Update to 0.05:

0.05    29 July 2013
        - add option to avoid automatic deployment (Phil Hallows)

(wiz)

OpenBSD renamed beagle to armv7.
>From patrick@openbsd via jmc@openbsd.

(wiz)

Updated meta-pkgs/modular-xorg-drivers to 1.14.1

(wiz)

On NetBSD, depend on xf86-video-ati6 instead of xf86-video-ati.
Requested by Jan Danielsson <jan.m.danielsson@gmail.com>.
Bump version.

(wiz)

Updated meta-pkgs/modular-xorg-protos to 1.4

(wiz)

Updated meta-pkgs/modular-xorg-apps to 1.14.1.2

(wiz)

Add a few, bump version.

(wiz)

Add xtrans and font-util, bump version.

(wiz)

Updated devel/mercurial to 2.7.1

(wiz)

Update to 2.7.1. WhatsNew page on wiki is overrun by spambots, so
no list of changes available.

(wiz)

+ cdialog-1.2.20130902, libgee0.8-0.8.8, libmicrohttpd-0.9.30.

(wiz)

Drop maintainership.

(wiz)

Updated print/poppler to 0.24.1

(wiz)

Update to 0.24.1. No shlib version change.

Release 0.24.1
        core:
        * SplashOutputDev: use getRGBLine images if available. Bug #66928
        * SplashOutputDev: Don't copy bitmap if we don't need to.
        * PSOutputDev: Fix regression in -eps -level1sep rendering. Bug #68321
        * Fix crash in malformed file 1026.asan.0.42.pdf
        * use copyString instead of strdup where memory is freed with gfree. Bug #67666

        utils:
        * pdfdetach: don't mention xpdfrc
        * pdftotext: Fix -bbox with stdin as input. Bug #45163
        * pdftohtml: Fix jpeg image export. Bug #48270
        * pdfimages: Fix typos in man page

        glib:
        * demo: Remove GTK_DISABLE_DEPRECATED compilation flag

        qt4:
        * Fix small typo in documentation

        qt5:
        * Fix small typo in documentation

(wiz)

Updated net/gupnp to 0.20.4

(wiz)

Update to 0.20.4:

0.20.4
======

Changes since 0.20.3:
- Fix warnings in gtk-doc run.
- Some minor improvements to the documentation.
- Make the user-agent ASCII-only to fix issues with discovery when
  g_get_application_name () returned a translated string.
- Force HTTP version to be 1.1 in SOAP requests.
- Fix potential crashes in XML parsing.
- Fix potential memory leak in gupnp_service_proxy_send_action_valist.
- Some codestyle fixes.
- Add gupnp_device_info_list_dlna_device_class_identifier () to retrieve the
  content of <dlna:X_DLNADOC> nodes.

Bugs fixed in this release:
- https://bugzilla.gnome.org/show_bug.cgi?id=704953
- https://bugzilla.gnome.org/show_bug.cgi?id=702555
- https://bugzilla.gnome.org/show_bug.cgi?id=704094
- https://bugzilla.gnome.org/show_bug.cgi?id=700947

All contributors to this release:
  Ludovic Ferrandis <ludovic.ferrandis@intel.com>
  Jens Georg <mail@jensge.org>
  S辿bastien Bianti <sebastien.bianti@linux.intel.com>
  Bastien Nocera <hadess@hadess.net>

(wiz)

Updated net/gssdp to 0.14.4

(wiz)

Update to 0.14.4:

0.14.4
======

- Fix gtk-doc warnings.
- Use SO_REUSEPORT on FreeBSD and Debian/kFreeBSD.
- Fix discovery issue with some devices when g_get_application_name returned a
  non-ASCII translated version of the name.
- Fix a compiler warning about parentheses.
- Use GNOME_COMPILE_WARNINGS for a safe set of strict compiler warnings.
- Remove the use of INCLUDES which has been deprecated for a while.

All contributors to this release:
  Jens Georg <jensg@openismus.com>
  Bastien Nocera <hadess@hadess.net>

(wiz)

Updated textproc/raptor2 to 2.0.10

(wiz)

Update to 2.0.10:

New functions, types and enums

Functions

  ● int raptor_string_escaped_write(const unsigned char *string, size_t len, const char delim, unsigned int flags, raptor_iostream *iostr)

  ● int raptor_term_escaped_write(const raptor_term *term, unsigned int flags, raptor_iostream* iostr)

  ● int raptor_uri_escaped_write(raptor_uri* uri, raptor_uri* base_uri, unsigned int flags, raptor_iostream *iostr)

Types

  ● raptor_escaped_write_bitflags

(wiz)

Updated games/puzzles to 10035

(wiz)

Update to 10035, bug fixes.

(wiz)

Updated security/gnupg2 to 2.0.21

(wiz)

Update to 2.0.21:

Noteworthy changes in version 2.0.21 (2013-08-19)
-------------------------------------------------

* gpg-agent: By default the users are now asked via the Pinentry
  whether they trust an X.509 root key.  To prohibit interactive
  marking of such keys, the new option --no-allow-mark-trusted may
  be used.

* gpg-agent: The command KEYINFO has options to add info from
  sshcontrol.

* The included ssh agent does now support ECDSA keys.

* The new option --enable-putty-support allows gpg-agent to act on
  Windows as a Pageant replacement with full smartcard support.

* Support installation as portable application under Windows.

(wiz)

Updated fonts/dejavu-ttf to 2.34

(wiz)

Update to 2.34:

Changes from 2.33 to 2.34
* Sans, SansMono, Serif: unlinked references of U+2596 for bug 50848
  (by Denis Jacquerye)
* Sans, SansMono, Serif: added U+A7AA (by Denis Jacquerye)
* Sans, SansMono, Serif: added U+2A6A, U+2A6B, U+2E1F based on U+223B
  (by Denis Jacquerye)
* Sans, Serif: removed superfluous ligature definitions for ffl und ffi
  (bug 55363) (by Gee Fung Sit 薛至峰)
* Sans, Serif: swapped glyphs for U+25D2 and U+25D3 (bug 55197)
  (by Gee Fung Sit 薛至峰)
* Sans, Serif: added U+A740, U+A741 (by Denis Jacquerye)
* Sans: added U+20BA Turkish Lira sign (by Denis Jacquerye)
* Sans: replaced Georgian Asomtavruli U+10A0-U+10C5 and Mkhedruli U+10D0-U+10FC
  with new version (by Besarion Gugushvili)
* Sans: added Georgian Nuskhuri U+2D00-U+U+2D25 (by Besarion Gugushvili)
* Sans: added Private Use Area glyphs for Georgian U+F400-U+F441
  (by Besarion Gugushvili)
* Sans: tweaked U+0250, U+0254 (by Denis Jacquerye)
* Sans: adjusted hinting of U+032C-U+032D, avoiding problem on some platforms
  (by Denis Jacquerye)
* Sans: added U+A7A0-U+A7A9, pre-1921 Latvian letters with oblique stroke
  (by Denis Jacquerye)
* Sans: added anchors to U+2C6D (by Denis Jacquerye)
* Sans: added cedilla anchor to some Latin characters (by Denis Jacquerye)
* Sans: added ogonek anchor to A, E, O, U, Y (by Denis Jacquerye)
* Sans: adjusted ogonek reference in U+0172, U+01EA, U+01EB
  (by Denis Jacquerye)
* Sans: added anchors to U+0104, U+0105 (by Denis Jacquerye)
* Sans: added U+1F600, U+1F611, U+1F615, U+1F617, U+1F619, U+1F61B, U+1F61F,
  U+1F626-U+1F627, U+1F62E-U+1F62F, U+1F634 (by Gee Fung Sit 薛至峰)
* Sans: replaced U+27A1 with mirror image of U+2B05 for consistency
  (by Gee Fung Sit 薛至峰)
* Sans: copied hints from U+14A3, U+14A7 to U+2142-U+2143
  (by Gee Fung Sit 薛至峰)
* Sans: added Lisu block (by Gee Fung Sit 薛至峰)
* Sans: typographical improvements to U+0166-U+0167, U+02A6, U+02AA
  (by Gee Fung Sit 薛至峰)
* Sans: slightly change hinting of "2" to fix bug 37395 (by Ben Laenen)
* Sans: fixed U+1444 which had wrong top dot that shouldn't be there
  (by Denis Jacquerye)
* Sans: added anchors for diacritics to U+01B7, U+01B8, U+01B9, U+0292
  (by Denis Jacquerye)
* Sans: added U+01B7, U+01B8 to context for case diacritics above
* SansMono: fixed U+0574 (by Ruben Hakobian)
* SansMono: added U+2016, U+27C2 (by Yoshiki Ohshima)
* SansMono: added U+02CE, U+02CF (by Denis Jacquerye)
* SansMono: added U+2148, U+27E6-U+27E7, U+2B05-U+2B0D, U+1D55A
  (by Gee Fung Sit 薛至峰)
* Serif: added U+02BA, U+02C2-U+02C5, U+02CA-U+02CB, U+02D7, U+02F3, U+02F7,
  U+046C-U+046D, U+0476-U+0477, U+1D7C-U+1D7F, U+20B8, U+2132, U+214E, U+2C7B
  to Serif (by Gee Fung Sit 薛至峰)
* Serif: typographic improvements to U+0194, U+01B1, U+0263, U+028A, U+02A6,
  U+02A8, U+02AA, U+02E0, U+03DC, U+1D3B, U+1D7B (by Gee Fung Sit 薛至峰)
* Serif: added small cap versions of q, x (in italic styles), delta, theta, xi,
  sigma, phi, omega, not wired in yet (by Gee Fung Sit 薛至峰)
* Serif: added anchors to U+0234-U+0236 (by Gee Fung Sit 薛至峰)
* Serif: added U+02EC, U+02EF, U+02F0, U+0360 (by Denis Jacquerye)

(wiz)

Updated misc/py-carddav to 0.5.1

(wiz)

Update to 0.5.1:
This is a bugfix release, if pyCardDAV is working fine for you,
there is no reason to upgrade.

(wiz)

Updated devel/doxygen to 1.8.5

(wiz)

Update to 1.8.5:

Changes

Doxygen's source code is now managed using git and GitHub. Automatic builds
    and regression tests are scheduled via Travis CI.
Configuration data for the config file, the documentation, and the wizard
    are now produced from a single source (thanks to Albert)
All translation files have been migrated to UTF-8 (thanks to Petr Prikryl)
Added black box testing framework and a set of tests.

New features

Added SOURCE_TOOLTIPS option for advanced tooltip support while source browsing.

Lots of bug fixes.

(wiz)

Updated devel/py-setuptools to 1.1

(wiz)

Update to 1.1:

---
1.1
---

* Issue #71 (Distribute Issue #333): EasyInstall now puts less emphasis on the
  condition when a host is blocked via ``--allow-hosts``.
* Issue #72: Restored Python 2.4 compatibility in ``ez_setup.py``.

---
1.0
---

* Issue #60: On Windows, Setuptools supports deferring to another launcher,
  such as Vinay Sajip's `pylauncher <https://bitbucket.org/pypa/pylauncher>`_
  (included with Python 3.3) to launch console and GUI scripts and not install
  its own launcher executables. This experimental functionality is currently
  only enabled if  the ``SETUPTOOLS_LAUNCHER`` environment variable is set to
  "natural". In the future, this behavior may become default, but only after
  it has matured and seen substantial adoption. The ``SETUPTOOLS_LAUNCHER``
  also accepts "executable" to force the default behavior of creating launcher
  executables.
* Issue #63: Bootstrap script (ez_setup.py) now prefers Powershell, curl, or
  wget for retrieving the Setuptools tarball for improved security of the
  install. The script will still fall back to a simple ``urlopen`` on
  platforms that do not have these tools.
* Issue #65: Deprecated the ``Features`` functionality.
* Issue #52: In ``VerifyingHTTPSConn``, handle a tunnelled (proxied)
  connection.

Backward-Incompatible Changes
=============================

This release includes a couple of backward-incompatible changes, but most if
not all users will find 1.0 a drop-in replacement for 0.9.

* Issue #50: Normalized API of environment marker support. Specifically,
  removed line number and filename from SyntaxErrors when returned from
  `pkg_resources.invalid_marker`. Any clients depending on the specific
  string representation of exceptions returned by that function may need to
  be updated to account for this change.
* Issue #50: SyntaxErrors generated by `pkg_resources.invalid_marker` are
  normalized for cross-implementation consistency.
* Removed ``--ignore-conflicts-at-my-risk`` and ``--delete-conflicting``
  options to easy_install. These options have been deprecated since 0.6a11.

(wiz)

Updated devel/meld to 1.7.5

(wiz)

Update to 1.7.5:

2013-9-1 meld 1.7.5
===================

  Features:

    * Open the version control console view when the exit code of a VC
      operation indicates that there was an error (Kai Willadsen)
    * Improve our handling of bad gconf setups, and add a file-system key
      to force a no-gconf fallback for persistent issues (Daniel Richard G)
    * Add a preference for whether to highlight the current line of a file
      comparison (Kai Willadsen)
    * Keyboard shortcut for the commit dialog (Kai Willadsen)

  Fixes:

    * Fix traversing symlink loops in version control comparisons (Kai
      Willadsen)
    * Minor fixes and cleanups (Boruch Baum, Sandro Bonazzola, Kai Willadsen)

  Translations:

    * Antonio Fernandes C. Neto (pt_BR)
    * Daniel Mustieles (es)
    * Fran Diéguez (gl)
    * Marek Černocký (cs)
    * Rafael Ferreira (pt_BR)
    * Piotr Drąg (pl)

(wiz)

Updated devel/orc to 0.4.18

(wiz)

Update to 0.4.18.

Maintenance release:

- Important bugfix in reading constants from bytecode. (Tim-Philipp M端ller
  and Sebastian Dr旦ge)
- Documentation and code cleanup (Stefan Sauer)
- Fix cache flushing on iOS (Andoni Morales Alastruey)

(wiz)

Updated sysutils/desktop-file-utils to 0.22

(wiz)

Update to 0.22:

============
Version 0.22
============

desktop-file-validate
  - add DBusActivatable key to allowed keys (Matthias Clasen)
  - warn about OnlyShowIn in Action groups (Matthias Clasen)
  - allow validating multiple desktop files (Matthias Clasen)
  - improve help output (Matthias Clasen)
  - update man page (Matthias Clasen)

update-desktop-database
  - list only once a desktop file per mime type (Vincent)

misc
  - honor NOCONFIGURE=1 (Colin Walters)

(wiz)

+ Mesa-9.2, MesaLib-9.2, capc-calc-2.12.4.11, dar-2.4.11,
  dejavu-ttf-2.34, desktop-file-utils-0.22, etm2-2.1.10 [etm
  successor], fotoxx-13.09.1, giflib-5.0.5, gnumeric112-1.12.6,
  gnutls-3.2.4, horde-5.1.3, imp-6.1.4, kronolith-4.1.3, libgdata-0.14.0,
  libgee-0.10.4, libsecret-0.16, meld-1.7.5, nut-19.0, poppler-0.24.1,
  puzzles-10035, py-beets-1.2.2, py-qt5-5.1.1, py-setuptools-1.1,
  raptor2-2.0.10, redis-2.6.16, stella-3.9.2, subversion-base-1.8.3,
  texmaker-4.0.4, turba-4.1.2, wine-devel-1.7.1 [wait until NetBSD
  gains OSS 4 [kern/46611] or bring OSS 3 support back],
  x264-devel-20130831, xchm-1.23.

(wiz)

Fix build with bison-3.0.
Compile-tested only.

(wiz)

Added archivers/libunrar version 5.0.7

(wiz)

+ libunrar.

(wiz)

Import libunrar-5.0.7 as archivers/libunrar.

This package contains the unrar library.

unRAR is a utility to extract, test and view the contents of archives
created with the RAR archiver, version 1.50 and above.  The unRAR
utility is a minor part of the RAR archiver and contains RAR
uncompression algorithm. UnRAR requires very small volume of memory to
operate.

The license for this software states that "the code may not be used to
develop a RAR (WinRAR) compatible archiver."

(wiz)

Stop installing library again.
Split parts of Makefile into Makefile.common for upcoming libunrar.
Remove unnecessary patch (as discussed with upstream).

(wiz)

Updated archivers/unrar to 5.0.7nb1

(wiz)

Install unrar library and header file.
Bump PKGREVISION.

(wiz)

Rename xf86-video-ati v6 to xf86-video-ati6 and register conflicts.

(wiz)

+ tex-algorithm2e-doc

(wiz)

Fix grammar for bison-3.0.

(wiz)

MesaLib's configure insists on glproto>=1.4.11, so depend on that version.
Might help on NetBSD-5.2_STABLE.

(wiz)

+ calibre-1.0.0, doxygen-1.8.5, gentoo-0.20.3, git-1.8.4,
  libmicrohttpd-0.9.29, orc-0.4.18, parallel-20130822, py-kombu-2.5.14,
  redis-2.6.15, x264-devel-20130824, xfe-1.35.

(wiz)

Added sysutils/ncdu version 1.10

(wiz)

+ ncdu.

(wiz)

Import ncdu-1.10 as sysutils/ncdu, packaged for wip by slitvinov.

ncdu (NCurses Disk Usage) is a curses-based version of the well-known
'du', and provides a fast way to see what directories are using
your disk space.

(wiz)

Disable gethostbyname_r detection on NetBSD to make this build on 6.99.23.

(wiz)

+ dd_rescue-1.40, eric5-5.3.6, gnupg2-2.0.21, highlight-3.15,
  p5-Quota-1.7.0 [pkg/48144], py-qt5-5.0.1, py-sip-4.15, synergy-1.4.12
  [pkg/48141], tinc-1.0.22, vttest-20130818, webmin-1.650,
  wesnoth-1.10.6 [pkg/48139], xf86-video-intel-2.21.15.

(wiz)

Fix typo in Size.

(wiz)

Updated x11/xcb-util-cursor to 0.1.0

(wiz)

Update to 0.1.0:

Release 0.1.0  (2013-08-21)
=========================

- Bugfix: Use xhot/yhot, don窶冲 hardcode the value
- Bugfix: fix build on FreeBSD/NetBSD
- add COPYING file

(wiz)

Remove p5-Digest-SHA, included in perl.
Ride update.

(wiz)

Remove Digest-SHA dependency, this module comes with perl since 5.9.3,
as obache pointed out.
Ride update.

(wiz)

Updated textproc/mdoclint to 1.23

(wiz)

Recognize DragonFlyBSD. From Jan Stary <hans@stare.cz>.
Bump version.

(wiz)

Updated databases/pgbuildfarm to 4.11

(wiz)

Fix dependencies after update.

(wiz)

Update to 4.11, provided by Nicolas Thauvin <nico@orgrim.net> on
pkgsrc-users.

Changes since 4.10:

  * Turn down module cleanup verbosity
  * Add check for rogue postmasters.
  * Add pseudo-branch targets HEAD_PLUS_LATEST and HEAD_PLUS_LATEST2.
  * Use Digest::SHA instead of Digest::SHA1.
  * Make directory handling more robust in git code.
  * Move web transaction into a module procedure.
  * Switch to using the porcelain format of git status.
  * Provide parameter for core file patterns.
  * Use a command file for gdb instead of the -ex option

The web transaction and Digest::SHA changes have allowed the removal of
a couple of long-standing uglinesses on the system. In almost all
cases, the config parameter "aux_path" and the separate
run_web_transaction.pl script are now redundant (the exception is older

(wiz)

Updated mail/postgrey to 1.33nb9

(wiz)

Untaint --inet as well.
>From Yasuhiro KIMURA via github upstream bug report.
Bump PKGREVISION.

(wiz)

Fix typo.

(wiz)

Pick up maintainership.

(wiz)

Reset maintainer on his request.

(wiz)

+ denemo-1.0.6, mined-2013.23, nmap-6.40, pgbuildfarm-4.11,
  py-billiard-2.7.3.32, py-kombu-2.5.13, py-setuptools-1.0,
  serf-1.3.1, vim-7.4, x264-devel-20130817, xf86-video-ast-0.98.0.

(wiz)

Set LICENSE.

(wiz)

Fix build with bison-3.0.
(just build-tested)

(wiz)

Updated x11/xf86-input-mouse to 1.9.0nb1

(wiz)

Add all patches from NetBSD xsrc, as sent upstream just a minute ago.
Bump PKGREVISION.

(wiz)

Remove unneeded symbol that causes conflict with X headers.
PR 48128.

(wiz)

regen patch with better context

(wiz)

+ amarok-2.8, calibre-0.9.44, capc-calc-2.12.4.10, dd_rescue-1.39,
  eigen3-3.2, etm2-2.1.8 [etm successor], gettext-0.18.3.1,
  gramps-4.0.1, i3-4.6, kde4-4.11, libstatgrab-0.90, openbox-3.5.2,
  poppler-0.24.0, racket-5.3.6, rdesktop-1.8.0, rudiments-0.42,
  tig-1.2, xye-0.12.2.

(wiz)

Updated devel/scmgit-base to 1.8.3.4nb1

(wiz)

Remove perllocal.pod from PLIST. PR 48108.

Bump PKGREVISION.

(wiz)

+ xf86-video-neomagic-1.2.8.

(wiz)

+ unworkable-0.53 [pkg/48124].

(wiz)

Fix build with bison-3.0.

(wiz)

Updated pkgtools/pkglint to 4.132

(wiz)

Recognize apache24. Noted by Patrick Welche.
Bump version.

(wiz)

Updated x11/pixman to 0.30.2

(wiz)

Update to 0.30.2:

        Alexander Troosh (1):
              Require GTK+ version >= 2.16

        Andrea Canciani (1):
              test: Fix build on MSVC

        Markos Chandras (1):
              Use AC_LINK_IFELSE to check if the Loongson MMI code can link

        Matthieu Herrb (1):
              configure.ac: Don't use '+=' since it's not POSIX

        Siarhei Siamashka (1):
              test: fix matrix-test on big endian systems

        S淡ren Sandmann Pedersen (2):
              Version bump to 0.30.1
              Pre-release version bump to 0.30.2

        ingmar@irsoft.de (1):
              Fix broken build when HAVE_CONFIG_H is undefined, e.g. on Win32.

(wiz)

Updated devel/py-virtualenv to 1.10.1

(wiz)

Update to 1.10.1:

1.10.1 (2013-08-07)
~~~~~~~~~~~~~~~~~~~

* **New Signing Key** Release 1.10.1 is using a different key than normal with
  fingerprint: 7C6B 7C5D 5E2B 6356 A926 F04F 6E3C BCE9 3372 DCFA
* Upgraded pip to v1.4.1
* Upgraded setuptools to v0.9.8

(wiz)

Updated sysutils/cdrtools to 3.01alpha17

(wiz)

Update to 3.01a17:

All:

- The Schily Makefilesystem no longer uses CC= for internal tasks.
People who now set $CC to a different value will fail in a similar
way as they would fail with other build systems.

- Added automatic support for ARMv5 and ARMv6. This makes compilation
on the RaspberryPI also possible with the non-automake aware gmake.

- Allow "static" compilation (a compilation that does not use the
dynamic defines from the Schily Makefilesystem) on Linux ARMv6 (which
is used by RaspberryPI).

- Allow a "static" compilation (a compilation without using dynamic -I
Paths) on Linux on ARMv5 and ARMv6 by adding static #includes for
Linux ARM in:

include/schily//align.h
include/schily//archdefs.h
include/schily//avoffset.h
include/schily//xconfig.h

This is needed if you like to install the schily include files
to /usr/include/schily and allow users to "manually" compile
programs that use e.g libparanoia.

To allow this, the autoconf results need to be installed as:

<schily/armv5l-linux-gcc/*.h>
and
<schily/armv6l-linux-gcc/*.h>

- Many sources have been modified to deal __CYGWIN32__ like __CYGWIN__
This is neded as newer Cygwin versions that run in 64 Bit mode
do no longer define __CYGWIN32__ but __CYGWIN__

Libmdigest:

- Sources have been modified to deal __CYGWIN32__ like __CYGWIN__
This is neded as newer Cygwin versions that run in 64 Bit mode
do no longer define __CYGWIN32__ but __CYGWIN__

Libscg:

- Sources have been modified to deal __CYGWIN32__ like __CYGWIN__
This is neded as newer Cygwin versions that run in 64 Bit mode
do no longer define __CYGWIN32__ but __CYGWIN__

Cdrecord:

- Sources have been modified to deal __CYGWIN32__ like __CYGWIN__
This is neded as newer Cygwin versions that run in 64 Bit mode
do no longer define __CYGWIN32__ but __CYGWIN__

Cdda2wav (Maintained/enhanced by J�rg Schilling, originated by Heiko Eiﾟfeldt heiko@hexco.de):

- Sources have been modified to deal __CYGWIN32__ like __CYGWIN__
This is neded as newer Cygwin versions that run in 64 Bit mode
do no longer define __CYGWIN32__ but __CYGWIN__

(wiz)

+ cdrtools-3.01a17, digikam-3.3.0, libkface-3.3.0, libkgeomap-3.3.0,
  libkvkontakte-3.3.0, libmediawiki-3.3.0, mng-2.0.2, mono-3.2.1,
  perl5-5.18.1, py-jinja2-2.7.1, py-virtualenv-1.10.1, x264-devel-20130810.

(wiz)

Updated devel/at-spi2-core to 2.8.0nb1

(wiz)

Use MAKE_DIRS to create installation directory for default config file.
Bump PKGREVISION.

(wiz)

Updated editors/emacs24 to 24.3nb5

(wiz)

Add option for gtk3.
Rename "gtk" option to gtk2.
Change default from gtk to gtk3.

Correct a comment while here.

Bump PKGREVISION.

(wiz)

Apply patch from Onno van der Linden to try fixing PR 48116.

(wiz)

+ Sigil-0.7.3 [needs qt5], bzr-2.6.0, calibre-0.9.42, dd_rescue-1.38,
  eigen3-3.1.4, fotoxx-13.08, libvdpau-0.7, mawk-1.3.4.20130803,
  phpmyadmin-4.0.5, pixman-0.30.2, samba-4.0.8, wine-devel-1.7.0
  wordpress-3.6, x264-devel-20130803, xf86-video-ati-7.2.0,
  xf86-video-intel-2.21.14, xosview-1.15.

(wiz)

Updated x11/xwd to 1.0.6

(wiz)

Update to 1.0.6:

This minor maintenance release provides a collection of build configuration
improvements and janitorial cleanups.

Alan Coopersmith (10):
      Move Pixel typedef from xwd.c to wsutils.h
      Optionally enable existing code to call XkbStdBell from libxkbfile
      Fix gcc warnings about discarding const qualifiers
      Add printf attributes as suggested by gcc -Wmissing-format-attribute
      Rename new_list variable to not shadow new_list function
      Remove unused bitmap_unit variable from ReadRegionsInList()
      Quiet gcc warnings about potentially uninitialized variables
      Fix some clang warnings about integer size/sign conversions
      Strip trailing whitespace
      xwd 1.0.6

(wiz)

Updated x11/xset to 1.2.3

(wiz)

Update to 1.2.3:

This minor maintenance release adds a -version flag, as well as some
build configuration improvements and janitorial cleanups.

Alan Coopersmith (9):
      Drop usleep fallbacks for ancient OS versions
      Combine usage message strings
      Combine error message strings
      Fix gcc warnings about discarded const qualifiers
      Add printf attribute to usage function and fix warnings it causes
      Fix integer sign/size conversion warnings from clang
      Add -version flag
      Pass buf size to on_or_off instead of just assuming it is big enough
      xset 1.2.3

Jon TURNEY (1):
      Remove unneeded include of windows.h on WIN32

(wiz)

Updated x11/xprop to 1.2.2

(wiz)

Update to 1.2.2:

This minor maintenance release fixes the handling of the "u" type for
setting properties and does a bit of code & build configuration cleanup.

Alan Coopersmith (8):
      Expand description in README a little
      config: Add missing AC_CONFIG_SRCDIR
      Mark FatalError() as taking printf-style arguments
      Replace strlen/malloc/memcpy set with strdup
      Add unicode argument to Format_Len_String to reduce code duplication
      Fix const conversion warnings from gcc
      Mark usage() as noreturn, as suggested by gcc -Wmissing-noreturn
      xprop 1.2.2

lolilolicon (1):
      Enable setting property of type UTF8_STRING.

(wiz)

Reset PKGREVISION after update (clusterssh)

(wiz)

Reset PKGREVISION after update.

(wiz)

Remove FETCH_USING=curl after putting distfile on nbftp.

(wiz)

Remove FETCH_USING=curl after putting distfile on nbftp.

(wiz)

Remove more traces of unfinished -O support.

(wiz)

Simplify PKGNAME for older make(1)s or other parsers.

(wiz)

Remove -O from SYNOPSIS. Noted by jmc@OpenBSD.

(wiz)

Updated x11/xload to 1.1.2

(wiz)

Update to 1.1.2:

This release, like the previous one, brings more modernization and cleanup
of code for long dead OS variants.

Alan Coopersmith (18):
      Combine usage message strings
      unifdef -UCRAY
      unifdef -UAIXV3
      unifdef -UMOTOROLA
      unifdef -Uumips
      unifdef -Usony
      unifdef -Ualliant
      unifdef -UUTEK
      unifdef -Uhcx
      unifdef -Usequent
      unifdef -U__UNIXOS2__
      unifdef -UX_NOT_POSIX
      Make getloadavg() the first choice on Unix systems that have it
      Cleanup trailing whitespace
      Use 'imdent' to make the deeply nested #ifdef levels easier to follow
      Use C99 struct initializers for XKeyboardControl values
      Add noreturn attribute to usage() as suggested by -Wmissing-noreturn
      xload 1.1.2

Eric S. Raymond (1):
      Eliminate use of tab stops.  Helps with translation to DocBook.

(wiz)

Updated converters/fribidi to 0.19.5nb2

(wiz)

Do not install empty man pages (they confuse makemandb).
Bump PKGREVISION.

Reported upstream at https://bugs.freedesktop.org/show_bug.cgi?id=67714

(wiz)

Updated devel/mercurial to 2.7

(wiz)

Update to 2.7:

Mercurial 2.7 (2013-08-01)

Regularly scheduled feature release. This release contains an
important fix for a merge ancestor calculation regression in the
2.6 series.

    ancestor: decrement ninteresting correctly (issue3984)
    bookmarks: allow bookmark command to take multiple arguments
    bookmarks: pull --update updates to active bookmark if it moved (issue4007)
    bookmarks: update only proper bookmarks on push -r/-B (issue 3973)
    changegroup: fix fastpath during commit
    checklink: work around sshfs brain-damage (issue3636)
    churn: split email aliases from the right
    commands: add checks for unfinished operations (issue3955)
    commit: enable --secret option
    convert: catch empty origpaths in svn gettags (issue3941)
    convert: fix bad conversion of copies when hg.startrev is specified
    convert: handle changeset sorting errors without traceback (issue3961)
    hgweb: fix incorrect way to count revisions in log (issue3977)
    hgweb: run search instead of showing wrong error for ambigious identifier
    histedit: don't clobber working copy on --abort if not on histedit cset
    histedit: refuse to edit history that contains merges (issue3962)
    import: cut commit messages at --- unconditionally (issue2148)
    largefiles: overridematch() should replace the file path instead of extending (issue3934)
    log: add a log style that is default+phase (issue3436)
    paper: add line wrapping switch to file source view
    paper: code selection without line numbers in file source view
    paper: highlight line which is linked to in source view
    progress: respect HGPLAIN
    rebase: allow aborting when descendants detected
    rebase: continue abort without strip for immutable csets (issue3997)
    rebase: don't clobber wd on --abort when we've updated away (issue4009)
    revert: make backup when unforgetting a file (issue3423)
    revlog: handle hidden revs in _partialmatch (issue3979)
    rollback: mark as deprecated
    splicemap: improve error handling when parsing (issue2084)
    splicemap: support paths with spaces in splicemap (issue3844)
    sslutil: force SSLv3 on Python 2.6 and later (issue3905)
    summary: augment output with info from extensions
    templater: add strip function with chars as an extra argument
    log: show style list when unknown style specified
    tip: deprecate the tip command
    update: add tracking of interrupted updates (issue3113)
    worker: check problem state correctly (issue3982)
    worker: properly report errors from worker processes (issue3982)

(wiz)

Updated meta-pkgs/modular-xorg-apps to 1.14.1.1

(wiz)

+ xfwp, bump version.

(wiz)

Added x11/xfwp version 1.0.3

(wiz)

+ xfwp

(wiz)

Import xfwp-1.0.3 as x11/xfwp, packaged for wip by rodent@.

xfwp proxies X11 protocol connections, such as through a firewall.

(wiz)

Fix default options in man page. From jmc@openbsd.

(wiz)

Whitespace, from jmc@openbsd.

(wiz)

+ Mesa-9.1.6, MesaLib-9.1.6, dd_rescue-1.37, libgee-0.10.3,
  openttd-1.3.2.

(wiz)

Fix build with bison-3.0.
Addresses PR 48103 by Greg Oster.

(wiz)

Remove obsolete configure argument.

(wiz)

Fix build with bison-3.0.

(wiz)

Updated shells/zsh to 5.0.2nb2

(wiz)

Add .ts files to completion list for mplayer. Bump PKGREVISION.

(wiz)

+ gnutls-3.2.3.

(wiz)

+ texmaker-4.0.3.

(wiz)

Fix build with bison-3.0.

(wiz)

Updated games/puzzles to 9976

(wiz)

Update to 9976, bug fixes.

(wiz)

Updated devel/p5-Data-UUID to 1.219

(wiz)

Update to 1.219:

1.219    2013-07-06
  - cygwin fixes (thanks, Reini Urban!)
  - Skip t/threads.t unless perl version is 5.13.4 or greater (thanks, VPIT)
  - compile with strict C89 compilers (thanks, VPIT)
  - more bugfixes (thanks, VPIT)

(wiz)

Added security/PACK version 0.0.3

(wiz)

+ PACK.

(wiz)

Import PACK-0.0.3 as security/PACK.

PACK (Password Analysis and Cracking Toolkit) is a collection of
utilities developed to aid in analysis of password lists and
enhancing cracking of passwords using smart rule generation. It
can be used to reverse word mangling rules, generate source words
and optimize password masks for the Hashcat family of tools.

NOTE: The toolkit itself is not able to crack passwords, but instead
designed to make operation of password crackers more efficient.

(wiz)

+ Warn about using REPLACE_PYTHON without including application.mk.

(wiz)

Updated textproc/markdown-mode to 2.0

(wiz)

Update to 2.0:

Markdown Mode 2.0 Release Notes

Markdown Mode version 2.0 was released on March 24, 2013. This
is a major new stable release with many new features, including some
changes to keybindings for element insertion and outline navigation.
This new stable version can be downloaded from the
Markdown Mode homepage. Below is a description of significant or
user-visible enhancements or bug fixes. See the Git repository
for a complete history of changes.

New Features

Keybinding Changes

Physical styles: element insertion commands prefixed by C-c
C-p have been removed in favor of their logical style
counterparts prefixed by C-c C-s.
Shift is now the preferred way to distinguish keybindings for two
similar elements:

Insert an inline link with C-c C-a l or a reference
link with C-c C-a L. The latter keybinding is new
and preferred over C-c C-a r, which is deprecated.
Insert an inline image with C-c C-a l or a reference
image with C-c C-a L. The latter keybinding and
insertion functionality is new.
The new automatic header insertion function (more below) is
bound to C-c C-t h, with C-c C-t H being a
variant that indicates a preference for Setext headings.
Insert an ATX level-one heading with C-c C-t 1 or a
Setext level-one heading with C-c C-t ! (!
is S窶�1). The latter keybinding is new and is
preferred over C-c C-t t, which is deprecated.
Insert an ATX level-two heading with C-c C-t 2 or a
Setext level-two heading with C-c C-t @ (@
is S窶�2). The latter keybinding is new and is preferred
over C-c C-t s, which is deprecated.

Footnotes:

Insert new footnote: C-c C-f n has been changed to
C-c C-a f.
Jump to footnote text: C-c C-f g has been superceded
by C-c C-j (markdown-jump).
Jump to footnote marker: C-c C-f b has been
superceded by C-c C-j (markdown-jump).
Kill footnote: C-c C-f k has been superceded by
C-c C-k (markdown-kill-thing-at-point).

Several other new keybdings have been introduced and are described
in more detail below.

Smarter Markup Insertion

Fast heading insertion with a single command which automatically
calculates the type (atx or setext) and level (which can now be
cycled quickly, see below). Prefix with C-u to promote
the heading by one level or C-u C-u to demote the
heading by one level. Headings with a specific level or type
can still be inserted quickly with specific keybindings.
ATX heading insertion will use current line as heading text if not
blank and there is no active region.
Setext heading insertion will prompt for heading title when there is
no active region.
When the point is at a heading, the heading insertion commands will
replace the heading at point with a heading of the requested level
and type.
When there is no active region, the bold, italic, code, link, and
image insertion commands will operate on the word at point, if any,
so that you don窶冲 have to have an active selection for simple
modifications.
Repeating the bold, italic, or code insertion commands when the
point is at an element of the corresponding type will remove the
markup.
Smart list item insertion with M-RET, with indentation
and marker determined by the surrounding context. Prefix with
C-u to decrease the indentation by one level or
C-u C-u to increase the indentation one level.
Indentation of preformatted text and blockquotes will be adjusted
automatically in contexts where more indentation is required, as in
nested lists. (For example, in Markdown, a preformatted text block
inside a first-level list item must have eight spaces of
indentation.)
Improved reference link insertion with label completion.

Use word at point as link text, if possible, when there is no
active region.
Tab completion of reference labels from the set of currently
defined references.
Reference link insertion no longer prompts for
a URL or title if the label is already defined.
If no URL is given, create an empty reference definition
and move the point there.
Customizable reference link location via
markdown-reference-location.

Basic reference-style image markup insertion.
Multiple horizontal rule styles, customizable as a list of strings,
which can be cycled through.
New URL insertion command for inserting plain URLs delimited by
angle brackets (C-c C-a u). Works on URL at point, if any,
when there is no active region.
Generally improved insertion commands with respect to insertion of
surrounding whitespace and point position after insertion
(e.g., ensuring blank lines before and after newly inserted headings
and horizontal rules).

Markup Removal

Easily kill an element (e.g., a link or reference definition) at the
point and store the most important part in the kill ring (e.g., the
link text or URL).

Markup Completion

Completion (C-c C-]) normalizes the markup for an
element:

balancing hash marks and removing extra whitespace for atx
headings,
equating the length of the underline with the text for setext
headings, and
ensuring that horizintal rule strings are chosen from a
pre-defined list of strings (markdown-hr-strings).

Markup Promotion and Demotion

Markup promotion and demotion via C-c C窶� and C-c
C-=, respectively. The sequences M-<up> and
M-<down> may also be used.
Certain types of markup may be promoted or demoted: atx headings,
setext headings, horizontal rules, and list items.
Bold and italic also be toggled between asterisks and
underscores.
Promotion and demotion commands also cycle when level boundaries are
reached (e.g. when demoting a level six heading, it cycles back a
level one heading).

List and Region Editing

Move list items uper and more
accurate. The pre block matching routines have been completely
rewritten, with unit tests. The previous implementation was
incorrect in some cases and generally slow. This should be much
faster and more accurate for pre blocks inside nesterule for nested list items:
in a list item of level n, preformatted text must be indented at
least 4(n + 1) spaces.
Font lock for title strings in inline links.
Subtle syntax highlighting for hard line breaks.
More inclusive blockquote regular expressiofor more accurate syntax highlighting.

New and Improved Movement Commands

Jumping:

Quickly jump between reference definitions and reference-style links
and between footnote markers and footnote text with
C-c C-j.
Create undefined references when jumping from a reference link.
When jumping from a reference definition used by multiple
reference links, show a temporary buffer with buttons for
selecting which link to jump to. Possible locations may be
cycled with TAB and S-TAB.

Revised outline navigation commands, following org-mode. This
frees up the sexp navigation keys C-M-f and
C-M-b which can be useful in Markdown documents which
have many matching delimiters, as well as the defun navigation keys
C-M-a and C-M-e.
Previous/next section movement with C-M-a and
C-M-e (in Emacs parlance, this is movement by defun).
Mark the current section with C-M-h.
Previous/next paragraph movement via M-{ and M-}.
Previous/next block movement with C-u M-{ and C-u M-}.

Link Following and Movement

Unified link following: open links in a browser and wiki links in a
new buffer with the same keybinding (C-c C-o). This
supercedes the separate wiki link following command
(C-c C-w).
Generalized link movement and following: move between and open all
link types (inline, reference, wiki, angle URIs) using the same key
bindings (M-n and M-p). Previously, these
commands only moved between wiki links, but with the above following
enhancement, moving between hyperlinks of all types is more useful.
Remove wiki link following with RET and
markdown-follow-wiki-link-on-enter setting.

Other Enhancements

Respect hard line breaks when filling paragraphs.
A downside of this is that paragraph movement commands will also
stop at hard line breaks, thus slightly changing the definition of
窶徘aragraph窶� in this mode. In those cases, use the new block
movement commands instead.
Use adaptive filling for list items and blockquotes.

Treat all list items (any marker type) the same way with respect
to filling.
Retain the > prefix when filling blockquotes.
Fill list items inside of blockquotes.

Add indentation positions: preceding list markers and pre block position.
imenu support.
Added before and after export hooks.

markdown-before-export-hook can make changes to the source
buffer (e.g., preprocessing the Markdown source before running
Markdown). Any changes made will be kept until the file is
exported, so they will be incorporated into the resulting XHTML
output, and the buffer will be restored to it窶冱 original state
after exporting.
markdown-after-export-hook may modify the resulting output
buffer after export but before the output file is saved (e.g.,
postprocessing the XHTML output).

Added a library of regression tests which currently contains 1cks for older Emacsen and Xemacs.
Numerous other internal improvements to make the code base more robust.
Use button-map for navigating undefined references, so that
references can be navigated via the keyboard by pressing
TAB and S-TAB.
Clean up whites:

Dynamic loading and unloading for math support with
refontification.
Allow underscores and colons in equation labels.

GFM (GitHub Flavored Markdown) Mode Improvements

Changes italic font lock behavior to match GFM specification
regarding underscores in words.
Syntax highlighting for GFM quoted code blocks with an optional
language keyword.
Insertion command (C-c C-s P) for GFM quoted code blocks.

Bug Fixes

Fix bug in heading visibility cycling introduced in version 1.9
where the level of headings was being calculated incorrectly.
Fix problems inserting ATX headings at end of buffer.
Support small Setext headings (with fewer than three characters).
Several improvements to inline code syntax highlighting.

No need to prohibit spaces after the leading backquotes.
Markdown allows more than two leading backquotes (but the same number
must appear at the end to close the code tag).
Ensure following character is not a backquote.
Fix a bug where code fragments separated by a single character
were not highlighted correctly. (This is difficult or impossible
to do with regular expressions in Emacs, which do not support
negative lookahead or negative lookbehind, so a matching function
is used now.)

Fix some edge cases regarding escaping, spaces, etc. for bold and
italic font lock.
Prohibit newlines and tabs immediately after opening bold and italic
delimiters. This fixes a bug where italics would not be
highlighted following a horizontal rule
Improved multi-line font lock performance for large files.
Improved multi-line font lock at beginning of buffer.
List items with any of the three markers are filled in the same way
(previously list items starting with + were not filled with
hanging indentation).
Fix end detection for empty list items. Don窶冲 skip over the
whitespace following the marker on the same line. Previously, empty
list items were not being detected properly by
markdown-cur-list-item-bounds as a result of this.
Don窶冲 exclude [^], which is a valid reference tag (but let窶冱
please stick to alphanumeric characters).
No longer highlight escaped wiki links.
Fix line number buttons for reference checking and make all buttons
clickable.
Fix killing of footnotes with no text.
Fix escaping in markdown-xhtml-standalone-regexp.
Fix a font-lock edge case involving footnote markers preceding
inline links.
More accurate font-lock for ATX headings in edge cases.
Fix killing of footnotes from footnote text.

1.9

markdown-mode 1.9

markdown-mode version 1.9 was released on
January 25, 2013. This is a major new stable release with important
bug fixes.

New features:

Support for setext-style headers in outline-mode.
Thanks to Shigeru Fukaya.
Font lock for tilde-fenced code blocks.
Reference link insertion (C-c C-a r).
Support both forms of aliased wiki links:
[[link text|PageName]] and [[PageName|link text]]
(markdown-wiki-link-alias-first).
Footnote support: font lock and insertion, deletion, and
navigation functions for footnotes (prefixed by C-c C-f;
following for wiki links is now C-c C-w).
Thanks to Joost Kremers.
Improved preview and export commands.
Thanks to Donald Ephraim Curtis.
imenu support. Thanks to Akinori Musha.
Added autoload token for gfm-mode.
Thanks to Max Penet and Peter Eisentraut for the suggestion.
Optional character set declaration in XHTML output.
Thanks to Franﾃｧois Gannaz for the suggestion.
Smart unindentation when pressing delete at the beginning of a
line. Thanks to Zhenlei Jia.
Optional prefix argument to C-c C-w for opening wiki links
in another window.
Open inline and reference links and inline URIs in browser
(C-c C-o). Thanks to Peter Jones.
Open files in a standaline previewer or editor (C-c C-c o).
Clean up numbered/ordered lists (C-c C-c n).
Thanks to Donald Ephraim Curtis.
Save Markdown output to kill ring (copy to clipboard) (C-c C-c w).
Thanks to Donald Ephraim Curtis.

Bug fixes:

Fixed a bug which caused unusual behavior in functions performing
search, replace, and/or matching operations.
Thanks to Christopher J. Madsen for the patch.
Fixed a bug which caused an incompatibility with orgtbl-mode.
Thanks to Vegard Vesterheim for the report and to Carsten Dominik
for a patch.
Fixed a bug where reference links at the beginning of a line
would be mistaken for reference definitions.
Improved font lock for headers.
Improved font lock for reference definitions. Thanks to Ian Yang.
Avoid byte compiler warning about region-exists-p in GNU Emacs.
Additional key bindings for cross-platform header cycling
compatibility.
Fix problem with externally modified files on disk where the user
would get stuck in a loop answering 窶徨eally edit the buffer?窶�
Thanks to Bryan Fink for a detailed report.

Other improvements:

Improve markdown-mode-hook docstring.
Thanks to Shigeru Fukaya for the more precise description.
Don窶冲 require Common Lisp extensions at run time.
Thanks to Shigeru Fukaya.
Prefer visual-line-mode, the replacement for longlines-mode,
when in gfm-mode. Thanks to Christopher J. Madsen.
Proper GitHub wiki link handling in gfm-mode. Thanks to Kevin
Porter.
XEmacs compatibility:

Avoid malformed list errors during font definitions.
Handle replace-regexp-in-string.
Use text properties instead of overlays.
Fall back to set-buffer-modified-p when
restore-buffer-modified-p is unavailable.
Many additional fixes. Thanks to Michael Sperber.

Handle wiki links in buffers not associated with a file.
Update autoload documentation to support byte compilation.
Option to use - instead of _ for wiki links in gfm-mode.
Add two tab stops to possible indentation positions following list
items.

Minor bug fixes:

Font lock fix for URLs with underscores.
Escape shell commands to handle filenames with spaces, etc.
Thanks to Marcin Kasperski for a patch.
Use :slant instead of unsupported :italic in font spec.
Fix typo in paragraph-fill regexp.

(wiz)

Updated textproc/mdoclint to 1.22

(wiz)

Bump version.

(wiz)

Fix -F.

(wiz)

Remove unused function.

(wiz)

Remove mentions of -O, noone's implemented it yet or will soon.

(wiz)

Make Bq and Er optional for errors.

(wiz)

Fix -A support.

(wiz)

+ glpk-4.52.1, gssdp-0.14.4, gupnp-0.20.4, milter-manager-2.0.0,
  nut-18.10, vinagre-3.8.3.

(wiz)

regen

(wiz)

Fix typo

(wiz)

Updated x11/libX11 to 1.6.1

(wiz)

Update to 1.6.1:

This release includes fixes to the code, docs & build system.
It makes both clang & doclifter happier.  It includes patches
upstreamed from Ubuntu, NetBSD, and Solaris.

Alan Coopersmith (10):
      Update README to reflect where to find the Xlib specs now
      specs/libX11: correct prototype for XListPixmapFormats/XImageByteOrder
      troff macro expansion in specs/libX11
      Require ANSI C89 pre-processor, drop pre-C89 token pasting support
      XSetModifierMapping: Use Data instead of GetReqExtra
      Refactor common code from XAddHost & XRemoveHost into single function
      omGeneric: remove space between struct name & member name
      Add ku_TR.UTF-8 (Kurdish language, Turkey region) to compose/locale.dir
      Fix undefined XCMSDIR error when building lint library
      libX11 1.6.1

Eric S. Raymond (2):
      Remove call to undefined macro.
      Remove call to undefined macro.

Kees Cook (2):
      libX11: check size of GetReqExtra after XFlush
      libX11: check "req" when calling GetReqExtra

Thomas Klausner (6):
      Deal with the limited range of VAX floating point numbers when compiling for VAX.
      Stop truncating source to destination length if it is larger.
      Use newer callback-based API for XIM.
      Check for symbol existence with #ifdef, not #if
      Fix out-of-range comparison in _XF86BigfontQueryFont
      Tighten out-of-range comparisons.

(wiz)

Updated meta-pkgs/modular-xorg-protos to 1.3

(wiz)

Add three more proto packages.

(wiz)

Updated devel/lua-mode to 20130419

(wiz)

Update to 20130419:

* highlight hash-bang line as comment

* make lua-mode-hook editable via customize

* fix several indentation bugs & quirks

* fix lua-send-proc not to send previous function when point is at the beginning of a function

* derive lua-mode from prog-mode for Emacs24

* add font-locking for builtins and numeric constants

* fix a bug causing exponential complexity in a keyword matching regexp

* add more unindentation cases for block-closing tokens

* improve multiline highlighting via font-lock-syntactic-keywords
  This should make font-locking of multiline literals more fluent & stable. And
  it becomes customizable via standard font-lock configuration

* properly fontify variable definitions in 'local ...' & 'for ...'
  Also, perform some basic syntax verification in those lines. Multi-line
  constructs not supported yet.

* fix indentation for blocks starting on continued lines
    local foo =
        {
          bar,
          baz
        } ^
          1. these lines should be indented properly now
    ^
    2. the following lines should be unindented properly now

* extend imenu-generic-expression
  Now it matches 'foo = function(...)' function definitions

(wiz)

Updated finance/gnucash to 2.4.13nb1

(wiz)

Fix starting it with guile-www-2.37.
Bump PGKREVISION since it builds fine without the fix, but doesn't start.

(wiz)

Fix configury for guile-www-2.37+.

(wiz)

Fix build with python != 2.7

(wiz)

Updated wm/spectrwm to 2.3.0

(wiz)

Update to 2.3.0, based on a patch by Lokesh Mandvekar <lsm5@fedoraproject.org>
on pkgsrc-users.

Changes:
This release adds the ability to move/resize floating windows beyond
region boundaries. It adds 'soft boundary' behavior to region
boundaries. When moving a window past the region boundary, the
window will 'snap' to the region boundary if it is less than
boundary_width distance beyond the edge. A new boundary_width
configuration option has been added. The 'soft boundary' behavior
can be disabled by setting this option to 0. The ability to set
tile_gap to negative values has been added, which makes it possible
for tiled windows to overlap. Set this to the opposite of border_width
to collapse borders.

(wiz)

+ blender-2.68a, liferea-1.10.1, phpmyadmin-4.0.4.2.

(wiz)

Updated mail/akonadi to 1.10.2

(wiz)

Update to 1.10.2:

1.10.2                            23-July-2013
----------------------------------------------
- Fix PostgreSQL support (once more)

1.10.1   22-July-2013
----------------------------------------------
- Fix PostgreSQL support
- Optimize appending flags to items
- Introduce CHANGEDSINCE parameter to FETCH command

(wiz)

Updated devel/guile-www to 2.37

(wiz)

Update to 2.37:

- 2.37 | 2013-07-28

  - dropped

    - example program wwwcat

      This trivial program is not so interesting.

    - module ‘(www main)’
    - proc ‘(www server-utils parse-request) read-first-line’
    - proc ‘(www server-utils parse-request) read-headers’
    - proc ‘(www server-utils parse-request) skip-headers’
    - proc ‘(www server-utils parse-request) read-body’
    - proc ‘(www http) http:head’
    - proc ‘(www http) http:get’

      These were announced in Guile-WWW 2.34 (2012-03-29) NEWS.

    - support for values in ‘receive-response’ spec

      This was announced in Guile-WWW 2.36 (2012-11-22) NEWS.

  - ‘(www url) url:address’ more strict
  - ‘(www url) url:unknown’ more strict

    Contrary to the Guile-WWW 2.34 (2012-03-29) NEWS blurb, these
    two procs are not being deleted.  Instead, they persist and now
    check their arg for the proper scheme (‘mailto’ and ‘unknown’,
    respectively) to better support the (thin) abstraction they
    provide over the underlying data structure.

  - bootstrap tools upgraded
    - GNU Automake 1.13.4
    - Guile-BAUX 20130705.0751.4969fb4

(wiz)

Updated devel/py-virtualenv to 1.10

(wiz)

Update to 1.10:

1.10 (2013-07-23)
~~~~~~~~~~~~~~~~~

* **BACKWARDS INCOMPATIBLE** Dropped support for Python 2.5. The minimum
  supported Python version is now Python 2.6.

* **BACKWARDS INCOMPATIBLE** Using ``virtualenv.py`` as an isolated script
  (i.e. without an associated ``virtualenv_support`` directory) is no longer
  supported for security reasons and will fail with an error.

  Along with this, ``--never-download`` is now always pinned to ``True``, and
  is only being maintained in the short term for backward compatibility
  (Pull #412).

* **IMPORTANT** Switched to the new setuptools (v0.9.7) which has been merged
  with Distribute_ again and works for Python 2 and 3 with one codebase.
  The ``--distribute`` and ``--setuptools`` options are now no-op.

* Updated to pip 1.4.

* Added support for PyPy3k

* Added the option to use a version number with the ``-p`` option to get the
  system copy of that Python version (Windows only)

* Removed embedded ``ez_setup.py``, ``distribute_setup.py`` and
  ``distribute_from_egg.py`` files as part of switching to merged setuptools.

* Fixed ``--relocatable`` to work better on Windows.

* Fixed issue with readline on Windows.

.. _Distribute: https://pypi.python.org/pypi/distribute

(wiz)

Updated devel/py-setuptools to 0.9.8

(wiz)

Update to 0.9.8:

-----
0.9.8
-----

* Issue #53: Fix NameErrors in `_vcs_split_rev_from_url`.

-----
0.9.7
-----

* Issue #49: Correct AttributeError on PyPy where a hashlib.HASH object does
  not have a `.name` attribute.
* Issue #34: Documentation now refers to bootstrap script in code repository
  referenced by bookmark.
* Add underscore-separated keys to environment markers (markerlib).

(wiz)

Updated parallel/py-billiard to 2.7.3.31

(wiz)

Update to 2.7.3.31:

2.7.3.31 - 2012-07-04
---------------------

- Recent changes broke support for running without C extension (Issue #57).

(wiz)

Updated misc/py-anki2 to 2.0.12

(wiz)

Update to 2.0.12:

If you use LaTeX for mathematical equations on your cards, and
study with AnkiMobile or AnkiDroid, please revert to 2.0.8 (available
on the Anki website) until the mobile clients receive an update.

o Changed the way pastes and drag&drops are handled. This should
hopefully fix various issues people were having with Google Images,
and will now automatically add any images in a chunk of selected
text to your collection if you have HTML stripping turned off.

o Fixed errors caused by importing notes with a changed schema

o Revert a change made in 2.0.9 where Anki would switch the active
review deck when a different deck was selected in the add screen.
This caused issues where the note type would change invisibly and
some people objected to the new behaviour.

o Work around a crash when closing Anki on Windows caused by a toolkit bug.

o Make sure we always show 3 buttons for cards in initial learning in a filtered deck.

o Improve the checking for and stripping of filename characters
that can't be stored on some filesystems.

o Fixed an error when pasting media with non-latin characters in the URL.

o Don窶冲 throw an error if user adds cloze reference to a non-cloze note type.

o Fix error shown when importing csv file with blank first line.

o Fix year and all time review graph.

o If the user窶冱 recycle bin is broken, just delete the file instead.

o Don窶冲 send old backups to the recycle bin.

o Fix duplicates not being detected when pasting text.

o Fix a bug where the clipboard was being clobbered after the first paste.

o When pasting images between fields, don窶冲 try to download them.

o Allow includegraphics in LaTeX

o You can now view stats on filtered decks.

o Warn when prefs.db is corrupt, and catch another corruption case

o Allow searches for decks with + in their name.

o Work around linux crash on paste of invalid image.

o Fix an issue where duplicates were not being detected on paste

o Ensure starting ease can't be set below 130%.

o Make sure we include px in stats (thanks to steveaw)

o Fix error shown when importing from command line (thanks to Hans)

o Fix incorrect instructions to "maintenance" menu (thanks to Soren)

o Unused media now shows "no missing media" message in a tooltip
(thanks to Julien)

(wiz)

+ akonadi-1.10.2, guile-www-2.37, py-ZopeInterface-4.0.3,
  py-anki2-2.0.12, py-billiard-2.7.3.31, py-lxml-3.2.3,
  py-setuptools-0.9.8, py-virtualenv-1.10, qemu-1.5.2, redis-2.6.14,
  serf-1.3.0, subversion-base-1.8.1, x264-devel-20130727,
  xf86-video-intel-2.21.13.

(wiz)

Updated textproc/py-docutils to 0.11

(wiz)

Update to 0.11:

Release 0.11 (2013-07-22)
=========================

* General

  - Apply [ 2714873 ] Fix for the overwritting of document attributes.
  - Support embedded aliases within hyperlink references.
  - Fix [ 228 ] try local import of docutils components (reader, writer, parser,
    language module) before global search.

* docutils/parsers/rst/directives/tables.py

  - Fix [ 210 ] Python 3.3 checks CVS syntax only if "strict" is True.

* docutils/writers/html4css1/__init__.py
  - Fix [ 3600051 ] for tables in a list, table cells are not compacted.
  - New setting `stylesheet_dirs` (see above).

    Now, it is easy to add a custom stylesheet to Docutils' default
    stylesheet with, e.g., ``--stylesheet_path='html4css1.css, mystyle.css'``

    Changed behaviour of the default settings:
      if there is a file ``html4css1.css`` in the working directory of the
      process at launch, it is used instead of the one provided by Docutils
      in the writer source directory.

  - New default for math-output_: ``HTML math.css``.
  - Avoid repeated class declarations in html4css1 writer
    (modified version of patch [ 104 ]).

.. _math-output: docs/user/config.html#math-output

* docutils/writers/latex2e/__init__.py

  - Drop the simple algorithm replacing straight double quotes with
    English typographic ones.
    Activate the SmartQuotes_ transform if you want this feature.
  - New setting `stylesheet_dirs`: Comma-separated list of directories
    where stylesheets are found. Used by `stylesheet_path` when expanding
    relative path arguments.

.. _SmartQuotes: docs/user/config.html#smart-quotes

* docutils/writers/manpage.py

  - Fix [3607063] handle lines starting with a period.
  - Fix option separating comma was bold (thanks to Bill Morris).

(wiz)

Updated devel/bison to 3.0

(wiz)

Update to 3.0:

* Noteworthy changes in release 3.0 (2013-07-25) [stable]

** WARNING: Future backward-incompatibilities!

  Like other GNU packages, Bison will start using some of the C99 features
  for its own code, especially the definition of variables after statements.
  The generated C parsers still aim at C90.

** Backward incompatible changes

*** Obsolete features

  Support for YYFAIL is removed (deprecated in Bison 2.4.2): use YYERROR.

  Support for yystype and yyltype is removed (deprecated in Bison 1.875):
  use YYSTYPE and YYLTYPE.

  Support for YYLEX_PARAM and YYPARSE_PARAM is removed (deprecated in Bison
  1.875): use %lex-param, %parse-param, or %param.

  Missing semicolons at the end of actions are no longer added (as announced
  in the release 2.5).

*** Use of YACC='bison -y'

  TL;DR: With Autoconf <= 2.69, pass -Wno-yacc to (AM_)YFLAGS if you use
  Bison extensions.

  Traditional Yacc generates 'y.tab.c' whatever the name of the input file.
  Therefore Makefiles written for Yacc expect 'y.tab.c' (and possibly
  'y.tab.h' and 'y.outout') to be generated from 'foo.y'.

  To this end, for ages, AC_PROG_YACC, Autoconf's macro to look for an
  implementation of Yacc, was using Bison as 'bison -y'.  While it does
  ensure compatible output file names, it also enables warnings for
  incompatibilities with POSIX Yacc.  In other words, 'bison -y' triggers
  warnings for Bison extensions.

  Autoconf 2.70+ fixes this incompatibility by using YACC='bison -o y.tab.c'
  (which also generates 'y.tab.h' and 'y.output' when needed).
  Alternatively, disable Yacc warnings by passing '-Wno-yacc' to your Yacc
  flags (YFLAGS, or AM_YFLAGS with Automake).

** Bug fixes

*** The epilogue is no longer affected by internal #defines (glr.c)

  The glr.c skeleton uses defines such as #define yylval (yystackp->yyval) in
  generated code.  These weren't properly undefined before the inclusion of
  the user epilogue, so functions such as the following were butchered by the
  preprocessor expansion:

    int yylex (YYSTYPE *yylval);

  This is fixed: yylval, yynerrs, yychar, and yylloc are now valid
  identifiers for user-provided variables.

*** stdio.h is no longer needed when locations are enabled (yacc.c)

  Changes in Bison 2.7 introduced a dependency on FILE and fprintf when
  locations are enabled.  This is fixed.

*** Warnings about useless %pure-parser/%define api.pure are restored

** Diagnostics reported by Bison

  Most of these features were contributed by Th辿ophile Ranquet and Victor
  Santet.

*** Carets

  Version 2.7 introduced caret errors, for a prettier output.  These are now
  activated by default.  The old format can still be used by invoking Bison
  with -fno-caret (or -fnone).

  Some error messages that reproduced excerpts of the grammar are now using
  the caret information only.  For instance on:

    %%
    exp: 'a' | 'a';

  Bison 2.7 reports:

    in.y: warning: 1 reduce/reduce conflict [-Wconflicts-rr]
    in.y:2.12-14: warning: rule useless in parser due to conflicts: exp: 'a' [-Wother]

  Now bison reports:

    in.y: warning: 1 reduce/reduce conflict [-Wconflicts-rr]
    in.y:2.12-14: warning: rule useless in parser due to conflicts [-Wother]
    exp: 'a' | 'a';
                ^^^

  and "bison -fno-caret" reports:

    in.y: warning: 1 reduce/reduce conflict [-Wconflicts-rr]
    in.y:2.12-14: warning: rule useless in parser due to conflicts [-Wother]

*** Enhancements of the -Werror option

  The -Werror=CATEGORY option is now recognized, and will treat specified
  warnings as errors. The warnings need not have been explicitly activated
  using the -W option, this is similar to what GCC 4.7 does.

  For example, given the following command line, Bison will treat both
  warnings related to POSIX Yacc incompatibilities and S/R conflicts as
  errors (and only those):

    $ bison -Werror=yacc,error=conflicts-sr input.y

  If no categories are specified, -Werror will make all active warnings into
  errors. For example, the following line does the same the previous example:

    $ bison -Werror -Wnone -Wyacc -Wconflicts-sr input.y

  (By default -Wconflicts-sr,conflicts-rr,deprecated,other is enabled.)

  Note that the categories in this -Werror option may not be prefixed with
  "no-". However, -Wno-error[=CATEGORY] is valid.

  Note that -y enables -Werror=yacc. Therefore it is now possible to require
  Yacc-like behavior (e.g., always generate y.tab.c), but to report
  incompatibilities as warnings: "-y -Wno-error=yacc".

*** The display of warnings is now richer

  The option that controls a given warning is now displayed:

    foo.y:4.6: warning: type clash on default action: <foo> != <bar> [-Wother]

  In the case of warnings treated as errors, the prefix is changed from
  "warning: " to "error: ", and the suffix is displayed, in a manner similar
  to GCC, as [-Werror=CATEGORY].

  For instance, where the previous version of Bison would report (and exit
  with failure):

    bison: warnings being treated as errors
    input.y:1.1: warning: stray ',' treated as white space

  it now reports:

    input.y:1.1: error: stray ',' treated as white space [-Werror=other]

*** Deprecated constructs

  The new 'deprecated' warning category flags obsolete constructs whose
  support will be discontinued.  It is enabled by default.  These warnings
  used to be reported as 'other' warnings.

*** Useless semantic types

  Bison now warns about useless (uninhabited) semantic types.  Since
  semantic types are not declared to Bison (they are defined in the opaque
  %union structure), it is %printer/%destructor directives about useless
  types that trigger the warning:

    %token <type1> term
    %type  <type2> nterm
    %printer    {} <type1> <type3>
    %destructor {} <type2> <type4>
    %%
    nterm: term { $$ = $1; };

    3.28-34: warning: type <type3> is used, but is not associated to any symbol
    4.28-34: warning: type <type4> is used, but is not associated to any symbol

*** Undefined but unused symbols

  Bison used to raise an error for undefined symbols that are not used in
  the grammar.  This is now only a warning.

    %printer    {} symbol1
    %destructor {} symbol2
    %type <type>  symbol3
    %%
    exp: "a";

*** Useless destructors or printers

  Bison now warns about useless destructors or printers.  In the following
  example, the printer for <type1>, and the destructor for <type2> are
  useless: all symbols of <type1> (token1) already have a printer, and all
  symbols of type <type2> (token2) already have a destructor.

    %token <type1> token1
          <type2> token2
          <type3> token3
          <type4> token4
    %printer    {} token1 <type1> <type3>
    %destructor {} token2 <type2> <type4>

*** Conflicts

  The warnings and error messages about shift/reduce and reduce/reduce
  conflicts have been normalized.  For instance on the following foo.y file:

    %glr-parser
    %%
    exp: exp '+' exp | '0' | '0';

  compare the previous version of bison:

    $ bison foo.y
    foo.y: conflicts: 1 shift/reduce, 2 reduce/reduce
    $ bison -Werror foo.y
    bison: warnings being treated as errors
    foo.y: conflicts: 1 shift/reduce, 2 reduce/reduce

  with the new behavior:

    $ bison foo.y
    foo.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
    foo.y: warning: 2 reduce/reduce conflicts [-Wconflicts-rr]
    $ bison -Werror foo.y
    foo.y: error: 1 shift/reduce conflict [-Werror=conflicts-sr]
    foo.y: error: 2 reduce/reduce conflicts [-Werror=conflicts-rr]

  When %expect or %expect-rr is used, such as with bar.y:

    %expect 0
    %glr-parser
    %%
    exp: exp '+' exp | '0' | '0';

  Former behavior:

    $ bison bar.y
    bar.y: conflicts: 1 shift/reduce, 2 reduce/reduce
    bar.y: expected 0 shift/reduce conflicts
    bar.y: expected 0 reduce/reduce conflicts

  New one:

    $ bison bar.y
    bar.y: error: shift/reduce conflicts: 1 found, 0 expected
    bar.y: error: reduce/reduce conflicts: 2 found, 0 expected

** Incompatibilities with POSIX Yacc

  The 'yacc' category is no longer part of '-Wall', enable it explicitly
  with '-Wyacc'.

** Additional yylex/yyparse arguments

  The new directive %param declares additional arguments to both yylex and
  yyparse.  The %lex-param, %parse-param, and %param directives support one
  or more arguments.  Instead of

    %lex-param  {arg1_type *arg1}
    %lex-param  {arg2_type *arg2}
    %parse-param {arg1_type *arg1}
    %parse-param {arg2_type *arg2}

  one may now declare

    %param {arg1_type *arg1} {arg2_type *arg2}

** Types of values for %define variables

  Bison used to make no difference between '%define foo bar' and '%define
  foo "bar"'.  The former is now called a 'keyword value', and the latter a
  'string value'.  A third kind was added: 'code values', such as '%define
  foo {bar}'.

  Keyword variables are used for fixed value sets, e.g.,

    %define lr.type lalr

  Code variables are used for value in the target language, e.g.,

    %define api.value.type {struct semantic_type}

  String variables are used remaining cases, e.g. file names.

** Variable api.token.prefix

  The variable api.token.prefix changes the way tokens are identified in
  the generated files.  This is especially useful to avoid collisions
  with identifiers in the target language.  For instance

    %token FILE for ERROR
    %define api.token.prefix {TOK_}
    %%
    start: FILE for ERROR;

  will generate the definition of the symbols TOK_FILE, TOK_for, and
  TOK_ERROR in the generated sources.  In particular, the scanner must
  use these prefixed token names, although the grammar itself still
  uses the short names (as in the sample rule given above).

** Variable api.value.type

  This new %define variable supersedes the #define macro YYSTYPE.  The use
  of YYSTYPE is discouraged.  In particular, #defining YYSTYPE *and* either
  using %union or %defining api.value.type results in undefined behavior.

  Either define api.value.type, or use "%union":

    %union
    {
      int ival;
      char *sval;
    }
    %token <ival> INT "integer"
    %token <sval> STRING "string"
    %printer { fprintf (yyo, "%d", $$); } <ival>
    %destructor { free ($$); } <sval>

    /* In yylex().  */
    yylval.ival = 42; return INT;
    yylval.sval = "42"; return STRING;

  The %define variable api.value.type supports both keyword and code values.

  The keyword value 'union' means that the user provides genuine types, not
  union member names such as "ival" and "sval" above (WARNING: will fail if
  -y/--yacc/%yacc is enabled).

    %define api.value.type union
    %token <int> INT "integer"
    %token <char *> STRING "string"
    %printer { fprintf (yyo, "%d", $$); } <int>
    %destructor { free ($$); } <char *>

    /* In yylex().  */
    yylval.INT = 42; return INT;
    yylval.STRING = "42"; return STRING;

  The keyword value variant is somewhat equivalent, but for C++ special
  provision is made to allow classes to be used (more about this below).

    %define api.value.type variant
    %token <int> INT "integer"
    %token <std::string> STRING "string"

  Code values (in braces) denote user defined types.  This is where YYSTYPE
  used to be used.

    %code requires
    {
      struct my_value
      {
        enum
        {
          is_int, is_string
        } kind;
        union
        {
          int ival;
          char *sval;
        } u;
      };
    }
    %define api.value.type {struct my_value}
    %token <u.ival> INT "integer"
    %token <u.sval> STRING "string"
    %printer { fprintf (yyo, "%d", $$); } <u.ival>
    %destructor { free ($$); } <u.sval>

    /* In yylex().  */
    yylval.u.ival = 42; return INT;
    yylval.u.sval = "42"; return STRING;

** Variable parse.error

  This variable controls the verbosity of error messages.  The use of the
  %error-verbose directive is deprecated in favor of "%define parse.error
  verbose".

** Renamed %define variables

  The following variables have been renamed for consistency.  Backward
  compatibility is ensured, but upgrading is recommended.

    lr.default-reductions      -> lr.default-reduction
    lr.keep-unreachable-states -> lr.keep-unreachable-state
    namespace                  -> api.namespace
    stype                      -> api.value.type

** Semantic predicates

  Contributed by Paul Hilfinger.

  The new, experimental, semantic-predicate feature allows actions of the
  form "%?{ BOOLEAN-EXPRESSION }", which cause syntax errors (as for
  YYERROR) if the expression evaluates to 0, and are evaluated immediately
  in GLR parsers, rather than being deferred.  The result is that they allow
  the programmer to prune possible parses based on the values of run-time
  expressions.

** The directive %expect-rr is now an error in non GLR mode

  It used to be an error only if used in non GLR mode, _and_ if there are
  reduce/reduce conflicts.

** Tokens are numbered in their order of appearance

  Contributed by Valentin Tolmer.

  With '%token A B', A had a number less than the one of B.  However,
  precedence declarations used to generate a reversed order.  This is now
  fixed, and introducing tokens with any of %token, %left, %right,
  %precedence, or %nonassoc yields the same result.

  When mixing declarations of tokens with a litteral character (e.g., 'a')
  or with an identifier (e.g., B) in a precedence declaration, Bison
  numbered the litteral characters first.  For example

    %right A B 'c' 'd'

  would lead to the tokens declared in this order: 'c' 'd' A B.  Again, the
  input order is now preserved.

  These changes were made so that one can remove useless precedence and
  associativity declarations (i.e., map %nonassoc, %left or %right to
  %precedence, or to %token) and get exactly the same output.

** Useless precedence and associativity

  Contributed by Valentin Tolmer.

  When developing and maintaining a grammar, useless associativity and
  precedence directives are common.  They can be a nuisance: new ambiguities
  arising are sometimes masked because their conflicts are resolved due to
  the extra precedence or associativity information.  Furthermore, it can
  hinder the comprehension of a new grammar: one will wonder about the role
  of a precedence, where in fact it is useless.  The following changes aim
  at detecting and reporting these extra directives.

*** Precedence warning category

  A new category of warning, -Wprecedence, was introduced. It flags the
  useless precedence and associativity directives.

*** Useless associativity

  Bison now warns about symbols with a declared associativity that is never
  used to resolve conflicts.  In that case, using %precedence is sufficient;
  the parsing tables will remain unchanged.  Solving these warnings may raise
  useless precedence warnings, as the symbols no longer have associativity.
  For example:

    %left '+'
    %left '*'
    %%
    exp:
      "number"
    | exp '+' "number"
    | exp '*' exp
    ;

  will produce a

    warning: useless associativity for '+', use %precedence [-Wprecedence]
    %left '+'
          ^^^

*** Useless precedence

  Bison now warns about symbols with a declared precedence and no declared
  associativity (i.e., declared with %precedence), and whose precedence is
  never used.  In that case, the symbol can be safely declared with %token
  instead, without modifying the parsing tables.  For example:

    %precedence '='
    %%
    exp: "var" '=' "number";

  will produce a

    warning: useless precedence for '=' [-Wprecedence]
    %precedence '='
                ^^^

*** Useless precedence and associativity

  In case of both useless precedence and associativity, the issue is flagged
  as follows:

    %nonassoc '='
    %%
    exp: "var" '=' "number";

  The warning is:

    warning: useless precedence and associativity for '=' [-Wprecedence]
    %nonassoc '='
              ^^^

** Empty rules

  With help from Joel E. Denny and Gabriel Rassoul.

  Empty rules (i.e., with an empty right-hand side) can now be explicitly
  marked by the new %empty directive.  Using %empty on a non-empty rule is
  an error.  The new -Wempty-rule warning reports empty rules without
  %empty.  On the following grammar:

    %%
    s: a b c;
    a: ;
    b: %empty;
    c: 'a' %empty;

  bison reports:

    3.4-5: warning: empty rule without %empty [-Wempty-rule]
    a: {}
        ^^
    5.8-13: error: %empty on non-empty rule
    c: 'a' %empty {};
            ^^^^^^

** Java skeleton improvements

  The constants for token names were moved to the Lexer interface.  Also, it
  is possible to add code to the parser's constructors using "%code init"
  and "%define init_throws".
  Contributed by Paolo Bonzini.

  The Java skeleton now supports push parsing.
  Contributed by Dennis Heimbigner.

** C++ skeletons improvements

*** The parser header is no longer mandatory (lalr1.cc, glr.cc)

  Using %defines is now optional.  Without it, the needed support classes
  are defined in the generated parser, instead of additional files (such as
  location.hh, position.hh and stack.hh).

*** Locations are no longer mandatory (lalr1.cc, glr.cc)

  Both lalr1.cc and glr.cc no longer require %location.

*** syntax_error exception (lalr1.cc)

  The C++ parser features a syntax_error exception, which can be
  thrown from the scanner or from user rules to raise syntax errors.
  This facilitates reporting errors caught in sub-functions (e.g.,
  rejecting too large integral literals from a conversion function
  used by the scanner, or rejecting invalid combinations from a
  factory invoked by the user actions).

*** %define api.value.type variant

  This is based on a submission from Michiel De Wilde.  With help
  from Th辿ophile Ranquet.

  In this mode, complex C++ objects can be used as semantic values.  For
  instance:

    %token <::std::string> TEXT;
    %token <int> NUMBER;
    %token SEMICOLON ";"
    %type <::std::string> item;
    %type <::std::list<std::string>> list;
    %%
    result:
      list  { std::cout << $1 << std::endl; }
    ;

    list:
      %empty        { /* Generates an empty string list. */ }
    | list item ";" { std::swap ($$, $1); $$.push_back ($2); }
    ;

    item:
      TEXT    { std::swap ($$, $1); }
    | NUMBER  { $$ = string_cast ($1); }
    ;

*** %define api.token.constructor

  When variants are enabled, Bison can generate functions to build the
  tokens.  This guarantees that the token type (e.g., NUMBER) is consistent
  with the semantic value (e.g., int):

    parser::symbol_type yylex ()
    {
      parser::location_type loc = ...;
      ...
      return parser::make_TEXT ("Hello, world!", loc);
      ...
      return parser::make_NUMBER (42, loc);
      ...
      return parser::make_SEMICOLON (loc);
      ...
    }

*** C++ locations

  There are operator- and operator-= for 'location'.  Negative line/column
  increments can no longer underflow the resulting value.

(wiz)

+ KeePass-2.23, SOGo-2.0.7, bison-3.0, dd_rescue-1.36, gaupol-0.24,
  libgdiplus-3.0, libreoffice-4.1, libreoffice4-bin-4.1, mksh-47,
  mono-3.2, mp3splt-2.6.0, mp3splt-gtk-0.9.0, parallel-20130722,
  pspp-0.8.0, py-docutils-0.11, wireshark-1.10.1.

(wiz)

Updated devel/meld to 1.7.4

(wiz)

Update to 1.7.4:

2013-7-28 meld 1.7.4
====================

  Features:

    * Commit dialog can now automatically wrap commit messages (Kai Willadsen)
    * Add Bazaar support for pushing changes, and improve sensitivity setting
      (Louis des Landes)
    * Support Subversion 1.8 (Yuri)
    * Type-ahead-find search now works in folder comparisons (Kai Willadsen)
    * Preference to swap pane order left = local (Kai Willadsen)

    * Arch, Codeville and RCS are no longer supported (Kai Willadsen)

  Fixes:

    * When opening new tabs from the command line, bring new tab to the front
      (Richard Simmons)
    * Current chunk highlight is now customisable using gtkrc (Louis des
      Landes)
    * Syncpoints now prevent chunks from being re-merged, syncpoints move with
      insertions, and disable automatic updating (Kai Willadsen)
    * Various version control updates and fixes (Louis des Landes, Kai
      Willadsen)
    * Sensitivity fixes (Kai Willadsen)
    * Translation string fixes (Marek Černocký)

  Translations:
    * Daniel Mustieles (es)
    * Dimitris Spingos (el)
    * Fran Diéguez (gl)
    * Marek Černocký (cs)
    * Matej Urbančič (sl)
    * Piotr Drąg (pl)

(wiz)

Updated meta-pkgs/modular-xorg-protos to 1.2

(wiz)

Add more xorg proto packages. Bump version.

(wiz)

spectrwm/scrotwm updates.

(wiz)

Remove scrotwm, replaced by spectrwm.

(wiz)

Added wm/spectrwm version 2.2.0

(wiz)

Import spectrwm-2.2.0 as wm/spectrwm.

Based on a package by Lokesh Mandvekar <lsm5@fedoraproject.org>
and wm/scrotwm.

Spectrwm is a small dynamic tiling window manager for X11. It tries
to stay out of the way so that valuable screen real estate can be
used for much more important stuff. It has sane defaults and does
not require one to learn a language to do any configuration. It
was written by hackers for hackers and it strives to be small,
compact and fast.

It was largely inspired by xmonad and dwm. Both are fine products
but suffer from things like: crazy-unportable-language-syndrome,
silly defaults, asymmetrical window layout, "how hard can it be?"
and good old NIH. Nevertheless dwm was a phenomenal resource and
many good ideas and code was borrowed from it. On the other hand
xmonad has great defaults, key bindings and xinerama support but
is crippled by not being written in C.

Spectrwm is a beautiful pearl! For it too, was created by grinding
irritation. Nothing is a bigger waste of time than moving windows
around until they are the right size-ish or having just about any
relevant key combination being eaten for some task one never needs.
The path of agony is too long to quote and in classical OpenBSD
fashion (put up, or hack up) a brand new window manager was whooped
up to serve no other purpose than to obey its masters. It was
written by Marco Peereboom & Ryan Thomas McBride and it is released
under the ISC license.

(wiz)

Updated security/libgcrypt to 1.5.3

(wiz)

Update to 1.5.3:

Noteworthy changes in version 1.5.3 (2013-07-25)
------------------------------------------------

* Mitigate the Yarom/Falkner flush+reload side-channel attack on
  RSA secret keys.  See <http://eprint.iacr.org/2013/448>.

(wiz)

Updated security/gnupg to 1.4.14

(wiz)

Update to 1.4.14:

Noteworthy changes in version 1.4.14 (2013-07-25)
-------------------------------------------------

    * Mitigate the Yarom/Falkner flush+reload side-channel attack on
      RSA secret keys.  See <http://eprint.iacr.org/2013/448>.

    * Fixed IDEA for big-endian CPUs

    * Improved the diagnostics for failed keyserver lockups.

    * Minor bug and portability fixes.

(wiz)

Update scrotwm status.

(wiz)

Set MAINTAINER per request on pkgsrc-users.

(wiz)

Fix checksum for patch-src_perl_tabbed (hi morr)

(wiz)