fullscreen has been added (thanks to Yannick Verdie).
              * Call cvSetWindowProperty(window_name, CV_WINDOW_FULLSCREEN,
                1) to make the particular window to fill the whole screen.
                This feature is not supported in the Cocoa bindings yet.
          * gstreamer backend has been improved a lot (thanks to Stefano
            Fabri)

  Bug Fixes
    * - about 200 bugs have been fixed

2.0

  (September, 2009)

  New functionality, features:
    * - General:
          * New Python interface officially in.
      - MLL:
          * The new-style class aliases (e.g. cv::SVM ~ CvSVM) and the
            train/predict methods, taking cv::Mat in addition to CvMat, have
            been added. So now MLL can be used more seamlesly with the rest
            of the restyled OpenCV.

2.0 beta

  (September, 2009)

  New functionality, features:
    * General:
          * The brand-new C++ interface for most of OpenCV functionality
            (cxcore, cv, highgui) has been introduced. Generally it means
            that you will need to do less coding to achieve the same results;
            it brings automatic memory management and many other advantages.
              * See the C++ Reference section in opencv/doc/opencv.pdf and
                opencv/include/opencv/*.hpp.
                    * The previous interface is retained and still supported.
          * The source directory structure has been reorganized; now all the
            external headers are placed in the single directory on all
            platforms.
          * The primary build system is CMake,

    * CXCORE, CV, CVAUX:
          * CXCORE now uses Lapack (CLapack 3.1.1.1 in OpenCV 2.0) in its
            various linear algebra functions (such as solve, invert, SVD,
            determinant, eigen etc.) and the corresponding old-style
            functions (cvSolve, cvInvert etc.
          * Lots of new feature and object detectors and descriptors have
            been added (there is no documentation on them yet), see cv.hpp
            and cvaux.hpp:
              * FAST - the fast corner detector, submitted by Edward Rosten
              * MSER - maximally stable extremal regions, submitted by Liu
                Liu
              * LDetector - fast circle-based feature detector
                    * by V. Lepetit (a.k.a. YAPE)

              * Fern-based point classifier and the planar object detector -
                    * based on the works by M. Ozuysal and V. Lepetit

              * One-way descriptor - a powerful PCA-based feature
                descriptor,
                    * S. Hinterstoisser, O. Kutter, N. Navab, P. Fua, and V.
                      Lepetit, "Real-Time Learning of Accurate Patch
                      Rectification". Contributed by Victor Eruhimov

              * Spin Images 3D feature descriptor
                    * based on the A. Johnson PhD thesis; implemented by
                      Anatoly Baksheev

              * Self-similarity features - contributed by Rainer Leinhar
              * HOG people and object detector - the reimplementation of
                Navneet Dalal framework
                    * (http://pascal.inrialpes.fr/soft/olt/). Currently, only
                      the detection part is ported, but it is fully
                      compatible with the original training code.
                    * See cvaux.hpp and opencv/samples/c/peopledetect.cpp.
              * LBP (Local Binary Pattern) features
              * Extended variant of the Haar feature-based object detector -
                implemented by Maria Dimashova. It now supports Haar
                features and LBPs, other features can be added in the same
                way.
              * Adaptive skin detector and the fuzzy meanshift tracker -
                contributed by Farhad Dadgostar, see cvaux.hpp and
                opencv/samples/c/adaptiveskindetector.cpp

          * The new traincascade application complementing the new-style
            HAAR+LBP object detector has been added. See
            opencv/apps/traincascade.
          * The powerful library for approximate nearest neighbor search
            FLANN by Marius Muja is now shipped with OpenCV, and the
            OpenCV-style interface to the library is included into cxcore.
            See cxcore.hpp and opencv/samples/c/find_obj.cpp
          * The bundle adjustment engine has been contributed by PhaseSpace;
            see cvaux.hp
          * Added dense optical flow estimation function based on the paper
              * "Two-Frame Motion Estimation Based on Polynomial Expansion"
                by G. Farnerback.
              * See cv::calcOpticalFlowFarneback and the C++ documentation
          * Image warping operations (resize, remap, warpAffine,
            warpPerspective) now all support bicubic and Lanczos
            interpolation.
          * Most of the new linear and non-linear filtering operations
            (filter2D, sepFilter2D, erode, dilate ...) support arbitrary
            border modes and can use the valid image pixels outside of the
            ROI (i.e. the ROIs are not "isolated" anymore), see the C++
            documentation.
          * The data can now be saved to and loaded from GZIP-compressed
            XML/YML files, e.g.: cvSave("a.xml.gz", my_huge_matrix);
    * MLL:
          * Added the Extremely Random Trees that train super-fast, comparing
            to Boosting or Random Trees (by Maria Dimashova).
          * The decision tree engine and based on it classes (Decision Tree
            itself, Boost, Random Trees) have been reworked and now:
              * they consume much less memory (up to 200% savings)
              * the training can be run in multiple threads (when OpenCV is
                built with OpenMP support)
              * the boosting classification on numerical variables is
                especially fast because of the specialized low-overhead
                branch.
          * mltest has been added. While far from being complete, it contains
            correctness tests for some of the MLL classes.
    * HighGUI:
          * [Linux] The support for stereo cameras (currently Videre only)
            has been added.
          * There is now uniform interface for capturing video from two-,
            three- ... n-head cameras.
          * Images can now be compressed to or decompressed from buffers in
            the memory, see the C++ HighGUI reference manual
    * Documentation:
          * The reference manual has been converted from HTML to LaTeX (by
            James Bowman and Caroline Pantofaru)

    * Samples, misc.:
          * Better eye detector has been contributed by Shiqi Yu, see
            opencv/data/haarcascades/*[lefteye|righteye]*.xml
          * sample LBP (Local Binary Pattern) cascade for the frontal face
            detection has been created by Maria Dimashova, see
            opencv/data/lbpcascades/lbpcascade_frontalface.xml
          * Several high-quality body parts and facial feature detectors have
            been
              * contributed by Modesto Castrillon-Santana,
              * see opencv/data/haarcascades/haarcascade_mcs*.xml

  Optimization:
    * Many of the basic functions and the image processing operations(like
      arithmetic operations, geometric image transformations, filtering
      etc.) have got SSE2 optimization, so they are several times faster.
    * The model of IPP support has been changed. Now IPP is supposed to be
      detected by CMake at the configuration stage and linked against
      OpenCV. (In the beta it is not implemented yet though).
    * PNG encoder performance improved by factor of 4 by tuning the
      parameters

1.1pre1

  (October, 2008)

  New functionality/features:
    * General:
          * Octave bindings have been added. See interfaces/swig/octave (for
            now, Linux only)
    * CXCORE, CV, CVAUX:
          * Speeded-up Robust Features (SURF), contributed by Liu Liu. see
            samples/c/find_obj.cpp and the documentation opencvref_cv.htm
          * Many improvements in camera calibration:
              * Added stereo camera calibration: cvStereoCalibrate,
                cvStereoRectify etc.
              * Single camera calibration now uses Levenberg-Marquardt
                method and supports extra flags to switch on/off
                optimization of individual camera parameters
              * The optional 3rd radial distortion parameter (k3*r^6) is now
                supported in every calibration-related function
          * 2 stereo correspondence algorithms:
              * very fast block matching method by Kurt Konolige (processes
                the Tsukuba stereo pair in <10ms on Core2Duo laptop)
              * slow but more accurate graph-cut based algorithm by
                Kolmogorov and Zabin
          * Better homography estimation algorithms (RANSAC and LMEDs)
          * new C++ template image classes contributed by Daniel Filip
            (Google inc.). see opencv/cxcore/include/cvwimage.h
          * Fast approximate nearest neighbor search (by Xavier Delacour)
          * Codebook method for background/foreground segmentation (by Gary
            Bradski)
          * Sort function (contributed by Shiqi Yu)
          * [OpenCV+IPP] Face Detection (cvHaarDetectObjects) now runs much
            faster (up to 2x faster) when using IPP 5.3 or higher.
          * Much faster (~4x faster) fixed-point variant of cvRemap has been
            added
    * MLL:
          * Python bindings for MLL have been added. There are no samples
            yet.
    * HighGUI:
          * [Windows, 32bit] Added support for videoInput library. Hence,
            cvcam is [almost] not needed anymore
          * [Windows, 32bit] FFMPEG can now be used for video
            decoding/encoding via ffopencv*.dll
          * [Linux] Added unicap support
          * Improved internal video capturing and video encoding APIs
    * Documentation:
          * OpenCV book has been published (sold separately :) see
            docs/index.htm)

    * New samples (opencv/samples):
          * Many Octave samples
          * find_obj.cpp (SURF), bgfg_codebook.cpp (Codebook BG/FG
            segmentation),
              * stereo_calib.cpp (Stereo calibration and stereo
                correspondence)

(wiz)

Correct a typo in a comment, no functional change.

(dsainty)

Revbump

(adam)

* Fix building on Mac OS X
* Fix building with Exiv2 0.21

(adam)

* Fix buildling on Mac OS X
* Fix buildiing with Exiv2 0.21

(adam)

Updated misc/py-anita to 1.18

(gson)

Update misc/py-anita to 1.18.  Changes from 1.17:

Deal with sysinst no longer asking for a password cipher.

Create the work directory in "test" mode so that it works without
running "install" first.

Fix incorrect description of the --workdir option in the man page, and
consistently use the term "work directory" throughout.

Reorganized the man page, adding a MODES section.

New mode "print-workdir".

Fix installation of historical versions of -current around
CVS date 2009.08.23.20.57.40.

(gson)

Fix building on Mac OS X

(adam)

Note update of these pacakges:

textproc/ruby-bluecloth 2.0.9
textproc/ruby-builder 3.0.0
textproc/ruby-coderay 0.9.6
textproc/ruby-haml 3.0.24
textproc/ruby-hpricot 0.8.3
textproc/ruby-stringex 1.2.0
textproc/ruby-treetop 1.4.9
textproc/ruby-xslt 0.9.8

(taca)

Update ruby-xslt pacakge to 0.9.8.

0.9.8 :
* Replace STR2CSTR by StringValuePtr

(taca)

Update ruby-treetop package to 1.4.9.

Changes are unknown.

(taca)

Update ruby-stringex package to 1.2.0.

Changes are unknown.

(taca)

Update ruby-hpricot package to 0.8.3

= 0.8.3
=== 3 November, 2010
* GH#8: Nil-check before downcasing attribute key
* GH#25: Proper ruby 1.9 encoding support
* GH#28. Use integers instead of ?? on 1.9, which is just a string.
* including noscript to ElementInclusions , so that hpricot wont fail
  when trying to parse a meta tag inside head section when noscript is
  present.
* latest changes from fast_xs mainline
* Fixes to get Hpricot running on Rubinius:
  * Use free, not XFREE
  * Remove RSTRUCT craziness, don't break Array#at

(taca)

Update ruby-haml package to 3.0.24.

3.0.24

Tagged on GitHub.

    * html2haml now properly generates Haml for silent script expressions
      nested within blocks.
    * IronRuby compatibility. This is sort of a hack: IronRuby reports its
      version as 1.9, but it doesn’t support the encoding APIs, so we treat
      it as 1.8 instead.

3.0.23

Tagged on GitHub.

    * Fix the error message for unloadable modules when running the
      executables under Ruby 1.9.2.
    * Fix an error when combining old-style and new-style attributes.

3.0.22

Tagged on GitHub.

    * Allow an empty line after case but before when.
    * Remove vendor/sass, which snuck into the gem by mistake and was causing
      trouble for Heroku users (thanks to Jacques Crocker).
    * Support the Rails 3.1 template handler API.

3.0.21

Tagged on GitHub.

    * Fix the permissions errors for good.

3.0.20

Tagged on GitHub.

    * Fix some permissions errors.

3.0.19

Tagged on GitHub.

    * Fix the :encoding option under Ruby 1.9.2.
    * Fix interpolated if statement when HTML escaping is enabled.
    * Allow the --unix-newlines flag to work on Unix, where it’s a no-op.

(taca)