Updated textproc/p5-Text-Xslate to 3.1.2

(wen)

Update to 3.1.2

Upstream changes:
3.1.2 2014-02-20 21:09:47+0900
    [TEST]
    - Add a test for github issue #105, which will fail on Perl 5.19.x

(wen)

Remove tcl-itcl-3.4.1, newer version (4.0.0) is in base tcl-8.6 and later.

(obache)

itcl (4.0.0) is in tcl 8.6 (and later).

(obache)

Do same as bootstrap script

(obache)

Updated textproc/p5-PDF-API2 to 2.021

(wen)

Update to 2.021

Upstream changes:
2.021    2014-02-20

    - Fixed numerous bugs in the string parsing code, including the
      one reported in [RT #63918] by Frank Doepper.

    - [RT #41049] Rewrote literal string parsing to prevent a stack
      overflow due to an inefficient regex (reported by Sergei
      Fetisov).

    - [RT #91822] Fix compression of GIF images to ensure that output
      codes don't exceed 12 bits (reported by Vadim Repin).

    - The RunLengthDecode filter didn't actually work.  Its code has
      been rewritten and now passes basic encoding and decoding tests.

    - Fix Code128 barcode switching from Code C to Code B in certain
      cases (reported by Doru Petrescu).

(wen)

Updated www/p5-HTTP-Tiny to 0.043

(wen)

Update to 0.043

Upstream changes:
0.043    2014-02-20 20:40:23-05:00 America/New_York

    [FIXED]

    - Does not send absolute request URI when tunneling SSL via proxy

    - Fixes regression in setting host name to verify SSL

    - Protects tests from https_proxy and all_proxy when doing mock testing

0.042    2014-02-18 11:23:17EST-0500 America/New_York

    [ADDED]

    - If IO::Socket::IP 0.25+ is installed, HTTP::Tiny will use it for
      transparent IPv4 or IPv6 support.

0.041    2014-02-17 13:07:54-05:00 America/New_York

    [no code change, only an amended Changes file]

    [INCOMPATIBLE CHANGES (from 0.039)]

    - The 'proxy' attribute no longer takes precedence over the
      'http_proxy' environment variable.  With the addition of http_proxy
      and https_proxy attributes (and corresponding environment variable
      defaults), the legacy 'proxy' attribute now maps to the
      all_proxy/ALL_PROXY environment variable and only takes effect when
      other proxy attributes are not defined.

    [ADDED (since 0.039)]

    - Added 'keep_alive' attribute for single-server persistent connections
      (Clinton Gormley)

    - Added support for Basic authorization with proxies

    - Added support for https proxies via CONNECT

    [FIXED (since 0.039)]

    - Requests are made with one less write for lower latency (Martin
      Evans)

0.040    2014-02-17 13:02:47-05:00 America/New_York

    [INCOMPATIBLE CHANGES]

    - The 'proxy' attribute no longer takes precedence over the
      'http_proxy' environment variable.  With the addition of http_proxy
      and https_proxy attributes (and corresponding environment variable
      defaults), the legacy 'proxy' attribute now maps to the
      all_proxy/ALL_PROXY environment variable and only takes effect when
      other proxy attributes are not defined.

    [ADDED]

    - Added support for Basic authorization with proxies

    - Added support for https proxies via CONNECT

(wen)

Updated devel/p5-Module-CoreList to 3.07

(wen)

Update to 3.07

Upstream changes:
3.07 Fri Feb 21 2014
- fix for is_core tests and perls ending in 0

3.06 Thu Feb 20 2014
- updated for 5.19.9
- fix Module::CoreList::is_core default perl version
- fix Module::CoreList::is_core version comparision

3.05
  - Prepared for v5.19.9

(wen)

Updated net/p5-SOAP-Lite to 1.11

(wen)

Update to 1.11

Upstream changes:
1.11 February 21, 2014
    ! #93080 bug when SOAP::Lite is trying to parse WSDL definition by https with custom settings

1.10 January 23, 2014
    ! Fixed uninitialized warning when Context-Length is not set [github/dbeusee]
    ! avoid string eval in as_base64 [github/otrosien]
    ! #88915 warnings issued with using importing with +trace

1.09 January 14, 2013
    ! #92025 tighten multipart content-type matching [dmn@debian.org]

(wen)

Updated sysutils/tkdesk to 2.0p2

(obache)

Update fkdesk to 2.0p2, apply upstream patches, with fixes build with TCL 8.6,
Itcl 4.0.0 is bundled.
XXX: but not works well yet.

2.0p1  04/16/2004
Compilation of the file libdesk/tixImgXpm.c fails when using a Tk version lower than 8.3.

2.0p2  04/20/2004
Some Linux distributions are using newer "unofficial" versions of Incr Tcl
(Itcl). Though TkDesk 2.0 works with the official Itcl 3.2.1 release, these
newer Itcl installations disable backwards compatibility for some older Itcl
syntax that TkDesk still requires. The typical symptom of this problem is this
error message when TkDesk is started:
Error in startup script: can't rename "itcl_class": command doesn't exist...

(obache)

Add LICENSE=cc0-1.0-universal.

(hiramatsu)

Drop tcl-itcl/buildlink3.mk, now tcl includes tcl-itcl.
(PKGREVISION++ not necessary, failing to build after recent revbump)
-.include "../../lang/tcl-itcl/buildlink3.mk"
+#include "../../lang/tcl-itcl/buildlink3.mk"
.include "../../lang/tcl/buildlink3.mk"

(mef)

Note update of lang/ruby193-base package to 1.9.3p484nb3.

(taca)

Fix patch for rubygems/dependency_installer.rb, noted by Joerg Sonnenberger
via private e-mail.

Bump PKGREVISION.

(taca)

Updated filesystems/openafs to 1.6.6 [gendalia 2014-02-22]

(wiz)

Add patches/patch-util_google_sparsehash_sparseconfig.h to distinfo

(wiz)

Updated sysutils/xentools42 to 4.2.4

(prlw1)

Update xentools42 to 4.2.4

Removed patch-libxc_xc_dom_h: commited as cb08944a

This fixes the following critical vulnerabilities:

- CVE-2013-2212 / XSA-60 Excessive time to disable caching with HVM guests with
PCI passthrough
- CVE-2013-1442 / XSA-62 Information leak on AVX and/or LWP capable CPUs
- CVE-2013-4355 / XSA-63 Information leaks through I/O instruction emulation
- CVE-2013-4361 / XSA-66 Information leak through fbld instruction emulation
- CVE-2013-4368 / XSA-67 Information leak through outs instruction emulation
- CVE-2013-4369 / XSA-68 possible null dereference when parsing vif ratelimiting
info
- CVE-2013-4370 / XSA-69 misplaced free in ocaml xc_vcpu_getaffinity stub
- CVE-2013-4371 / XSA-70 use-after-free in libxl_list_cpupool under memory press
ure
- CVE-2013-4375 / XSA-71 qemu disk backend (qdisk) resource leak
- CVE-2013-4416 / XSA-72 ocaml xenstored mishandles oversized message replies
- CVE-2013-4494 / XSA-73 Lock order reversal between page allocation and grant t
able locks
- CVE-2013-4553 / XSA-74 Lock order reversal between page_alloc_lock and mm_rwlo
ck
- CVE-2013-4551 / XSA-75 Host crash due to guest VMX instruction execution
- CVE-2013-4554 / XSA-76 Hypercalls exposed to privilege rings 1 and 2 of HVM gu
ests
- CVE-2013-6375 / XSA-78 Insufficient TLB flushing in VT-d (iommu) code
- CVE-2013-6400 / XSA-80 IOMMU TLB flushing may be inadvertently suppressed
- CVE-2013-6885 / XSA-82 Guest triggerable AMD CPU erratum may cause host hang
- CVE-2014-1642 / XSA-83 Out-of-memory condition yielding memory corruption duri
ng IRQ setup
- CVE-2014-1891 / XSA-84 integer overflow in several XSM/Flask hypercalls
- CVE-2014-1895 / XSA-85 Off-by-one error in FLASK_AVC_CACHESTAT hypercall
- CVE-2014-1896 / XSA-86 libvchan failure handling malicious ring indexes
- CVE-2014-1666 / XSA-87 PHYSDEVOP_{prepare,release}_msix exposed to unprivilege
d guests
- CVE-2014-1950 / XSA-88 use-after-free in xc_cpupool_getinfo() under memory pre
ssure

Apart from those there are many further bug fixes and improvements.

(prlw1)

Updated sysutils/xenkernel42 to 4.2.4

(prlw1)

Update xenkernel42 to 4.2.4

This fixes the following critical vulnerabilities:

- CVE-2013-2212 / XSA-60 Excessive time to disable caching with HVM guests with PCI passthrough
- CVE-2013-1442 / XSA-62 Information leak on AVX and/or LWP capable CPUs
- CVE-2013-4355 / XSA-63 Information leaks through I/O instruction emulation
- CVE-2013-4361 / XSA-66 Information leak through fbld instruction emulation
- CVE-2013-4368 / XSA-67 Information leak through outs instruction emulation
- CVE-2013-4369 / XSA-68 possible null dereference when parsing vif ratelimiting info
- CVE-2013-4370 / XSA-69 misplaced free in ocaml xc_vcpu_getaffinity stub
- CVE-2013-4371 / XSA-70 use-after-free in libxl_list_cpupool under memory pressure
- CVE-2013-4375 / XSA-71 qemu disk backend (qdisk) resource leak
- CVE-2013-4416 / XSA-72 ocaml xenstored mishandles oversized message replies
- CVE-2013-4494 / XSA-73 Lock order reversal between page allocation and grant table locks
- CVE-2013-4553 / XSA-74 Lock order reversal between page_alloc_lock and mm_rwlock
- CVE-2013-4551 / XSA-75 Host crash due to guest VMX instruction execution
- CVE-2013-4554 / XSA-76 Hypercalls exposed to privilege rings 1 and 2 of HVM guests
- CVE-2013-6375 / XSA-78 Insufficient TLB flushing in VT-d (iommu) code
- CVE-2013-6400 / XSA-80 IOMMU TLB flushing may be inadvertently suppressed
- CVE-2013-6885 / XSA-82 Guest triggerable AMD CPU erratum may cause host hang
- CVE-2014-1642 / XSA-83 Out-of-memory condition yielding memory corruption during IRQ setup
- CVE-2014-1891 / XSA-84 integer overflow in several XSM/Flask hypercalls
- CVE-2014-1895 / XSA-85 Off-by-one error in FLASK_AVC_CACHESTAT hypercall
- CVE-2014-1896 / XSA-86 libvchan failure handling malicious ring indexes
- CVE-2014-1666 / XSA-87 PHYSDEVOP_{prepare,release}_msix exposed to unprivileged guests
- CVE-2014-1950 / XSA-88 use-after-free in xc_cpupool_getinfo() under memory pressure

Apart from those there are many further bug fixes and improvements.

(prlw1)

skip checking portability on installer/misc/makedeb.sh.

(mrg)

Updated pkgtools/pbulk to 0.52

(jperkin)

pbulk-0.52: ensure sort(1) is called with LC_ALL=C

Without this, a LANG of e.g. en_US.UTF-8 would sort "- *" before "+ <file>",
resulting in no packages being uploaded.

(jperkin)

+scipy-0.13.3

(rodent)

Upgrade OpenAFS to 1.6.6.
Remove unused options bos-new-config, fast-restart, & largefile.
Remove patches fixed upstream.

OpenAFS 1.6.6

  All platforms

    * As of this release, OpenAFS no longer ships uncompressed source tarballs.
      Tarballs are still shipped with both compression formats, gzip and bzip2.
      (10131)

    * Documentation improvements (10136 10314 10601)

    * Improved diagnostics and error messages (9412 10085 10274)

    * Avoid redefining "assert" in our public header files, which could
      cause failures when building some applications using them. (10096)

    * Fixes for parallel builds (10005 10309 10337)

    * Added a -s switch to afscp (not installed by default) to help simulate
      a slow client. (9416 9417)

    * Added a -probe switch to vlclient test program (not installed by default)
      to ping all vlservers in a cell in parallel. (9570)

  All server platforms
    * The fileserver now ignores any vice partitions with a NeverAttach flag
      file present in the root directory. (RT #130561) (9470 9471)

    * Restrict forcing CPS ("Current Protection Subdomain") recalculation in
      the fileserver to administrators. Also fixed a bug that could cause this
      operation to be incomplete. (9485 9487)

    * Allow non-DAFS fileservers to attach unusable volumes, restoring pre-1.6
      behaviour. (RT #131505) (9499)

    * Restored the pre-1.6 behaviour when running vos examine for a volume
      currently in a transaction, showing the volume as busy again rather than
      offline. (9685 9915 9916)

    * Reduced the minimum time a bos salvage takes from 5 seconds to 1. (9476)

    * Fixed buserver to not segfault when started with the -servers option.
      (RT #131706) (10166)

    * Salvager fixes, addressing a wide variety of possible problems from
      unnecessary salvaging to aborts (9282 9283 9457 9458 9459 9461 9462 9480
      9481 10165 10167)

    * Fixed a bug that could cause saved state information to be discarded
      when restarting a large or busy fileserver, which negatively impacted
      performance. (9683)

    * Fixed a bug that could have caused undefined behaviour in the vlserver
      in rare cases when a fileserver registered its addresses in the VLDB.
      (9429)

    * Added the -preserve-vol-stats switch to volserver, allowing it to keep
      the access statistics across volume restore and reclone operations
      instead of resetting them. (9477)

    * Inserted an exponential delay between retries when bosserver attempts to
      restart a server process. (9571 10199)

    * Improved vldb_check (not installed by default) to cope with broken
      vlentry names and volids, and provide more output to aid debugging.
      (10268)

    * Releasing a volume after adding a new RO site no longer touches any of
      the existing RO sites, if the RW data hasn't changed since the last
      release. (10174)

    * Make the copyDate field for RO clones have the same meaning as for
      remote RO volumes. Previously, the copyDate field for clones was updated
      every time we released. (9451)

    * Fixed potentially undefined behaviour in ptserver when too many pts
      ids are allocated. (10124)

    * Note that the server side NAT pings feature present in the prereleases
      was removed before the final release, since no positive feedback
      was provided during prerelease testing. (9420 10135)

  Linux servers

    * Start bosserver with -nofork in the systemd unit file, to allow systemd
      to track its state (10093)

  All client platforms

    * No longer track file locks on read-only volumes. Write locks can't
      succeed, read locks always will. Avoids log messages about this kind
      of lock. (8910)

    * Added the "fs flushall" subcommand, which makes the client discard all
      cached data. This was previously available on Windows only. (9065 9388
      9389 9390)

    * Fixed a bug that could make the client incorrectly believe its cache
      is up to date. This change could negatively impact AFS <-> DFS
      translators, should those still be running anywhere. (8898)

    * Several changes to avoid panicing in certain error conditions.
      (9131 9287 10354 10355 10356 10357) (partially addressing RT #131747)

    * Added the -rxmaxfrags switch to afsd, allowing to limit the number
      of UDP fragments sent or received per RX packet. (9430)

    * Build fixes for aklog on several platforms (RT #131716) (9917 10107 10275)

    * Require that the AFS mountpoint specified in the cacheinfo file is
      an absolute path. Relative paths result in a client that basically
      works but is not fully functional. (10253)

    * Fixed a bug that could cause one of the afsd threads to enter an infinite
      loop (10431 .. 10436)

  Linux clients

    * Support Linux kernels up to 3.13 (10241)

    * Fixed a bug that made readv/writev calls in AFS space fail with Linux
      kernels where generic_file_aio_read exists but those operations have
      not been switched to using aio_read/aio_write. This was a regression
      introduced with release 1.6.3 and affected at least RHEL 5.9 kernels.
      (10248)

    * Fixed a similar bug making core dumps fail in AFS space, affecting
      a much wider range of kernels including the most recent ones.
      (RT #131729) (10254)

    * Enhanced the keyring code to make PAGs work correctly on kernels with a
      distribution specific change to the Linux keyring code. This affected at
      least SLES 11 SP3 kernels. (10252)

    * Fixed a bug that could make failures during PAG instantiation go
      unnoticed. (10255)

    * Fixed a bug that made compilation fail for Linux kernels without
      keyring support. This affected at least the SLE 10 SDK and an
      OEM version of SLES 11 SP1. (10325)

    * Fixed build for kernels with user namespace support enabled. Likely
      to be required for Ubuntu 14.04 and eventually other distributions.
      (10456 10457 10458 10518 10472)

    * Support RHEL 6.5 kernels, and possibly others with changes backported
      from recent mainline kernels that touch getname/putname, by no longer
      using those functions. Previously, the client could cause a kernel
      panic when syscall auditing was enabled. (10578)

    * Make tmpfs usable as the cache filesystem again. This had been broken
      since kernel 3.1 (9950 10193)

    * When starting the client fails, clean up the backing device information
      created in sysfs, to avoid error messages during a subsequent start
      and possible system instability later on (10454)

    * Update Red Hat packaging to support Fedora >= 20, RHEL >= 7 and
      ELrepo kernels (10597 10619 10622 10703 10704)

  OS X Clients

    * Support OS X 10.9 "Mavericks" (10519 10541 10542 10543 10548 10549)

  AIX clients

    * Fixed a bug that caused the 1.6 AIX client to never receive any RX
      packets in the kernel. (RT #131725)

  FUSE client

    * Support Solaris 11 (9454 9455)

    * Allow other users to access filesystems mounted by root. (9452)

  FreeBSD

    * Build tvolser and dvolser on this platform (10122)
    * Several fixes to catch up with newer releases (10374 .. 10381)

  NetBSD

    * Build tsalvaged, tvolser and dvolser on this platform (10121)
    * Fixed build on NetBSD 5 and newer. (10138)

(gendalia)

Fix gtk-doc fallout.

(joerg)

Fix signature of main. Fix missing string.h. Request compat from TCL.

(joerg)

Fix build with newer libgcrypt.

(joerg)

Reapply libc++ support.

(joerg)

Fix build with newer bison.

(joerg)

Request C++11 support for Clang, so that linking works.

(joerg)

Fix build with Clang. Use normal cmake rules.

(joerg)

Annotate format string use.

(joerg)

Fix build with Clang's stricter literal format string logic.

(joerg)

Disable container forward definitions for libc++, it doesn't agree with
the use of inline namespaces.

(joerg)

Fix build with libc++.

(joerg)

gtk-doc update added new files. Bump revision.

(joerg)

Fix build with TCL 8.6.

(joerg)

Don't force forward declarations of containers.

(joerg)

Add special case for dependency checking in 'non-bootstrap' builds,
don't redefine 'pkg_info' if it already exists.
This is notably for pkg_rolling_replace -u in case pkg_install
was revbump'd and openssl happens to be in PKG_DEFAULT_OPTIONS.

(richard)

add utf8 [default] option to ghostscript-gpl

(richard)

Updated pkgtools/x11-links to 0.83

(jperkin)

x11-links-0.83: Assume Xorg if x11.pc exists.

This fixes detection on RHEL-based distributions which have the X11 libraries
and development files installed but not the X server.

(jperkin)

Updated www/firefox to 27.0.1nb1

(ryoon)

Bump PKGREVISION

Fix SUBST string.
>From tsutsui@, thank you.

(ryoon)

Pullup ticket #4332.

(tron)

Pullup ticket #4332 - requested by obache
multimedia/adobe-flash-plugin11: security update

Revisions pulled up:
- multimedia/adobe-flash-plugin11/Makefile                      1.25
- multimedia/adobe-flash-plugin11/distinfo                      1.23

---
  Module Name: pkgsrc
  Committed By: obache
  Date: Fri Feb 21 08:50:20 UTC 2014

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

  Log Message:
  Update adobe-flash-plugin11 to 11.2.202.341 for APSB14-07.

(tron)

Needs gmake. Not MAKE_JOBS_SAFE.

(fhajny)

Not a "root", it's really fhajny!

(obache)

Updated textproc/groonga to 4.0.0

(obache)

Update groonga to 4.0.0.

Release 4.0.0 - 2014/02/09
--------------------------

* Bump version to 4.0.0!

Improvements
^^^^^^^^^^^^

* [normalizer] Supported to show "checks" which is used for calculating next character position.
  Use WITH_CHECKS flag to enable this feature.
* [deb] Dropped Ubuntu 13.04 support.

Fixes
^^^^^

* Fixed a crash bug that an object in grn_expr is used after it is freed.
  Normally Groonga server users aren't affected this bug. This bug mainly affects Rroonga users.
  Because this bug is occured by specifying column name including pseudo column name -
  such as '_key' - Rroonga users may use the usage.
* Fixed not to execute unexpected cascade delete which is introduced Groonga 3.0.8 release.
  If source's range and index's domain are different, Groonga doesn't execute cascade delete.
  [groonga-dev,02073] [Reported by yoku]
* Fixed not to publish grn_snip structure. Use grn_obj instead of grn_snip.
  If you use grn_snip_close, please replace grn_snip_close to grn_obj_close.
* [snippet_html] Fixed a crash bug when --query is empty.
  [groonga-dev,02097] [Reported by Naoya Murakami]
* [snippet_html] Fixed to suppress ALERT level message when contents of column is empty text.
  [groonga-dev,02097] [Reported by Naoya Murakami]
* [groonga-httpd] Fixed a bug "off" is used as path name in groonga_query_log_path.
  [groonga-dev,02113] [Reported by Ryoji Yamamoto]

Thanks
^^^^^^

* yoku
* Naoya Murakami
* Ryoji Yamamoto

(obache)

gtk-doc-1.2 is responsible for recent PLIST change, record updated dependency

(richard)

update PLIST with missing '*-insensitive.png' files

(richard)

Updated multimedia/adobe-flash-plugin11 to 11.2.202.341

(obache)

Update adobe-flash-plugin11 to 11.2.202.341 for APSB14-07.

(obache)

Updated lang/nodejs to 0.10.26

(fhajny)

Update nodejs to 0.10.26.

2014.02.18, Version 0.10.26 (Stable)
* uv: Upgrade to v0.10.25 (Timothy J Fontaine)
* npm: upgrade to 1.4.3 (isaacs)
* v8: support compiling with VS2013 (Fedor Indutny)
* cares: backport TXT parsing fix (Fedor Indutny)
* crypto: throw on SignFinal failure (Fedor Indutny)
* crypto: update root certificates (Ben Noordhuis)
* debugger: Fix breakpoint not showing after restart (Farid Neshat)
* fs: make unwatchFile() insensitive to path (iamdoron)
* net: do not re-emit stream errors (Fedor Indutny)
* net: make Socket destroy() re-entrance safe (Jun Ma)
* net: reset `endEmitted` on reconnect (Fedor Indutny)
* node: do not close stdio implicitly (Fedor Indutny)
* zlib: avoid assertion in close (Fedor Indutny)

(fhajny)

note update of security/netpgp to 20140220

(agc)

Update netpgp to version 20140220

Changes from previous version:

+ portability fixes from xtraeme for his Linux distribution:
+ add search for ar(1) into autoconf (I fixed this differently
  so as not to rely on an automake check)
+ define __printflike if it's not already defined
+ fix missing asprintf (I fixed this differently)

(agc)

Actually use my DIOCGDISKINFO ioctl implementation by regenerating configure.

(prlw1)

Add a missing DESTDIR. From Leonardo Taccari <iamleot@gmail.com>
on pkgsrc-bulk.

(wiz)

Support X11BASE?=/usr for native X11 on RedHat and derivatives.

(jperkin)

Remove patches whose checksums were removed from distinfo during the
last update.

(wiz)

Sort.

(wiz)

Updated textproc/p5-Search-Xapian to 1.2.17.0

(schmonz)

Update to 1.2.17.0. From the changelog:

- Bump the version to indicate compatibility with Xapian 1.2.17.
- t/symbol-test/Makefile.PL: Pass $CPPFLAGS and $CXXFLAGS from the
  environment through to the sub-build, fixing symbol-test.t failure
  when built with CPPFLAGS=-D_GLIBCXX_DEBUG. [Olly Betts]

(schmonz)

Updated textproc/xapian-omega to 1.2.17

(schmonz)

Update to 1.2.17. From the changelog:

documentation:

* docs/overview.html: Add Abiword as an example use of --filter, based on patch
  from Frank J Bruzzaniti (fixes#383).

portability:

* Fix "no previous declaration" warning on platforms which don't have
  mkdtemp().

indexers:

* omindex:

  + Fix off-by-one when finding documents to delete which would sometimes cause
    omindex to fail to delete documents from the database when they weren't
    refound during an index update.

  + Decode dates in xlsx files.

  + Ignore extensions 'adm', 'cur', and 'ico' by default.

  + Group-readable files which are owner-readable but not world-readable should
    still get a "readable by owner" term added.  Reported by Emmanuel Garette.

build system:

* Compress source tarballs with xz instead of gzip.

* configure: Sync compiler warning flag machinery against xapian-core.  The
  changes are special handling for clang, passing -fshow-column where
  supported, and handling for new warning flags in GCC 4.6 and 4.7.

(schmonz)

Updated textproc/xapian to 1.2.17

(schmonz)

Update to 1.2.17. From the changelog:

API:
* Enquire::set_sort_by_relevance_then_value() and
  Enquire::set_sort_by_relevance_then_key(): Fix sense of reverse parameter.
  Reported by "boomboo" on IRC.
* BM25Weight: Fix case where (k1 == 0 || b == 0) but k2 != 0.  Reported by
  "boomboo" on IRC.
* Unicode::tolower(): Fix to give correct results for U+01C5, U+01C8, U+01CB,
  and U+01F2 (previously these were left unchanged).
* PositionIterator,PostingIterator,TermIterator,ValueIterator: Don't segfault
  if skip_to() or check() is called on an iterator which is already at_end().
  Reported by David Bremner.
* ValueCountMatchSpy: get_description() on a default-constructed
  ValueCountMatchSpy object no longer fails when xapian-core is built with
  --enable-log.
* ValueCountMatchSpy: get_total() on a default-constructed ValueCountMatchSpy
  object now returns 0 rather than segfaulting.

testsuite:
* Automatically probe for and hook in eatmydata to the testsuite using the
  wrapper script it now includes.
* Fix apitest to build when brass, chert or flint are disabled.
* If -v/--verbose is specified more than once to a test program, show the
  diagnostic output for passing tests as well as failing/skipped ones.
* queryparsertest: Change qp_scale1 to time 5 repetitions of the large query to
  help average out variations.
* queryparsertest: Add test coverage for explicit synonym of a term with a
  prefix (e.g. ~foo:search).
* apitest: Remove code from registry* testcases which tries to test the
  consequences of throwing an exception from a destructor - it's complex to
  ensure we don't leak memory while doing this (it seems GCC doesn't release
  the object in this case, but clang does), and it's generally frowned upon,
  plus C++11 makes destructors noexcept by default.
* Fix "make check" to actually removed cached databases first, as is
  intended.

brass backend:
* Fix handling of invalid block sizes passed to Xapian::Brass::open() - the
  size gets fixed as documented, but the uncorrected size was passed to the
  base file (and abort() was called if 0 was passed).
* Validate "dir_end" when reading a block.  (fixes #592)
* When moving a cursor on a read-only table, check if the block we want is in
  the internal cursor.  We already do this for a writable table, as it is
  necessary for correctness, but it's a cheap check and may avoid asking the
  OS for a block we actually already have.
* Correctly report the database as closed rather than 'Bad file descriptor'
  in certain cases.
* Reuse a cursor for reading values from valuestreams rather than creating
  a new one each time.  This can dramatically reduce the number of blocks
  redundantly reread when sorting by value.  The rereads will generally get
  served from VM cache, but there's still an overhead to that.

chert backend:
* Fix handling of invalid block sizes passed to Xapian::Chert::open() - the
  size gets fixed as documented, but the uncorrected size was passed to the
  base file (and abort() was called if 0 was passed).
* Validate "dir_end" when reading a block.  (fixes #592)
* When moving a cursor on a read-only table, check if the block we want is in
  the internal cursor.  We already do this for a writable table, as it is
  necessary for correctness, but it's a cheap check and may avoid asking the
  OS for a block we actually already have.
* Correctly report the database as closed rather than 'Bad file descriptor'
  in certain cases.
* Reuse a cursor for reading values from valuestreams rather than creating
  a new one each time.  This can dramatically reduce the number of blocks
  redundantly reread when sorting by value.  The rereads will generally get
  served from VM cache, but there's still an overhead to that.

flint backend:
* Fix handling of invalid block sizes passed to Xapian::Flint::open() - the
  size gets fixed as documented, but the uncorrected size was passed to the
  base file (and abort() was called if 0 was passed).
* Validate "dir_end" when reading a block.  (fixes #592)
* When moving a cursor on a read-only table, check if the block we want is in
  the internal cursor.  We already do this for a writable table, as it is
  necessary for correctness, but it's a cheap check and may avoid asking the
  OS for a block we actually already have.
* Correctly report the database as closed rather than 'Bad file descriptor'
  in certain cases.

build system:
* configure: Improve reporting of GCC version.
* Use -no-fast-install on platforms where -no-install causes libtool to emit a
  warning.
* docs/Makefile.am: Fix handling of MAINTAINER_NO_DOCS.
* Include UnicodeData.txt and the script to generate the unicode tables from
  it.
* Compress source tarballs with xz instead of gzip.
* Split XAPIAN_LIBS out of XAPIAN_LDFLAGS so that -l flags for libraries
  configure detects are needed appear after -L flags specified by the user
  that may be needed to find such libraries.  (fixes#626)
* XO_LIB_XAPIAN now handles the user specifying a relative path in
  XAPIAN_CONFIG, e.g.: "./configure XAPIAN_CONFIG=../xapian-core/xapian-config"
* Adjust XO_LIB_XAPIAN to strip _gitNNN suffix from snapshot versions.
* configure: Handle git snapshot naming when calculating REVISION.
* configure: Enable -fshow-column for GCC - things like vim's quickfix mode
  will then jump to the appropriate column for a compiler error or warning, not
  just the appropriate line.
* configure: Report GCC version in configure output.

documentation:
* postingsource.rst: Clarify a couple of points (reported by "vHanda" on IRC).
* The API documentation shipped with the release is now generated with
  doxygen 1.8.5 instead of 1.5.9, which is most evident in the different
  HTML styling newer doxygen uses.
* Document how Utf8Iterator handles invalid UTF-8 in API documentation.
* Improve how descriptions of deprecated features appear in the API
  documentation.
* docs/remote_protocol.rst: Correct error in documentation of REPLY_DOCDATA
  message.
* docs/overview.rst: Correct documentation for how to specify "prog" remote
  databases in stub files.
* Direct users to git in preference to SVN - we'll be switching entirely in
  the near future.

portability:
* Protect the ValueIterator::check() method against Mac OS X SDK headers
  which define a check() macro.
* Fix warning from xlC compiler.
* Avoid use of grep -e in configure, as /usr/bin/grep on Solaris doesn't
  support -e.
* Fix check for flags which might be needed for ANSI mode for compilers called
  'cxx'.
* configure: Improve handling of Sun's C++ compiler - trick libtool into not
  adding -library=Cstd, and prefer -library=stdcxx4 if supported.  Explicitly
  add -library=Crun which seems to be required, even though the documentation
  suggests otherwise.
* configure: clang doesn't support -Wstrict-null-sentinel or -Wlogical-op, so
  don't pass it these options.
* Fix build errors and warnings with mingw.
* Suppress "unused local typedef" warnings from GCC 4.8.
* If the compiler supports C++11, use static_assert to implement
  CompileTimeAssert.
* tests/zlib-vg.c: Fix two warnings when compiled with clang.
* Fix failure when built with -D_GLIBCXX_DEBUG - we were modifying the top()
  element of a heap before calling pop(), such that the heap comparison
  operation (which is called when -D_GLIBCXX_DEBUG is on to verify the heap is
  valid) would read off the end of the data.  In a normal build, this issue
  would likely never manifest.
* configure: When generating ABI compatibility checks in xapian/version.h, pass
  $CXXFLAGS and $CPPFLAGS to $CXXCPP as they could contain options which affect
  the ABI (such as -fabi-version for GCC).  (Fixes #622)
* Microsoft GUIDs in binary form have reversed byte order in the first three
  components compared to standard UUIDs, so the same database would report a
  different UUID on Windows to on other platforms.  We now swap the bytes to
  match the standard order.  With this fix, the UUIDs of existing databases
  will appear to change on Windows (except in rare "palindronic" cases).
* Fix a couple of issues to get Xapian to build and work on AIX.
* common/safeuuid.h: Remove bogus take-address-of from uuid handling code for
  NetBSD and OpenBSD.
* Under cygwin, use cygwin_conv_path() if using a new enough cygwin version,
  rather than the now deprecated cygwin_conv_to_win32_path().  Reported by
  "Haroogan" on the xapian-devel mailing list.
* common/safeuuid.h: Add missing '#include <cstdlib>' and qualify free with std.
* Fix 'unused label' warning when chert backend is disabled.
* xapian.h: Add check for Wt headers being included before us and defining
  'slots' as a macro - if they are, give a clear error advising how to work
  around this (previously compilation would fail with a confusing error).

tools:
* xapian-chert-update: Fix -b to work rather than always segfaulting (reported
  in http://bugs.debian.org/716484).
* xapian-chert-update: The documented alias --blocksize for -b has never
  actually been supported, so just drop mentions of it from --help and the man
  page.
* xapian-check:
  + Fix chert database check that first docid in each doclength chunk is more
    than the last docid in the previous chunk - previously this didn't actually
    work.
  + Fix database check not to falsely report "position table: Junk after
    position data" whenever there are 7 unused bits (7 is OK, *more* than 7
    isn't).
  + Fix to report block numbers correctly for links within the B-tree.
  + If the METAINFO key is missing, only report it once per table.
  + Fix database consistency checking to always open all the tables at the same
    revision - not doing this could lead to false errors being reported after a
    commit interrupted by the process being killed or the machine crashing.
    Reported by Joey Hess in http://bugs.debian.org/724610

examples:
* quest: Add --check-at-least option.

debug code:
* Fix assertion failure for when an OrPostList decays to an AndPostList - the
  ordering of the subqueries by estimated termfreq may not be the same as it
  was when the OrPostList was constructed, as the subqueries may themselves
  have decayed.  Reported by Michel Pelletier.
* Fix -Wcast-qual warning from GCC 4.7 when configured with --enable-log.

(schmonz)

Updated textproc/gtkspell to 2.0.16nb19

(wiz)

Add some files that are now installed, I guess by gtk-doc-1.20, and
depend on it. Bump PKGREVISION.

(wiz)

Updated databases/p5-App-Sqitch to 0.991

(schmonz)

Update to 0.991. From the changelog:

0.991  2014-01-16T23:24:33Z
    - Greatly simplified determining the Firebird ISQL client. It no longer
      tries so hard to find a full path, but does seach through the path list
      for a likely candidate between fbsql, isql-fb, and isql (or equivalents
      ending in .exe on Windows).
    - Removed a bunch of inappropriately pasted stuff from the Firebird
      tutorial, and updated it a bit.
    - `HEAD` and `ROOT` are now recognized as aliases for `@HEAD` and
      `@ROOT`, when querying the database, too. That means that `revert --to
      HEAD` now works the same as `revert --to @HEAD`, as had been expected
      in v0.990.
    - Eliminated "use of uninitialized value" warnings when database
      connections fail.
    - Reduced the minimum required DBD::Firebird to v1.11.
    - Fixed the `--verbose` option to the `target` command.
    - Eliminated more user-configuration issues in tests, thanks to
      chromatic.
    - Fixed test failures when the `$PGPASSWORD` environment variable is set,
      thanks to Ioan Rogers's test smoker.

0.990  2014-01-04T01:14:24Z
    [New Features]
    - Added new command and feature: `target`. Use it to manage multiple
      database targets, each with an associated URI and, optionally, a
      registry name and command-line client. Inspired by Git remotes.
    - Added Firebird engine. Three cheers to ��tefan Suciu for this
      contribution!
    - Added support for the generation of arbitrary scripts from templates to
      the `add` command. Just add template files to subdirectories of the
      `templates` directory, and scripts will be created in a directory of
      the same name based on those templates.
    - Added `--open-editor` option (and aliases) to the `add` and `rework`
      commands. This option will open the newly-added change scripts in the
      preferred editor. Thanks to Thomas Sibley for the patch!

    [Improvements]
    - Improved database driver loading to ensure the proper version of the
      driver is required.
    - Non-fatal but possibly unexpected messages -- which correspond to exit
      value 1 -- now send their messages to STDOUT instead of STDERR, and
      respect the `--quiet` option. Thanks to @giorgio-v for the report!
    - Added or replaced the `--target` option to commands that connect to a
      database to specify the name of target managed by the new `target`
      command or a database URI.
    - `HEAD` and `ROOT` are now recognized as aliases for `@HEAD` and
      `@ROOT`, respectively, since they are disallowed as change names,
      anyway, and folks often use them out of habit from Git.

    [Internals]
    - Replaced the engine-specific connection attributes with three
      attributes use by every engine:
      * `target`: The name of a target managed by the new `target` command.
        Defaults to a value stored for the `core.$engine.target`
        configuration variable. If that variable does not exist, the target
        falls back on the stringification of `uri`.
      * `uri`: a database URI with the format `db:{engine}:{dbname}` or
        `db:{engine}://{user}:{password}@{host}:{port}/{dbname}`. If its
        value is not passed to the constructor, a `uri` value is looked up
        for the associated `target`. If `target` is not passed or configured,
        or if it has no URI associated with it, the `config.$engine.uri`
        configuration variable is used. If that value does not exist, the URI
        defaults to `db:$engine:`. In any of these cases, if any of the
        `--db-*` options are passed, they will be merged into the URI.
      * `registry`: the name to use for the Sqitch registry schema or
        database, where Sqitch's own data will be stored, as appropriate to
        each engine. If its value is not passed to the constructor, a
        `registry` value is looked up for the associated `target`. If
        `target` is not passed or configured, or if it has no registry
        associated with it, the `config.$engine.registry` configuration
        variable is used. If no value is found there, it defaults to an
        engine-specific value, usually "sqitch".

    [Bug Fixes]
    - Fixed a bug when installing under local::lib. Thanks to Thomas Sibley
      for the pull request!
    - Eliminated "Wide character in print" warnings when piping the `log`
      command.
    - Documented that reworked changes do not have their verify tests run by
      the `verify` command. They do run when using the `--verify` deploy
      option.
    - Removed the documentation for the `add.with_deploy`, `add.with_revert`,
      and `add.with_verify` configuration variables, which were never
      implemented.

    [Deprecations]
    - Deprecated engine-specific connection attributes and configuration
      variables. See the "Internals" section for their replacements. The
      deprecated options are:
      * `core.$engine.username`
      * `core.$engine.password`
      * `core.$engine.db_name`
      * `core.$engine.host`
      * `core.$engine.port`
      * `core.$engine.sqitch_schema`
      * `core.$engine.sqitch_db`
    - Deprecated all command-specific options with the string "target" in
      them, such as `--to-target`, `--upto-target`, etc. They have been
      replaced with options containing the string "change", instead, such as
      `--to-change` and `--upto-change`. Few people used these options,
      preferring their shorter aliases (`--to`, `--upto`, etc.).
    - Deprecated the `--deploy-template`, `--revert-template`, and
      `--verify-template` options to the `add` command. They are replaced
      with a single option, `--use` which takes a key/value pair for the
      script name and template path. This makes it useful for arbitrary
      script generation, as well.
    - Deprecated the `--deploy`, `--revert`, and `--verify` options to the
      `add` command, as well as their `--no-*` variants. They are replaced
      with two new options, `--with` and `--without`, to which a script name
      is passed. These are useful for arbitrary script generation, as well.
    - Deprecated the `add.deploy_template`, `add.revert_template`, and
      `add.verify_template` configuration settings. They have been replaced
      with a section, `add.templates`, which is more general, and supports
      arbitrary script generation, as well.

    [Incompatibilities]
    - Removed the undocumented `--test` option to the `add` command.
    - Changed the meaning of `--target` from specifying a change to
      specifying a deployment target. Use the new `--change` option to
      specify a change.

0.983  2013-11-21T21:50:12Z
    - Fixed "Use of uninitialized value" in the MySQL engine. Thanks to
      Jean-Michel REY for the report.
    - All tests now protect against failures due to the presence of the
      `$SQITCH_CONFIG` environment variable (issue #114).
    - The installer now respects the `distdir` option to `Build.PL` when
      searching for existing templates. Important for packaging.
    - Fixed the error "Table 'sqitch.changes' doesn't exist" when deploying
      to a MySQL database that exists but has not been initialized. Thanks to
      Jean-Michel REY for the report!
    - Refactored the handling of the C<--log-only> option so it sets an
      engine attribute, rather than passing the flag to a whole stack of
      method calls.
    - Fixed "Argument "en_us" isn't numeric" error on Windows.
    - Now using `LC_ALL` instead of `LC_MESSAGES` when setting the locale, as
      the latter is not present on Windows.
    - The sqitch-pg RPM now requires DBD::Pg 2.0.0 or higher.
    - Improved handling of invalid command names so that the error message is
      less ambiguous when triggered by a Perl parse error.
    - Added `-m` as an alias for `--note`, for you Git folks out there.
    - Added exception handling to the Postgres and Oracle engines to avoid
      unexpected errors when deploying to a database that has not been
      deployed to before.
    - Updated detection of an uninitialized database to double-check with the
      engine that it really thinks it's uninitialized, not just that the
      "changes" table is missing. This should catch the case where the
      database has its own "changes" table unrelated to Sqitch.

0.982  2013-09-11T18:26:07Z
    - Errors thrown by Template toolkit are no longer silently ignored.
    - Variables passed to change templates are now cloned before the
      execution of each template. This prevents one template from deleting
      variable values another template might also need.
    - Fixed "The getpwnam function is unimplemented" errors on Win32.
    - No longer runs revert scripts when deploying with `--log-only` and a
      verify script fails, as that could lead to data loss (yikes!). Thanks
      to BryLo for the report (issue #112).

0.981  2013-09-06T00:22:26Z
    - Now use Encode::Locale to try to decode the user's full name from the
      system encoding when fetched from the system on all OSes. Note that
      this is not necessary if the `user.name` config is explicitly set, as
      recommended. Issue #107.
    - Removed the special-case handling of the user's full name fetched from
      the system on OS X.
    - Added call to `sleep` to test in an attempt to fix SQLite failures.
    - The SQLite engine now requires that the SQLite client be 3.3.9 or
      later, for support of the `-bail` option.
    - Bug fix: The MySQL engine now properly uses the host, port, and
      password options when connecting to the database. Thanks to vreb87 for
      the report!

0.980  2013-08-28T21:40:00Z
    - Changed the default SQLite Sqitch database name from
      `$dbname-sqitch.$suffix` to `sqitch.$suffix`. The `$suffix` still
      comes from the destination database name. This breaks compatibility
      with previous releases. If you need the old name, set it with
      `sqitch config core.sqlite.sqitch_db $dbname`.
    - Fixed encoding of the user's full name when fetched from the system on
      OS X. Thanks to Tomohiro Hosaka for the pull request!
    - Fixed test failures when DBD::SQLite is installed but compiled with
      SQLite 3.7.10 or lower.
    - Fixed a bug where declaring a dependency on a reworked change would
      incorrectly result in the error "Key "foo" matches multiple changes".
      Thanks to BryLo for the report (issue #103).
    - Modified tests to allow them to run in parallel without stomping on
      each other.
    - Bundling of options, such as `-vvv`, now works properly (issue #108).
    - Added alias `--get-regexp` for `--get-regex` to the `config` command.
      This brings it in line with the documentation for the `config` command
      (Issue #110).
    - Fixed all of the `config` command actions that contain a dash so that
      they actually work. Thanks to ��tefan Suciu for the report (issue #110).
    - All leading and trailing white space is now trimmed from plan notes,
      rather than just vertical white space. Thanks to Ronan Dunklau for the
      report (issue #106).
    - The `status` command now notices if the specified database is
      uninitialized and says as much, rather than dying with an SQL error
      (issue #109).
    - When reading the user's username from the system Sqitch now uses
      Encode::Locale to try to decode the value from the system encoding.
      Issue #107.
    - Compatibility change: Changed the location and name of script template
      files. Previously they were called `deploy.tmpl`, `revert.tmpl`, and
      `verify.tmpl`, and they lived in the `templates` subdirectory of the
      system-wide and user-specific configuration directories. They now live
      in subdirectories of the `templates` directory named for each action
      (deploy, revert, and verify), and with file names matching engine names
      (`pg.tmpl`, `sqlite.tmpl`, `oracle.tmpl`, and `mysql.tmpl`). The
      installer will move old files from the system-wide config directory
      (`sqitch --etc-path`) to their new homes, named `pg.tmpl` and
      `sqlite.tmpl`. It assumes no customizations exist for Oracle. If that's
      not true in your case, simply copy the `pg.tmpl` files to
      `oracle.tmpl`.
    - Added the `--template-name` option to the `add` command. By default, it
      looks for templates named for the current engine. The option allows for
      the user of task-specific templates. For example, if you create
      templates named `createtable.tmpl` in the `deploy`, `revert`, and
      `verify` subdirectories of `~/.sqitch/templates`, You can specify
      `--template-name createtable` to use those templates when adding a
      change.
    - Added the `--exists` option to the `show` command.
    - Fixed the `--set` option to the `add` command so that duplicate keys
      have their values passed to the template as an array, as documented.
    - If Template::Toolkit is installed, the `add` command will use it for
      processing templates instead of Template::Tiny. This makes it easy to
      upgrade the templating environment just by installing a module.

0.973  2013-07-03T13:47:22Z
    - Now Require DBD::SQLite compiled with SQLite 3.7.11 or higher. It
      always has, but now it throws a meaningful exception if an older
      version is compiled into DBD::SQLite. Thanks to Damon Buckwalter for
      the report.
    - When a deploy fails because of missing dependencies, the list of
      missing dependencies no longer contains duplicates. Thanks to Damon
      Buckwalter for the report.

0.972  2013-05-31T23:26:52Z
    - Fixed test failures on Windows.
    - Fixed locale configuration on Windows so that `sqitch` will actually
      run, rather than exiting with an error about `LC_MESSAGES` not being
      set.
    - Fixed a test hang on Windows when DBD::Oracle is installed but the
      Oracle libraries (`OCI.dll`) are not or cannot be found. This was
      triggering a UI dialog that did not dismiss itself. Using Win32::API
      to work around this issue. Thanks to Jan Dubois for the fix.

0.971  2013-05-18T21:08:51Z
    - Removed most uses of the smartmatch operator, since as of Perl 5.17.11
      it is marked as experimental, and silenced the warning where it is
      still used.
    - Added 0.1s sleep between logging changes back-to-back in the engine
      tests, mostly to try to get SQLite to generate different timestamps.
      Pretty sure the recent test failures have been due to the passage of
      less than a millisecond between the two inserts.
    - Added the `shell` and `quote_shell` methods to Sqitch.pm for shelling
      out a command.
    - Sqitch now shells out to an editor when opening a file for the user to
      edit. For example, if the `$EDITOR` environment variable is set to
      `"emacs -nw"`, it will now work. Thanks to Florian Ragwitz for the
      report (issue #80).
    - Removed the pod-checking tests from the distribution.

0.970 2013-05-09T00:21:06Z
    - Fixed the default ordering of changes displayed by the `plan` command.
      They are now ascending by default.
    - Switched to PerlIO::utf8_strict for fast character encoding and
      decoding.
    - The help emitted when an unknown option is passed to `sqitch` now
      consists of a usage statement and brief table of options, rather than
      the entire man page.
    - Added the project name in a header to the output of the `plan` command.
    - Added the Oracle engine.
    - Added `sqitchtutorial-oracle.pod`, a Oracle-specific variant of
      `sqitchtutorial.pod`.
    - Added missing version declaration to the App::Sqitch::Plan::* modules.
    - Devel::StackTrace 1.30 is now properly required (it was previously
      recommended).
    - The `--show-tags` and `--show-changes` options to the `status` command
      now show the changes when the project plan cannot be found (issue #90).

0.965  2013-04-23T16:25:59Z
    - Fixed failing test due to line-ending character variations on Windows.
      Many thanks to Jan Dubois for the testing help.
    - Replaced all uses of `$/` in output to `"\n"`. Thanks to Jan Dubois for
      pointing out the incorrect use of `$/`.
    - Fixed build error that prevented installation on Perl 5.10 when the
      parent module was not installed.

0.964  2013-04-15T18:47:30Z
    - Fixed test failures on Perl versions lower than 5.14 when DBD::SQLite
      or DBD::Pg is not installed.
    - Removed DBD::SQLite from the list of build dependencies.
    - Fixed test failures due to encoded (wide-character) warnings on
      triggered on systems with non-english locales. Thanks to Alexandr
      Ciornii for the smoke testing that revealed this issue.
    - Removed overriding of Throwable's `previous_exception` in
      App::Sqitch::X on Throwable 0.200007 and higher, where it is no longer
      needed.
    - Changed test comparing file contents that fails on Windows to do a
      looser comparison and hopefully fix the test failure.

0.963  2013-04-12T19:11:29Z
    - Fixed a test failure when Git is in the execution path and the test is
      not run from a Git checkout.
    - Added `plan` to `sqitchchanges`, the contents of which are shown when
      Sqitch is run with no command.
    - Removed the unique constraint on tag names in the database, as it
      prevented two projects from having the same tag name. Replaced it with
      a unique constraint on the project and tag names. Folks with production
      PostgreSQL installs should run these queries:
          ALTER TABLE sqitch.tags DROP CONSTRAINT tags_tag_key, ADD UNIQUE(project, tag);
          COMMENT ON COLUMN sqitch.tags.tag IS 'Project-unique tag name.';
    - Fixed failing tests when DBD::SQLite is not installed.
    - Removed dependency on Git::Wrapper. The `checkout` command does things
      very simply, and we already have tools for running command-line
      applications. So we just take advantage of that. The code is no more
      complicated than it was before.
    - Added the `core.vcs.client` configuration setting. Defaults to `git`
      (or `git.exe` on Windows).

0.962  2013-04-10T17:10:05Z
    - Fixed failing test on Perl 5.12 and lower.
    - Fixed the French translation by re-encoding it in UTF-8 (Ronan
      Dunklau).
    - Fixed the loading of the editor with placeholder text to properly
      encode that text as UTF-8 (Ronan Dunklau).

(schmonz)

Add and enable p5-URI-{db,Nested}.

(schmonz)

Added www/p5-URI-db version 0.12

(schmonz)

Add p5-URI-db, a class for database URIs. They're inspired by JDBC
URIs and PostgreSQL URIs, though they're a bit more formal.

(schmonz)

Added www/p5-URI-Nested version 0.10

(schmonz)

Add p5-URI-Nested, a Perl interface for nested URIs -- that is,
URIs that contain other URIs. The basic format is:

    {prefix}:{uri}

Some examples:

* `jdbc:oracle:scott/tiger@//myhost:1521/myservicename`
* `db:postgres://db.example.com/template1`

(schmonz)

Add and enable p5-PerlIO-utf8_strict.

(schmonz)

Added devel/p5-PerlIO-utf8_strict version 0.004

(schmonz)

Add p5-PerlIO-utf8_strict, a fast and correct UTF-8 PerlIO layer.
Unlike perl's default :utf8 layer it checks the input for correctness.

(schmonz)

from upstream: fix the fix for CVE-2013-6885
makes the workaround for AMD CPU erratum 793 work not only on
64-bit hypervisors but also for 32bit
bump PKGREV
(compile tested only)

(drochner)

Updated textproc/p5-Text-Markdown-Discount to 0.11

(schmonz)

Update to 0.11. From the changelog:

- upgraded Discount to 2.1.7, which fixed an issue generating <em> tags
  from _<nonascii>_ and *<nonascii>*
- doc typo fix (by @harupong)

(schmonz)

Updated mail/p5-Mail-Ezmlm to 0.08.2

(schmonz)

Update to 0.08.2. From the changelog:

- fixed handling of the 'owner' setting for ezmlm-idx > v5
- updated ezmlm-idx version detection
- allow "@" in the path of a mailing list
- add modules Mail::Ezmlm::GpgKeyRing and Mail::Ezmlm::GpgEzmlm
- fixed issues of Mail::Ezmlm::GpgEzmlm with ezmlm-idx v0.4x lists
- added check for external dependency to the test script

(schmonz)

+ cdialog-1.2.20140219, cvs-fast-export-1.8,
  libreoffice4-4.2.1, libreoffice4-bin-4.2.1, reposurgeon-3.6,
  webkit-gtk-2.2.5, zoom-1.1.5.

(wiz)

Remove unused patch after update.

(wiz)

Updated biology/bioperl to 1.6.923

(fhajny)

Update biology/bioperl to 1.6.923.

Way too many changes (0.7.0 was released in 2001!).

See changelog at:

https://github.com/bioperl/bioperl-live/blob/master/Changes

(fhajny)

Updated www/firefox-l10n to 27.0.1

(ryoon)