Updated devel/p5-IPC-Run3 to 0.045

(rhaen)

Updated to 0.045

ChangeLog:

0.045 2011-12-23
  skip a test that tickles an OpenBSD bug and hangs the process (thanks,
  BinGOs)

(rhaen)

Updated devel/p5-AnyEvent to 6.13

(rhaen)

Updated devel/p5-AnyEvent to 6.13

ChangeLog:
6.13 Thu Jan 12 07:27:01 CET 2012
- AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
          causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
- AnyEvent::Handle::push_read (line => did pass $1 directly, so
          regex-matching inside the callback would change the parameter, despite
          $1 being dynamically scoped per-block. this perl bug is now being
          worked around (testcase by Cindy Wang).

6.12 Mon Dec 12 13:21:10 CET 2011
- $! was clobbered by subsequent calls in tcp_connect, due to
          the postpone not saving/restoring it (analyzed by Richard Garnish).

6.11 Tue Nov 22 10:36:05 CET 2011
- Tk cannot create windows when tainted unless you set a title, so
          set a dummy title for AnyEvent's dummy mainwindow. How dummy.
- escape any nonprintable/nonascii characters when stringifying
          backtraces.
        - log the reason loading an interface module fails to load at level debug.

6.1  Tue Oct  4 19:44:30 CEST 2011
- INCOMPATIBLE CHANGE: the default log level is now "4" (error
          and above), and some messages inside AnyEvent have been elevated
          to higher log levels to print by default.
        - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
        - work around yet another signal race bug, newly introduced by POE. My
          how that module sucks.
        - add google public ipv6 dns as fallback servers, so AE::DNS can finally
          support ipv6-only operation even with defaults (also use ::1 instead
          of 127.0.0.1).
- give better guidelines on how to chose a logging level.
        - AnyEvent::Log can now cap (limit) log message priority by context.
        - AE::log will now save and restore $! and $@, so callers don't have to.
        - AE::Strict now checks that registered fds are still valid at regular
          intervals.
        - skip frequently called 'require' calls if the module appears loaded.
- increase signal latency test timer considerably for poor little osx,
          which can't be bothered with basic posix functionality such as signal
          delivery.
        - mention some debugging aids in the tutorial.
        - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
        - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
          of the context as title.
        - AnyEvent::Strict would not restore read-write access to $_ when a callback
          died - this is a user bug, but since AE::Strict exists to help dbeugging...
- detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
          to work with it - the author is well aware of the issues but refuses
          to fix them. This extreme step has been taken because the module actively
          attacks the goals of AnyEvent and makes it harder for module authors to use
          AnyEvent.

6.02  Fri Aug 26 20:08:31 CEST 2011
        - log_to_syslog now accepts facility strings.
- log_to_syslog would not allow facilities - "$facility|$level"
          *literally* means that string. sheesh.
        - the CPAN parser got confused and thought we provide Sys::Syslog, try
          to work around.
        - renamed AnyEvent::Impl::FLTK2 to FLTK.

6.01  Fri Aug 26 07:04:11 CEST 2011
        - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
          the TTL as fourth element - this affects ->resolve, ->request
          and ->dns_unpack indirectly. Convenience functions such as
          AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
          AnyEvent::Socket.
        - INCOMPATIBLE CHANGE: backend authors now should not implement
          one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
          methods.
        - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
          AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
        - DEPRECIATION: please change
          "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
          "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
          in your programs, for future compatibility.
        - *blush* idle watcher emulation was borked.
        - AnyEvent::Strict now checks the AE::xxx functions as well.
        - querying too long domain names would barf deep inside AnyEvent::DNS - since
          this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
          for too long domain names explicitly and passes an undef to the callback
          instead (reported by James Bromberger).
        - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
        - new functions: AnyEvent::postpone, AnyEvent::log.
        - new module: AnyEvent::Log, for simple logging needs.
        - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
        - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
          a debug shell.
        - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
          watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
        - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
          /etc/hosts can be found.
        - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
          configuration.
        - env variables can now be specified using AE_xyz instead of
          PERL_ANYEVENT_xyz, subject to some rules.
        - lots of new AnyEvent::Debug functionality.
        - AnyEvent::Strict now checks for modifications $_ by the callback
          (by setting it to readonly while executing the callback, if possible).
        - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
          Invest extra overhead to work around this peculiar behaviour.
        - split the pure perl backend into a loop module and an impl
          module (AnyEvent::Loop and AnyEvent::Impl::Perl).
        - some read types (regex, netstring, json, storable) could
          errornously keep an AE::Handle object alive on their own
          (reported by Mohammad Toossi).
        - AnyEvent::Socket::resolve_sockaddr and all functions using it now
          supports /etc/hosts overrides.
        - add more workarounds around child watcher bugs in IO::Async
          and (newly introduced) in POE.
        - use glib child watchers and try to work around its limitations,
          if possible (based on a patch by Kevin Ryde).
- add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
        - many read types in AnyEvent::Handle would malfunction during
          transitions to TLS because of a "delete $self->{rbuf}".
          (reported by Gerald Galster).
        - improve condvar blocking wait performance (probably only noticable
          with faster backends), also make it easier for event loops that
          do not support blocking waits.
        - PERL_ANYEVENT_MODEL now also allows full module names, although
          the usefulness of this feature is not yet clear.
        - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
        - add t/80_ssltext.t.
        - no longer use AUTOLOAD for the initial loading (for no very good reason).
        - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
        - do not run event loop detection multiple times when programs call their cached
          copies of AnyEvent::detect.
        - parse_hostport now accepts absolute paths als unix domain sockets.
        - suppress the idiotic warning inside POE by patching the POE::Kernel
          run flags instead of asking the user to jump through contortions.
        - add Task::Weaken as dependency (as recommends only at this time).
        - replace PF_UNSPEC by 0, for good old minix.

6.0 (not officially released, but was on CPAN for about 15 minutes :/)

(rhaen)

update to 4.0.8
changes:
-deal with non-utf8 encoded PPD files
-improved printer search
-minor build fixes and cleanup

(drochner)

update to 4.0.9
changes:
-bugfixes (command injection fix was integrated)
-minor feature extensions (color profile support, ghostscript speedup)

(drochner)

Updated devel/p5-EV to 4.10

(rhaen)

Updated devel/p5-EV to 4.10

ChangeLog:
4.10 Thu Jan 19 18:54:23 CET 2012
        - (libev) fix a race where the workaround against the epoll fork
          bugs caused signals to not be handled anymore.
        - (libev) correct backend_fudge for most backends, and implement
          a windows specific workaround to avoid looping because we call
          both select and Sleep, both with different time resolutions.
- e_new wasn't declared static (causing very minor .so bloat).
        - replace more old api names by new ones.

(rhaen)

Updated www/py-django to 1.3.1nb1

(obache)

distutils pkg, register egg-info.

Bump PKGREVISION.

(obache)

update to a recent snapshot
chances: new printer definitions

(drochner)

add patches from upstream to fix 2 security problems:
-data injection attack for certain protocols (CVE-2012-0036)
-SSL CBC IV vulnerability (OpenSSL related, CVE-2011-3389)
bump PKGREV

(drochner)

Apply patch 79d1d3311319f3390f540f547becaba9d957f84c
from qemu upstream:

Fill in word 64 of IDENTIFY data to indicate support for PIO modes 3 and 4.
This allows NetBSD guests to use UltraDMA modes instead of just PIO mode 0.

With this patch I can no longer reproduce PR 42455.
Bump package revision.

(cegger)

fix an indent.

(obache)

Updated databases/py-psycopg2 to 2.4.4

(obache)

Update py-psycopg2 to 2.4.4.
While here, let to register egg-info.

What's new in psycopg 2.4.4
---------------------------

  - 'register_composite()' also works with the types implicitly defined
    after a table row, not only with the ones created by 'CREATE TYPE'.
  - Values for the isolation level symbolic constants restored to what
    they were before release 2.4.2 to avoid breaking apps using the
    values instead of the constants.
  - Named DictCursor/RealDictCursor honour itersize (ticket #80).
  - Fixed rollback on error on Zope (ticket #73).
  - Raise 'DatabaseError' instead of 'Error' with empty libpq errors,
    consistently with other disconnection-related errors: regression
    introduced in release 2.4.1 (ticket #82).

What's new in psycopg 2.4.3
---------------------------

  - connect() supports all the keyword arguments supported by the
    database
  - Added 'new_array_type()' function for easy creation of array
    typecasters.
  - Added support for arrays of hstores and composite types (ticket #66).
  - Fixed segfault in case of transaction started with connection lost
    (and possibly other events).
  - Fixed adaptation of Decimal type in sub-interpreters, such as in
    certain mod_wsgi configurations (ticket #52).
  - Rollback connections in transaction or in error before putting them
    back into a pool. Also discard broken connections (ticket #62).
  - Lazy import of the slow uuid module, thanks to Marko Kreen.
  - Fixed NamedTupleCursor.executemany() (ticket #65).
  - Fixed --static-libpq setup option (ticket #64).
  - Fixed interaction between RealDictCursor and named cursors
    (ticket #67).
  - Dropped limit on the columns length in COPY operations (ticket #68).
  - Fixed reference leak with arguments referenced more than once
    in queries (ticket #81).
  - Fixed typecasting of arrays containing consecutive backslashes.
  - 'errorcodes' map updated to PostgreSQL 9.1.

(obache)

+ cc65-2.13.2, curl-7.24.0, dasm-2.20.11, etm-885, fuse-ntfs-3g-2012.1.15,
  gdb-7.4, gthumb-2.14.2, gtk2-2.24.9, kde4-4.8, lftp-4.3.5,
  libmicrohttpd-0.9.18, mimedefang-2.73, opendnssec-1.3.5,
  openmsx-0.8.2, opera-11.61, parallel-20120122, pure-ftpd-1.0.35
  [pkg/45867], puzzles-9388, remind-3.1.12, samba-3.6.2, snort-2.9.2.1,
  tmux-1.6, viewvc-1.1.13, webmin-1.580 [security].

(wiz)

Updated archivers/p5-Archive-Tar to 1.82

(rhaen)

Updated archivers/p5-Archive-Tar to 1.82

ChangeLog from upstream:
* important changes in version 1.82 21/11/2011 (CDRAKE)
- Adjustments to handle files >8gb (>0777777777777 octal)
- Feature to return the MD5SUM of files in the archive

(rhaen)

Updated www/py-webpy to 0.36nb1

(obache)

* no compiler languages is required to build.
* distutils pkg, register egg-info.

Bump PKGREVISION.

(obache)

Use SET_LIBDIR with packages that want to use to lib64

(sbd)

Add SET_LIBDIR variable which if non-empty will set GNU_CONFIGURE_LIBDIR to
the default libdir (and there for be passed to ./configure), this is the
best way to deal with the 'lib64' problem on x86_64-linux systems (with
most some packages).

Also add SET_LIBDIR, GNU_CONFIGURE_LIBDIR, and  GNU_CONFIGURE_LIBSUBDIR to
_PKG_VARS.gnu-configure

(sbd)

In the section of configure that decides whether a libdir suffix should be
used, force it to be an empty (i.e no suffix).

(sbd)

Updated net/zeromq to 2.1.11

(obache)

Update zeromq to 2.1.11.

0MQ version 2.1.11 (Stable), released on 2011/12/18
===================================================

Bug fixes
---------

* Fixed issue 290, zmq_poll was using system time instead of monotonic
  clock (Mika Fischer).

* Fixed issue 281, crash on heavy socket creation - assertion failure in
  mutex.hpp:91. (Mika Fischer).

* Fixed issue 273, O_CLOEXEC flag used in ip.cpp:192 is supported only
  on Linux kernels 2.6.27+

* Fixed issue 261, assertion failure in kqueue.cpp:76.

* Fixed issue 269, faulty diagnostic code in 2.1.10.

* Fixed issue 254, assertion failure at tcp_socket.cpp:229 on ENOTCONN.

Changes
-------

* Now builds properly on AIX 6.1 (AJ Lewis).

* Builds using libdcekt on HP-UX (AJ Lewis).

* New upstream OpenPGM maintenance release 5.1.118.

* Enabled debugging on assertion failure on Windows (Paul Betts).

(obache)

Remove the section from configure that adds the gcc multi_os_directory to
the libdir.

(sbd)

For proper DESTDIR support it should be passed to make as 'DESTDIR=${DESTDIR}'
not 'prefix=${DESTDIR}${PREFIX}'

(sbd)

Split ifdef with && into two, to avoid parse errors on the second
clause when the first is not satisfied.  (Apparently &&, while not
evaluating, still has to parse.)  Fixes build on OS X.

(gdt)

Per PR#45871 Solaris derived OSs haven't DEFFILEMODE defined
so define it if it is not defined.

No bump, since this only changes a case where the package did not build.

Also, for the record, this version does not include the code vulnerable
to http://secunia.com/advisories/44104/ as that feature is not yet here.

(tez)

MAKE_JOBS_SAFE=no

Non-repeatable build error reported by joerg@ - disable parallel
builds for now until this can be debugged.

(gdt)

Updated emulators/qemu to 0.15.1nb1

(jakllsch)

Enable dependency on curl for http://-backed block devices.
Additionally qemu "curses" support wants terminfo support.
pkgrev++.

(jakllsch)

Add note about use of unsafe_depends_loose, from discussion with
obache@.

Note that there is an issue with unsafe_depends not getting set when
options change.

(gdt)

Updated www/py-curl to 7.19.0

(obache)

Update py-curl to 7.19.0.
(change to register distutils egg-info)

Version 7.19.0 [requires libcurl-7.19.0 or better]
--------------

        * Added CURLFILE, ADDRESS_SCOPE and ISSUERCERT options,
          as well as the APPCONNECT_TIME info.

        * Added PRIMARY_IP info (patch by
          Yuhui H <eyecat at gmail.com>).

        * Added support for curl_easy_reset through a
          new 'reset' method on curl objects
          (patch by Nick Pilon <npilon at oreilly.com>).

        * Added support for OPENSOCKET callbacks.
          See 'tests/test_opensocket.py' for example
          usage (patch by Thomas Hunger <teh at camvine.com>).

Version 7.18.2
--------------

        * Added REDIRECT_URL info and M_MAXCONNECTS option
          (patch by Yuhui H <eyecat at gmail.com>).

        * Added socket_action() method to CurlMulti objects.
          See 'tests/test_multi_socket_select.py' for example
          usage (patch by Yuhui H <eyecat at gmail.com>).

        * Added AUTOREFERER option.

        * Allow resetting some list operations (HTTPHEADER,
          QUOTE, POSTQUOTE, PREQUOTE) by passing an empty
          list to setopt (patch by Jim Patterson).

Version 7.18.1
--------------

        * Added POST301, SSH_HOST_PUBLIC_KEY_MD5,
          COPYPOSTFIELDS and PROXY_TRANSFER_MODE options.

        * Check for static libs in setup.py to better detect
          whether libcurl was linked with OpenSSL or GNUTLS.

        * PycURL is now dual licensed under the LGPL and
          a license similar to the cURL license (an MIT/X
          derivative).

(obache)

Updated www/zopeedit to 0.7nb4

(obache)

* no compiler languages is required to build
* convert to use distutils.mk, register egg-info.

Bump PKGREVISION.

(obache)

no compiler languages is requred to build.

(obache)

Updated wm/py-tyle to 0.7.5nb1

(obache)

Convert to use python/distutils.mk, register egg-info.

Bump PKGREVISION.

(obache)

Updated wm/icecc to 2.9nb13

(obache)

* using vim version specific directory, so change dependency pattern on vim.
* no need to buildlink with py-qt3-base, simple DEPENDS is sufficient.
* using qt3-libs, so buildlink with it and qt3-tools.

Bump PKGREVISION.

(obache)

Update MASTER_SITES and HOMEPAGE, original site is dead.

(obache)

Updated net/powerdns-* to 2.9.22.5nb2.

(ghen)

Previous PLIST change was wrong due to a libdir bug now fixed by sbd.  Revert
PLIST and bump PKGREVISION again for powerdns + modules.

(ghen)

Added net/p5-Net-Frame version 1.07.
Added net/p5-Net-Frame-Dump version 1.09.
Added net/p5-Net-Frame-Simple version 1.04.

(hiramatsu)

Note update of net/yale-tftpd.

(he)

Update patches for this package so that it builds with -Werror and
-Wall with the gcc in NetBSD 5.1, i.e. gcc 4.1.3.  Fixes prompted
by reports that a build outside of pkgsrc for "64-bit Linux" (amd64)
but using this set of patches, gets a segfault, and this fixes that
problem.  Bump pkgrevision.

(he)

Added p5-Net-Frame, p5-Net-Frame-Dump and p5-Net-Frame-Simple.

(hiramatsu)

Import Net::Frame::Simple into net/p5-Net-Frame-Simple. This package is
provided in PR pkg/45720.

This module is part of Net::Frame frame crafting framework. It is totally
optional, but can make playing with the network far easier.

Basically, it hides the complexity of frame forging, sending, and receiving,
by providing helper methods, which will analyze internally how to assemble
frames and find responses to probes.

For example, it will take care of computing lengths and checksums, and matching
a response frame to the requesting frame.

Status:

Vendor Tag: TNF
Release Tags: pkgsrc-base

(hiramatsu)

Import Net::Frame::Dump into net/p5-Net-Frame-Dump. This package is
provided in PR pkg/45720.

Net::Frame::Dump is the base class for all dump modules. With them, you can
open a device for live capture, for offline analysis, or for creating a pcap
file.

See Net::Frame::Dump::Offline, Net::Frame::Dump::Online,
Net::Frame::Dump::Writer for specific usage.

Status:

Vendor Tag: TNF
Release Tags: pkgsrc-base

(hiramatsu)

Import Net::Frame into net/p5-Net-Frame. This package is provided in
PR pkg/45720.

Net::Frame is a fork of Net::Packet. The goal here was to greatly simplify
the use of the frame crafting framework. Net::Packet does many things
undercover, and it was difficult to document all the thingies.

Also, Net::Packet may suffer from unease of use, because frames were assembled
using layers stored in L2, L3, L4 and L7 attributes. Net::Frame removes all
this, and is splitted in different modules, for those who only want to use part
of the framework, and not whole framework.

Finally, anyone can create a layer, and put it on his CPAN space, because of
the modularity Net::Frame offers. For an example, see Net::Frame::Layer::ICMPv4
on my CPAN space.

Net::Frame does ship with basic layers, to start playing.

Status:

Vendor Tag: TNF
Release Tags: pkgsrc-base

(hiramatsu)

LDFLAGS setting only needed when gtk option is turned on, so move it there.
Tested by John Marino.

(wiz)

x11/xfce4-notes-plugin: Fix indirect linking error on DragonFly only

(marino)

comms/efax-gtk: Fix indirect linking error on DragonFly only

(marino)

benchmarks/iozone: Unbreak DragonFly (again)

The update from version 3.308 to 3.397 broke DragonFly 11 days after I
fixed it due to missing a macro change.  An integer redeclaration was
also removed.

(marino)

chat/telepathy-idle: Fix indirect linking error on DragonFly only

(marino)

cad/gwave: Fix indirect linking error on DragonFly only

(marino)

cad/gsmc: Fix indirect linking error on DragonFly only

(marino)

audio/padevchooser: Fix indirect linking error on DragonFly only

(marino)

audio/jack-rack: Fix indirect linking error on DragonFly only

(marino)

audio/gqmpeg-devel: Fix indirect linking error on DragonFly only

(marino)

audio/audacious: Fix indirect linking error on DragonFly only

(marino)

audio/alsa-utils: Fix indirect linking error on DragonFly only

(marino)

x11/xvidcap-gtk2: Fix indirect linking error on DragonFly only

(marino)

x11/xvidcap: Fix indirect linking error on DragonFly only

(marino)

Fix mdoc syntax. Bump revision.

(joerg)

Fix mdoc syntax. Bump revision.

(joerg)

Updated net/py-boto to 1.9bnb2

(gson)

Include patch to fix boto issue 310:
<http://code.google.com/p/boto/issues/detail?id=310>.

(gson)

Fix mdoc markup. Bump revision.

(joerg)

Fix mdoc markup. Bump revision.

(joerg)

Kill reundant .TP statements. Bump revision.

(joerg)

Kill empty section in man page. Bump revision.

(joerg)

Convert file creating patch into plain file. Fix man page markup.
Bump revision.

(joerg)

Fix .so usage in man page. Bump revision of effected package.

(joerg)

fix patch checksum

(hans)

Updated geography/gdal-lib to 1.8.1

(gdt)

Update to 1.8.1.

= GDAL/OGR 1.8.1 release notes =

The 1.8.1 release is a routine bug fix release on the stable GDAL/OGR 1.8
branch.

== General ==
* Fix for windows build of cpl_vsil_gzip.cpp (#3908)
* Fix OGRDataSource.Open() signature so utf8 translation works in C# (#3766).
* Fix cpl_config.h.vc so it also works with mingw compiler on windows (#3960)
* Fix GEOS configuration checking affecting GEOS 3.0.4 (#3990)
* Support optional nmake.local file for local MSVC config (r22529).

== GDAL / Raster ==

* GDALReplicateWord(): fix error with missing one pixel (#4090)
* Fix issue with duplicating metadata (#3961)
* VRT: Fix issues with complex data types (#3977)
* VRT: Fix issue with nodata values and uninitialized buffers (#4045)
* VRT/.aux.xml: Fix performance with large color tables (#3961)
* VRT: Fix for VRTs smaller than the default block size (#4137)
* IDRISI/RST: Fix seg fault with missing fields in .ref and .rdc (#4100)
* IDRISI: Fix problem with spherical inverse flattening (#3757)
* GTIFF: Fix support for external overviews with COPY_SRC_OVERVIEWS (#3938)
* GTIFF: Correct handling of linear geokeys in feet (#3091)
* GTIFF: Fix edge anomoly issues with jpeg tiles (#4096)
* GTIFF: Don't skip driver even if library versions differ (#4101)
* SAGA: Fix reading/writing .sdat files larger than 2GB. (#4104)
* PCIDSK: Various fixes particularly for linked files (r21513)
* PCIDSK: Fixes for georeferencing parsing (#3913)
* ECW: Added/fixed french ECW coordinate systems (#3868)
* ECW: Force sign of y pixel size when needed (#393)
* ECW: Support EPSG:n style coordinate systems (#3955)
* ECW: Fixes related to geotransform overrides (#4131)
* ERS: Handle .ers files in case insensitive way (#3974)
* PNG: Add support for libpng >= 1.5.0 (#3914)
* JP2KAK: Support reading and writing resolution info (#3847)
* MRSID: Fix for configuring and building on MacOS X (#3910)
* HFA (.img): Fix filename computations and updating for renaming files (#3897)
* HFA (.img): Support reading 2bit compressed .img files (#3956)
* HFA (.img): Fix for rewriting projection/datum parameters (#3969)
* NITF: Correct hemisphere setting for ICORDS=U with world file (#3931)
* NITF: Added NITF_DISABLE_RPF_LOCATION_TABLE_SANITY_TESTS option (#3930)
* NITF: Make attribute section size hack safer (#4020)
* EHDR: Improved floating point file detection (#3933)
* ASRP: Approximated support for Azimuthal Equidistant in poles (#3946)
* HDF5/BAG: fix handle and memory leaks (#3953)
* HDF5: Avoid returning NULL from GetProjectionRef() (#4076)
* HDF5: Improve preservation of floating point precision in metadata (r22531).
* JPIPKAK: Fix problems with overviews, windowing and errors (#3967)
* USGSDEM: Fixing locale dependent parsing of numbers (#3886, #3979)
* RS2: Support subdataset oriented metadata and overviews (#4006)
* AIGRID: Support uncompressed integer images (ArcGIS 10) (#4035)
* gdalwarp: Include DstDensity mask in memory calculations (#4042)
* GeoRaster: Full backport for fixes and features (#4039,#3266,#4071,#4072,#4132)
* ENVISAT: Fix gcp computations for stripline products (#3160,#3709)
* ENVISAT: Fix gcp computations for MERIS products (#4086)
* SAR_CEOS: Fix potential crash with odd filenames. (#4065)
* SAR_CEOS: Skip datasets with wrong record length. (#4146)
* ENVI: Support ESRI style coordinate system strings (#3312)
* JAXAPALSAR: Fix datatype of 1.5 level product (#4136)
* gdalfillnodata: working file is created as bigtiff if possibly needed (#4088)

== OGR / Vector ==

* LIBKML: Fix build issue against static libkml (#3909)
* LIBKML: Improve checking to see if files are actually kml (#4003)
* GML: Fix crash with invalid geometries when axis order is lat/long (#3935)
* GML: Support curves in directEdges (#3934)
* GML: Do not force 3D linestrings for directed edges (#3936)
* GPX/GeoRSS: Fix expat library order issue on unix (#3948)
* PG/PGDUMP: Fix handling of NaN in zero width fields (#2112)
* PGDUMP: Add CREATE_SCHEMA and DROP_TABLE layer options (#4033)
* PGDUMP: Support wkbNone geometry better, and other subtle bugs (#4040)
* OGR SQL: Allow double quoting column name in SELECT DISTINCT (#3966)
* OGR SQL: Fix crashes with some uses of SUBSTR and CAST (#4055)
* OGR SQL: Fix crash on IS NULL test on floating point fields (#4091)
* OGR SQL: Fix error with SQL joins (#4112)
* MSSQL: Fix for IDENTITY INSERT problem (#3992)
* MSSQL: Fix for schema handling (#3951)
* MSSQL: Fix for geometry handling with ExecuteSQL() (#4149)
* OCI: Fix writing of tables with no general attributes (#4063)
* GeoJSON: Fix assertion on ESRI json (#4056, #4057)
* GeoJSON: Improve numeric field determination (#4082)
* SHAPE: Fix decoding of triangle fans in multipatch files (#4081)
* SHAPE: Fix multipolygon detection issue due to isClockwise failure (#4122)
* MITAB: Fix handling of filenames with two dots in them. (#4123)
* MITAB: Fix problems with null datetime values (#4150)
* ogr2ogr: Improve -overwrite/-append handling (#4012)
* ogr2ogr: Ensure -where fields are address in SetIgnoredFields (#4015)

(gdt)

Updated devel/SDL to 1.2.14nb9

(wiz)

Another fix for keyboard handling, from Opolo Remo,
and a fix for the color patterns, also from him.
Bump PKGREVISION.

(wiz)

no compiler language is required.

(obache)

Updated mail/getmail to 4.24.0nb1

(obache)

distutils pkg, register egg-info.

Bump PKGREVISION.

(obache)

Updated mail/archivemail to 0.9.0

(obache)

Update archivemail to 0.9.0.

Notable changes in archivemail 0.9.0:

  * IMAP: support for international mailbox names containing non-ASCII
    characters.

Notable changes in archivemail 0.8.0:

  * Removed the feature to setuid to the mailbox owners when run as root.
    This was a bad idea; it's really hard to do safely, if at all possible.
    Obsoletes: patch #2783134.
  * New option --archive-name, or short -a, to hard-code an archive filename.
    Like the --suffix and --prefix options, it is expanded with strftime().
    This option conflicts with archiving multiple mailboxes.  Closes: feature
    request #1306538.
  * New option --prefix, or short -p, to specify an archive name prefix.  Like
    a suffix specified with the --suffix option, the prefix is expanded with
    strftime().  Specifying this option disables the default archive name
    suffix.  Obsoletes: feature request #604281.  (Thanks Serafeim Zanikolas
    for an initial patch)
  * New option --all to archive all messages in a mailbox.  Closes: #1764846.
  * archivemail now expands wildcards in IMAP mailbox names.  For example, the
    url imaps://user@server/foo/* will expand to all subfolders of foo.
    Closes: feature request #1978540. Obsoletes: patch #1918937.
  * To determine the delivery date of a message, archivemail now looks for the
    timestamp of the latest 'Received' header before resorting to
    'Resent-Date' or 'Date'.  This should give much better results when there
    is no 'Delivery-date' header, which is still checked first.
    (Thanks Andrew Ferrier & Christian Brabandt)
    Closes: #1481316, #1764855, Debian bug #272666.
  * We now omit the dotlock if we don't have sufficient permissions to create
    it in the mbox directory. (The file is still locked with lockf.)
    Together with more changes under the hood (see below) this means
    archivemail can now operate on mbox files in the system mail spool.
    Closes: #855269.
  * Replaced some simple minded file operation security checks with more
    decent ones.  This means we can safely operate in /tmp, for example.  The
    price is that we no longer accept symlinked files.  Obsoletes: patch
    #1874868.
  * The archive now also gets locked while archivemail updates it.
  * mbox locking got completely rewritten.  Switched from flock to lockf
    locking, which is NFS-safe and portable, and we now lock with lockf first,
    then with a dotlock, instead of the other way around.  (This is makes
    archivemail compatible with Debian systems. ;)

(obache)

Updated mail/py-sendmail-admin to 1.0nb1

(obache)

distutils pkg, register egg-info.

Bump PKGREVISION.

(obache)

Updated mail/py-libgmail to 0.1.11

(obache)

Update py-libgmail to 0.1.10.

while here,
* LICENSE=gnu-gpl-v2
* register egg-info.

== Version 0.1.11 ==
libgmail.py
  * Fixed bug that broke attachment support (SF bug #2034927)
  * added .author_fullname field for messages
  * Don't crash on threads with google chat log (Debian bug #502458)

== Version 0.1.10 ==
libgmail.py
  * Use mechanize instead of ClientCookie [Patch #2014779]
  * Very basic Unicode support [Patch #1926861]

gmail_transport.py
  * New version that uses mechanize
    (owing again to Jose Rodriguez)

NOTE: libgmail now depends on mechanize, which
can be downloaded from:
  http://wwwsearch.sourceforge.net/mechanize/#download
  (in Debian/Ubuntu as python-mechanize, and an easy_install
  installer is also available)

== Version 0.1.9 ==
libgmail.py
    * Fixed login that was broken for a bunch of new
      gmail accounts, thanks to a patch by rhauer

NOTE: libgmail now depends on ClientCookie, which
can be downloaded from:
  http://wwwsearch.sourceforge.net/ClientCookie/#download

== Version 0.1.8 ==
libgmail.py
    * Added 'search' method to contactLists that returns
      an array of contacts who match a given search term
      (at some point, the contacts API is long overdue
      for a revamp, but for now, hey, why not)
      This is a patch by Alex Chiang    --WD--
    * libgmail now asks for the old Gmail interface,
      so that it isn't broken by the new Gmail updates.
      (Thanks to Aaron and Stu for work on this)
      (Fixes SF bug #1822662)

== Version 0.1.7 ==
libgmail.py
gmail_transport.py
    * Applied patch that adds proxy support, both
      for passwordless and password-ful proxies
      (is that a word?), by Jose Rodriguez --WD+SZ--

== Version 0.1.6.2 ==
libgmail.py
    * Bugfix for attachment problems --WD--
      (SF Bug #1793026, Patch #1799605 by 'stephster')
archive.py
    * Protect messages with a "from" line in them --WD--
      (SF Patch #1790809 by 'scop')

== Version 0.1.6.1 ==
libgmail.py
    * Bugfix for login problems --WD--

== Version 0.1.6 ==
libgmail.py
    * Added support for "Gmail Apps" aka "Gmail For Your Domain" --WD--

(obache)

audio/snd: Fix indirect linking error on DragonFly only

(marino)

audio/esound: Fix indirect linking error on DragonFly only

Since fixing the gnu linker v2.22 indirect linking errors for all platforms
have been met with controversy, this will be fixed with LDFLAGS.DragonFly
rather than LDFLAGS.  An identical LDFLAGS.IRIX already exists for this
package, and often this is the case for LDFLAGS.SunOS.  When pkgsrc
definitively decides to support unmodified gnu linker v2.22 behavior, we can
stop using LDFLAGS.DragonFly to address this issue.

(marino)

Recursive dependency bump for databases/gdbm ABI_DEPENDS change.

(sbd)

Use the correct version number for BUILDLINK_ABI_DEPENDS.gdbm

(sbd)

Updated net/tacacs-shrubbery to 4.0.4.22

(pettai)

PR/44791 -- apply markd's patch to unbreak build with older heimdal.

(shattered)

F4.0.4.21
        - do_auth.py - better Nexus support, better AV replacement, and only
          send roles to Nexus
        - fix bug in checking the return value of regexec() for login and
          enable ACLs.
        - do_auth.py - better Nexus support, better AV replacement, and only

F4.0.4.22
        - check of regexec() return value inverted

(pettai)

Add gettext-lib buildlink to gdbm's buildlink3.mk

Bump ABI_DEPENDS

(sbd)

Shut-up g++ error about invalid conversion.

(sbd)

When building in sim/ucsim/cmd.src:
  * cmdlex.cc needs to include <cstdio> for the EOF macro.
  * Search $(srcdir) before directorys in CXXFLAGS so that the "FlexLexer.h"
    in $(srcdir) is found before the one under PREFIX (from CXXFLAGS).

(sbd)

Updated lang/erlang to 14.1.4nb1

(sbd)

Add PLIST.Linux

Bump PKGREVISION

(sbd)