Updated gupnp to 1.0.0.

1.0.0
=====

- Only version changed

0.99.0
======
- Fix build on OS X

All contributors to this release:
- Tomasz Pajor <tomek@polishgeeks.com>
- Jens Georg <mail@jensge.org>

0.20.18
=======

- Retry description URL on failure.

Bugs fixed in this release:
- https://bugzilla.gnome.org/show_bug.cgi?id=694450

All contributors to this release:
- Jens Georg <mail@jensge.org>
- Jozef Šiška <jozef.siska@streamunlimited.com>

0.20.17
=======

- Update autogen.sh from GNOME template
- Cache user agents by IP
- Remove intltool
- Use newTargetValue in light example

Bugs fixed in this release:
- https://bugzilla.gnome.org/show_bug.cgi?id=653894
- https://bugzilla.gnome.org/show_bug.cgi?id=763945
- https://bugzilla.gnome.org/show_bug.cgi?id=764050

All contributors to this release:
- Jens Georg <mail@jensge.org>
- Philip Withnall <philip@tecnocode.co.uk>

0.20.16
=======

- Don't hardcode python path in gupnp-binding-tool
- Check for IFA_FLAGS to not bind ourselves to a recent kernel version
- Add missing m4 files
- Fix issues with introspection
- Make GUPnPServiceAction boxed
- Some JavaScript examples
- Distribute VAPI file in tarball

Bugs fixed in this release:
- https://bugzilla.gnome.org/show_bug.cgi?id=760217
- https://bugzilla.gnome.org/show_bug.cgi?id=760430
- https://bugzilla.gnome.org/show_bug.cgi?id=761265

All contributors to this release:
- Jens Georg <mail@jensge.org>
- Ting-Wei Lan <lantw@src.gnome.org>

0.20.15
=======

- Remove gnome-common
- Fix issue with gupnp_service_signals_autoconnect() blocking endlessly.
- Fix return type for BIN_HEX.
- Update some m4 macros.
- Fix typos in API docs.

NetworkManager context manager:
- Fix use-after-free issue.
- Fix various static code analysis findings.

Netlink context manager:
- Use proper macros for RTM_*ADDRESS messages.
- Use device name from netlink.
- Add possibility to dump received netlink packets for debugging.
- Fix issues with losing contexts after DHCP renew happens.
- Some style fixes.

Added/updated dependencies:
- GSSDP >= 0.14.13

Bugs fixed in this release:
- https://bugzilla.gnome.org/show_bug.cgi?id=736365
- https://bugzilla.gnome.org/show_bug.cgi?id=741257
- https://bugzilla.gnome.org/show_bug.cgi?id=750936
- https://bugzilla.gnome.org/show_bug.cgi?id=752490
- https://bugzilla.gnome.org/show_bug.cgi?id=759000

All contributors to this release:
- Jens Georg <mail@jensge.org>
- Debarshi Ray <debarshir@gnome.org>
- Olivier Crête <olivier.crete@collabora.com>
- Guillaume Marquebielle <guillaume.marquebielle@parrot.com>
- Bastien Nocera <hadess@hadess.net>

0.20.14
=======

Changes since 0.20.13:

Added/updated dependencies:
GLib 2.0 >= 2.40.0

- Simplify UUID handling.

Bugs fixed in this release:
- https://bugzilla.gnome.org/show_bug.cgi?id=743072
- https://bugzilla.gnome.org/show_bug.cgi?id=744221

All contributors to this release:
- Jens Georg <mail@jensge.org>

(wiz)

Updated net/gssdp to 1.0.0

(wiz)

Updated gssdp to 1.0.0.

0.99.0
======

- Don't leak arp lookup symbol
- Minor documentation updates

All contributors to this release:
- Jens Georg <mail@jensge.org>

0.14.16
=======

- Fill the interface index for bionic.
- Remove unused include.
- Fix interface lookup on Win32

Bugs fixed in this release:
- https://bugzilla.gnome.org/show_bug.cgi?id=762455
- https://bugzilla.gnome.org/show_bug.cgi?id=765688
- https://bugzilla.gnome.org/show_bug.cgi?id=765709

All contributors to this release:
- Aur辿lien Chabot <aurelien.chabot@parrot.com>
- Jens Georg <mail@jensge.org>
- Florian Zwoch <fzwoch@gmail.com>

0.14.15
=======

- Actually use the TTL that was set by the user
- Use if_nametoindex instead of IOCTL when available
- Implement a User Agent cache based on SSDP packages
- Fix potential strncpy overflow
- Add missing m4 macros
- Disable PKTINFO on OS X
- Do not leak host_addr in clients.

Bugs fixed in this release:
- https://bugzilla.gnome.org/show_bug.cgi?id=653894
- https://bugzilla.gnome.org/show_bug.cgi?id=762324
- https://bugzilla.gnome.org/show_bug.cgi?id=764677

All contributors to this release:
- Jens Georg <mail@jensge.org>
- Philip Withnall <philip.withnall@collabora.co.uk>
- Philip Withnall <philip@tecnocode.co.uk>

0.14.14
=======

- Fix warning if the IP address is not known.
- Fix issue with discovery.

Bugs fixed in this release:
- https://bugzilla.gnome.org/show_bug.cgi?id=760702

All contributors to this release:
- Jens Georg <mail@jensge.org>
- Philip Withnall <philip@tecnocode.co.uk>

0.14.13
=======

- Make "host-ip" of GSSDPClient a construction property to give the caller the
  chance to bypass the IP autodetection during client creation.

All contributors to this release:
- Jens Georg <mail@jensge.org>

0.14.12.1
=========

- Update m4 macros from autoconf-archive.

Bugs fixed in this release:
- https://bugzilla.gnome.org/show_bug.cgi?id=758778

All contributors to this release:
- Jens Georg <mail@jensge.org>

0.14.12
=======

- Fix issues with names of loopback device on *BSD.
- Move gtest code one level up and examples to their own folder.
- Fix a possible NULL derefence in gssdp-device-sniffer.
- Remove use of gnome-common for building.

Bugs fixed in this release:
- https://bugzilla.gnome.org/show_bug.cgi?id=743070

All contributors to this release:
- Jens Georg <mail@jensge.org>
- Thomas Klausner <tk@giga.or.at>

(wiz)

Updated fonts/fira to 4.202

(wiz)

Updated fira to 4.202.

Adds FiraSansCondensed and FiraMono-Medium.

(wiz)

Updated print/cups to 2.2.0

(wiz)

Updated cups to 2.2.0.

CHANGES.txt - 2.2.0 - 2016-09-13
--------------------------------

CHANGES IN CUPS V2.2.0

- Normalized the TLS certificate validation code and added additional
  error messages to aid troubleshooting.
- http*Connect did not work on Linux when cupsd was not running
  (Issue #4870)
- The --no-remote-any option of cupsctl had no effect (Issue #4866)
- http*Connect did not return early when all addresses failed
  (Issue #4870)

CHANGES IN CUPS V2.2rc1

- Updated the list of supported IPP Everywhere media types.
- The IPP backend did not validate TLS credentials properly.
- The printer-state-message attribute was not cleared after a
  print job with no errors (Issue #4851)
- The CUPS-Add-Modify-Class and CUPS-Add-Modify-Printer operations did
  not always return an error for failed adds (Issue #4854)
- PPD files with names longer than 127 bytes did not work (Issue #4860)
- Updated localizations (Issue #4846, PR #4858)

CHANGES IN CUPS V2.2b2

- Added Upstart support (PR #4825)
- CUPS now supports Let's Encrypt certificates on Linux.

CHANGES IN CUPS V2.2b1

- All CUPS commands now support POSIX options (Issue #4813)
- The scheduler now restarts faster (Issue #4760)
- Improved performance of web interface with large numbers of jobs
  (Issue #3819)
- Encrypted printing can now be limited to only trusted printers and
  servers (<rdar://problem/25711658>)
- The scheduler now advertises PWG Raster attributes for IPP Everywhere
  clients (Issue #4428)
- The scheduler now logs informational messages for jobs at LogLevel
  "info" (Issue #4815)
- The scheduler now uses the getgrouplist function when available
  (Issue #4611)
- The IPP backend no longer enables compression by default except for
  certain raster formats that generally benefit from it
  (<rdar://problem/25166952>)
- The scheduler did not handle out-of-disk situations gracefully
  (Issue #4742)
- The LPD mini-daemon now detects invalid UTF-8 sequences in job,
  document, and user names (Issue #4748)
- The IPP backend now continues on to the next job when the remote
  server/printer puts the job on hold (<rdar://problem/24858548>)
- The scheduler did not cancel multi-document jobs immediately
  (<rdar://problem/24854834>)
- The scheduler did not return non-shared printers to local clients
  unless they connected to the domain socket (<rdar://problem/24566996>)
- The scheduler now reads the spool directory if one or more job cache
  entries point to deleted jobs (<rdar://problem/24048846>)
- Added support for disc media sizes (<rdar://problem/20219536>)
- The httpAddrConnect and httpConnect* APIs now try connecting to
  multiple addresses in parallel (<rdar://problem/20643153>)
- The cupsd domain socket is no longer world-accessible on macOS
  (<rdar://problem/7542560>)
- Interface scripts are no longer supported for security reasons
  (<rdar://problem/23135640>)
- Added a new cupsHashData API and support for hashed job passwords
  (<rdar://problem/20221502>)
- Localization fixes (<rdar://problem/25292403>,
  <rdar://problem/25461517>, Issue #4041, Issue #4796)
- Documentation changes (Issue #4624, Issue #4781)
- Packaging fixes (PR #4832)

(wiz)

Add missing dependency.

(wiz)

Add upstream bug report URL.

(wiz)

Work around build problem in C locale.

(wiz)

Updated graphics/ImageMagick to 7.0.3.1

(wiz)

Updated ImageMagick to 7.0.3.1.

2016-09-20  7.0.3-1 Cristy  <quetzlzacatenango@image...>
  * Release ImageMagick version 7.0.3-1, GIT revision 18851:ad91ea8:20160920.

2016-09-16  7.0.3-1 Dirk Lemstra <dirk@lem.....org>
  * Added layer RLE compression to the PSD encoder.
  * Added define 'psd:preserve-opacity-mask' to preserve the opacity mask
    in a PSD file.
  * Fixed issue where the display window was used instead of the data window
    when reading EXR files (reference
    https://www.imagemagick.org/discourse-server/viewtopic.php?f=3&p=137849).

(wiz)

Updated graphics/adwaita-icon-theme to 3.22.0

(wiz)

Updated adwaita-icon-theme to 3.22.0.

3.22.0
  - simplified input-table-symbolic
  - opacity fixes for networking-vpn-acquring
  - invisible go-down-symbolic

3.21.91
  - user-available-symbolic transparency fix
  - calculator app icon update
  - fixed cursors
  - folder-download-symbolic style fixes
  - vertical visual centering for go-up, go-down-symbolic icons

3.21.2
  - network-wireless-connected improvements
  - document-edit, text-editor, modem-symbolic shape improvement
  - baseline-align folders
  - reshape folders
  - add a context-menu cursor
  - improve grab and dnd-no-drop cursors

(wiz)

Adapt patch for better upstreaming.

(wiz)

Updated print/abcm2ps to 8.12.4

(wiz)

Updated abcm2ps to 8.12.4.

---- Version 8.12.4 - 2016-09-25 ----

%%gutter added
(asked by Paul Hardy)
Check if spaces in PostScript font names
(reported by Dave Marcus)
Fix loss of computed PostScript values on page break with SVG output
(reported by David Lacroix)
Fix 'ghosts' at bottom of page when justified or centered text in page
(reported by Hudson Lacerda)

---- Version 8.12.3 - 2016-08-25 ----

Fix bad shape of cresc/dimin on next line after continuation
(reported by David Lacroix)
Fix bad scan of %%tuplets
(reported by David Lacroix)

---- Version 8.12.2 - 2016-08-24 ----

Handle long decorations on second repeats
(reported by Timm Reasbeck)
Add 4th parameter to %%tuplets ('where')
(asked by Ben)
Treat back the long decorations in overlayed voices ending
on a measure bar
(reported by David Lacroix)
Fix 'ghosts' at top of page when justified or centered text in page
(fixed by Hudson Lacerda)
Fix bad scale with '-s' when %%scale inside ABC
(reported by Hudson Lacerda)

---- Version 8.12.1 - 2016-08-05 ----

Fix bad alternate heads since 8.12.0

---- Version 8.12.0 - 2016-08-05 ----

Handle glissendo on chords
(reported by Timm Reasbeck)
Fix bad SVG output when text decoration before glissendo
Fix bad note head replacement when defined as a string since 8.5.0
Fix bad tie error when crossing a meter change
(reported by David Lacroix and Jean-Luc Zins)
Clash of ties on second interval chord
(reported by David Lacroix)

---- Version 8.11.7 - 2016-06-24 ----

Loss of decoration when same long decorations
in 2 voices of the same staff
(reported by Hudson Lacerda)
Do the overlay voice inherit of the note map (%%voicemap)
(reported by Gerhard Schacherl)
Crash when more than 16 lyric lines (w:)
(reported by Eric Reinbold)
Accept %%format in tune headers
(reported by Timm Reasbeck)

---- Version 8.11.6 - 2016-05-30 ----

%%rbdbstop added
(asked by Hudson Lacerda)
Stop back again the repeat brackets on ||, [| and |]
(asked by Bernhard Weichel)
Don't move the symbols of an empty staff if the staff is displayed
elsewhere in the music line
Don't shift whole notes on voice overlap
%%graceword added
(asked by Manavasu)
Fix clash of notes of 1st and 3rd voices when on a same staff
(reported by David Lacroix)

---- Version 8.11.5 - 16/05/01 ----

Adjust the endpoints of the glissendo
Fix loss of tie on new repeat when previous note is tied
Fix bad end of repeat bracket after %%score with change of voice role
Fix bad colors when %%voicecolor and %%voicescale
(reported by Guido Gonzato)
Fix array overflow in draw_keysig()
(reported by Hanno Böck)
New command line option '-p'
(patch from Stephen Beitzel)

---- Version 8.11.4 - 16/04/10 ----

Fix bad glissendo
Fix bad generation of "::" when at end of source line since 8.10.1
(reported by Rob Lindauer)

---- Version 8.11.3 - 16/04/09 ----

Fix loss of measure bars when %%score with less staves
further in the music line
Set back the decoration function '4' and move the glissendo under '8'
(reported by Gerhard Schacherl)
Don't scale the decorations on grace notes
(reported by Manavasu)
Fix bad position of P: when on a combined rest
(reported by Atte Jensen)
Fix bad key signature when defined between two measure bars
(reported by Stu)
Change the PS squiggly glissando to the same as the SVG one
(reported by Manavasu)

---- Version 8.11.2 - 16/03/23 ----

Fix bad length of ties when at end of line and notes in an other voice
(reported by David Lacroix)
Fix crash when long decoration without any note/rest in the music line
Fix bad glissendos when notes with accidental
(reported by Manavasu)
Add source annotations for multi-rests
(asked by Jan Wybren de Jong)

---- Version 8.11.1 - 16/02/25 ----

Decorations !-(!, !-)!, !~(! and !~)! added
(asked by Manavasu)
Fix bad ties on right repeat
Fix bad tie when over dotted bar
(reported by David Lacroix)
Center the measure repeat signs
(reported by David Lacroix)
Better vertical offset of rests when many voices (again!)
(reported by David Lacroix)
Handle PS "dlw" definition for SVG output
(reported by David Lacroix)

---- Version 8.11.0 - 16/02/05 ----

Better vertical offset of rests when many voices
(reported by David Lacroix)
Fix bad setting of grace note spacing
Permit ties from grace notes to chord
(asked by Hudson Lacerda)
Erroneous meter on many staves when last staff empty
(reported by Hudson Lacerda)
Close the last repeat bracket at end of tune
(asked by Hudson Lacerda)
Set %%cancelkey by default !! compatibility !!

---- Version 8.10.1 - 16/01/18 ----

On K:, put the natural accidentals after the other ones
when less sharps or flats
(reported by Nils Liberg)
!rbend! added
Better repeat brackets
(reported by Nils Liberg)

---- Version 8.10.0 - 16/01/09 ----

Change the length of the repeat brackets
(asked by ABC people)
Fix bad length unit when M:none in tune header and no L:
(reported by David Lacroix)
Add %%rbmax and %%rbmin
(asked by ABC people)
Fix bad offset of some decorations on grace notes with beam
Use the POSIX C extended regular expressions for selection

---- Version 8.9.4 - 15/12/21 ----

Extend the repeat bars on next line up to 4 measures instead of 2
(asked by Rickard Blixt)
Fix bad line width in user decorations with SVG output since 8.9.1
(reported by David Lacroix)
Fix bad microtone accidentals after transposition
(reported by Hudson Lacerda)
Fix bad accidental of note tied from inflected note after measure bar
and transposition
(reported by Hudson Lacerda)
Fix lack of transposition of guitar chords after tabulation

---- Version 8.9.3 - 15/11/22 ----

Fix bad length of long decorations when %%staffscale or %%voicescale
(reported by Gerhard Schacherl)
Fix bad heads when %%voicecombine and head decorations
(reported by Johann Schatzer)
%%voicecombine may be global
Fix bad handling of "$n" in lyrics
(reported by Gerhard Schacherl)
Don't draw a line above a measure bar up to the upper staff
when this upper staff has no bar
Draw the measure bars a bit thicker
(asked by Guido Gonzato)

---- Version 8.9.2 - 15/10/26 ----

Fix bad display of double repeat bars when crossing 2 music lines
since 8.8.2
(reported by Jean-Luc Zins)
Fix bad measure duration since 8.9.1
Fix bad slur when ending on first symbol from the previous music line
and in a second voice
(reported by Frédéric Boulanger and Jean-Luc Zins)
Update the INSTALL documentation
Fix bad scale with EPS output
(reported by Martin Tarenskeen)
Fix error in SVG when non-default stroke width and voice color
(reported by David Lacroix)
Fix bad setting of the configure variables when done in the command line
(reported by Martin Tarenskeen)
Change the default meter from 4/4 to none when M: is absent,
in accordance with the standard
Set the duration of a measure to a non null value when M:none
and avoid strange rendering of multi-rests ("Z")
(reported by Jakob Englhauser)

---- Version 8.9.1 - 15/10/15 ----

Fix bad height of fingering in groups of grace notes
Set back the PS out in 72 PPI
('gv' problem reported by Jean-Luc Zins)
Fix crash when some cases of voice unison
(reported by Jean-Luc Zins)
Fix stroke width in SVG images
(after tests with Larry Myerscough)
Fix bad conversion of "\oe" into œ
(reported by Guido Gonzato)

---- Version 8.9.0 - 15/10/10 ----

%%pagescale added
Change PPI from 72 to 96
Fix abnormal line to the circle-x note head with PS output
(reported by Atte Jensen)
Fix bad staff of slur continuation in secondary voices when many staves
(reported by Jean-Luc Zins)
Fix too much space after accidental when unison
(reported by David Lacroix)

---- Version 8.8.5 - 15/09/24 ----

Handle "\*" as the character '*' in lyrics (w:)
(reported by John Chambers)
Adjust the vertical room of rests
Fix clash of rests when more than 2 voices per staff
(reported by David Lacroix)
Fix bad head of top note when !head-xx! decoration on chord
(reported by David Lacroix)
Fix bad tie when clef change just later
(reported by Jean-Luc Zins)
Fix double accidental when unison
(reported by Jean-Luc Zins)
Fix bad string width computation when XML character references and SVG output
Add %%decoerr
Change '%%dynamic'... to '%%pos dyn'...
Add '%abc-2.1' as the first line in chinese.abc
(reported by Seymour Shlien)
Fix bad slur when between chord and non chord

---- Version 8.8.4 - 15/08/24 ----

Fix loss of tuplet when starting with grace notes
(reported by Lou Montulli)
Fix clash of accidentals when voice overlap
(reported by David Lacroix)
Fix lack of !head-xx! decorations with U: letters on single notes

---- Version 8.8.3 - 15/07/13 ----

Fix bad horizontal offset of tremolos when the notes are shifted
(reported by David Lacroix)
Fix bad accidentals in overlaid voices when %%transpose
(reported by David Lacroix)
Give more information about errors in %%transpose
Set smaller vertical height of tuplets
Fix double tuplet number when nested tuplets at start of a secundary voice
(reported by David Lacroix)

---- Version 8.8.2 - 15/06/17 ----

Fix bad output (PS) or crash (SVG) when the last symbol of a secondary voice
is alone in a music line
Fix crash when special glyph and no <defs> on SVG output
Adjust the '8's in octava clefs
(reported by David Lacroix)

---- Version 8.8.1 - 15/06/14 ----

Fix bad start of ties in second repeats at start of line
Fix lack of ties in second repeats at start of line when many voices
(reported by Gerhard Schacherl)
Fix offset of note dots when scaled staff or voice
Fix offset of '8' in octava clefs
Fix loss of line break when occuring just at minimal spacing
Fix bad offset of some decorations with %%deco function 3 since 8.8.0
(reported by David Lacroix and Gerhard Schacherl)

---- Version 8.8.0 - 15/05/21 ----

Fix again bad handling of %%ornament with %%deco function 3
(reported by Lou Montulli)
Extend the syntax of %%map
(asked by David Webber)
Replace "all,<any_note>" by "*" in %%map
Fix crash when PS output with SVG glyph definition and no "stroke-width:"
Add "%%combinevoices 3"
(asked by Alan Snyder)
Fix clash of accidentals in chords with unisons
(reported by David Lacroix)

---- Version 8.7.8 - 15/05/15 ----

Fix crash when %%score/%%staves before generation restart
Fix lack of stem in sample4.abc
Fix clash of accidentals in complex chords
(reported by Hudson Lacerda)
Fix music line too wide when line split on a beamed sequence
Fix crash when trying to cut a music line at end of tune
(reported by Hudson Lacerda)
Fix bad offset of tuplets and slurs when voice appearing
in a %%score/%%stave sequence
Fix SVG path to PS translation (in %%beginsvg sequence)
Accept XML character references and predefined entities
("&xxx;") in any string

---- Version 8.7.7 - 15/04/29 ----

Fix abnormal key space when many %%transpose in tune body and K:none
(reported by David Lacroix)
Fix bad linewidth when set by PS sequence since 8.7.4
(reported by David Lacroix)
Accept many invisible notes under beams
Set the number alone in tuplet closer to beam
(reported by Alan Snyder)

---- Version 8.7.6 - 15/04/20 ----

Fix bad SVG definition of !tenuto! when SVG output
(reported by David Lacroix)

---- Version 8.7.5 - 15/04/19 ----

Change some decorations of deco.abc
Fix annotation-like offset of decorations with function 0 (near notes)
Added files free.abc and bravura.abc
Accept (some) SVG paths by <defs>..</defs> for PS output
Accept glyphs by <defs>..</defs> for SVG output
Fix bad left margin when %%leftmargin inside tune body since 8.7.4
(reported by David Lacroix)
Don't adjust the pitch of notes when the key is 'P' (percussion)
(reported by Willem Vree)
Fix bad output when internal output buffer overflow
(reported by Jan Wybren de Jong)
Fix bad unison display according to %%shiftunison value
(reported by David Lacroix)
Fix bad vertical offset of "@0,0xx" guitar chords when after other
guitar chord/annotation on the same note since 7.1.1
(reported by Hudson Lacerda)
Fix abnormal key space when %%transpose in tune header before K:none
(reported by David Lacroix)
Don't set the %%voicecolor to the clef, key and time signatures
(asked by David Lacroix)

---- Version 8.7.4 - 15/04/09 ----

Fix "buffer overflow" messages with big tunes
(reported by Frédéric Boulanger).
Fix win32 compilation problems
(reported by Jean-Luc Zins)
Fix 'key,' in %%map which does not depend on the key mode
(reported by Hudson Lacerda).
Fix bad head color when %%voicecolor since 8.7.0
(reported by Hudson Lacerda).
Fix crash on %%voice since 8.7.0
(reported by Hudson Lacerda).
Better handling of %%splittune odd / even.
Fix bad tie direction after voice combine.

---- Version 8.7.3 - 15/04/05 ----

'all,' added in %%map.
%%voicecombine added.
Fix %%tune not working since 8.7.0
(reported by Hudson Flavio Meneses Lacerda).

---- Version 8.7.2 - 15/03/31 ----

Fix loss of measure bar after [I:setbarnb x] since 8.7.0.
Fix warning in accordion.abc
(reported by Matthias Rustler).
Fix bad PS/SVG output with some decorations since 8.7.0.
Fix bad note octave when note mapping and clef change
(reported by David Lacroix).
Extend ps_func of %%deco with different heads for stem up/down.
Extend %%map with different heads for stem up/down.

---- Version 8.7.1 - 15/03/28 ----

Fix bad directions/positions since 8.7.0
(reported by David Lacroix).

---- Version 8.7.0 - 15/03/27 ----

Fix crash when new music line occuring on a voice out of %%score/%%staves
(reported by Frédéric Boulanger).
Fix loop when no end of chord.
K:P added.
map= in V:/K: added for compatibility.
%%map and %%voicemap added.
Fix loss of text in %%begintext sequences after '%'.
Fix bad position of key signatures when declared before a clef
or a measure bar.
Fix bad ties when %%combinevoices
(reported by David Lacroix).
Fix bad transposition of K:none since 8.5.1
(reported by David Lacroix).

---- Version 8.6.2 - 15/03/12 ----

Fix bad handling of %%splittune odd / even
(reported by Hudson Flavio Meneses Lacerda).
Revert the changes in 8.6.1 about the '-0' and '-1' behaviour.

---- Version 8.6.1 - 15/03/10 ----

Fix bad staff offset of some decorations
(reported by David Lacroix).
Fix some little errors in SVG output.
Fix bad colors on SVG output with custumized decorations
(reported by David Lacroix).
Fix command line option '-1' and set oneperpage to 'true'
Fix command line option '-0' and set splittune to 'true'
(reported by Hudson Flavio Meneses Lacerda).

---- Version 8.6.0 - 15/03/06 ----

Fix bad !trem2! since 8.5.2.
Reduce the width of dotted unisons with note on a staff line
(asked by Graham Alsop).
Extend %%splittune with 'odd' / 'even'
(asked by Mike Moroney and Simon Wascher).
Accept annotation-like decorations on 'near the note' and chord decorations.
Fix bad handling of %%ornament with %%deco function 3.
Fix multi display of P: when specified in many voices.
Fix bad P:'s when many ones in a music line.
Don't shift a lyric word smaller than 3 characters starting with a digit
(asked by Helmut Eberl for fingering in lyrics)
Fix SVG error when '&', '<' or '>' in T:.
Add %%abc.
Fix bad tie at end of line when tie at end of a grace note sequence
with the same starting and ending notes since 8.5.2
(reported by David Lacroix).
Fix bad position of '8' in octave clefs.
Fix some bugs in the PS interpreter as bad 'translate'/'rotate' operators.
Fix bad measure number offset at start of line.
Fix bad boxes around parts
(reported by Hudson Flavio Meneses Lacerda).
Don't add index elements of titles which contain '--'
(reported by David Lacroix).
Fix bad music line splits giving measure bars at start of line
(reported by David Lacroix).
Fix lack of notes when voice out of its staff and original staff is empty.
Fix some more accidental clashes when voice overlap
(reported by David Lacroix).

---- Version 8.5.2 - 15/01/28 ----

Fix bad display when bad ties to grace notes
(reported by Willem).
Change the default fonts to serif and sans-serif
instead of Times and Helvetica when SVG output.
Fix bad guitar box when many guitar chords on a note.
Fix bad beams when tremolo inside beamed notes
(reported by David Lacroix).
Fix loss of ABC lines after %%abc-include when %%select did not
match the last included tune.
Fix clash of accidentals with stems in some cases of full voice overlap.
Fix chinese.abc for correct rendering with pango.
Fix some compilation warnings
(reported by Daniel Branning).
Fix bad measure number when the bar followed by a clef change since 8.3.5
(reported by Frédéric Boulanger).
Fix loss of first note when the tune starts with %%ps since 8.5.0
(reported by David Lacroix).
Fix bad font-family with SVG output since 8.5.1.

---- Version 8.5.1 - 15/01/05 ----

Fix bad handling of voice overlap when at leat one unison.
Fix bad extra space at end of line when clef change by K:.
Add file glyphs.abc in the package.
Fix loss of font change before non-ASCII characters in strings.
Fix bad offsets and fonts of annotation-like decorations.
Fix loop on %%voice after %%voice since 8.3.2.
Fix erroneous display of rests in empty staves when "%%staffnonote 0".

---- Version 8.5.0 - 15/01/02 ----

Fix bad handling of font-style when no '-' in font name.
Extend %%deco with annotation-like strings.
Extend "%%setfont-x" up to 9.
Add %%voicecolor
(asked by Frédéric Boulanger).
Fix "setgray" in SVG output.
Add some styles (preserve white spaces, svg as block) on XHTML output ('-X').
Remove '<p>' from '-v -O -' output (many SVG images on stdout).
Do smaller shifts when voice overlap
(asked by David Lacroix).
Fix bad unisons when different accidentals
(reported by David Lacroix).
Fix loss of key signatures and meters when voice in empty staff
(reported by Willem).
Fix some cases of music lines cut inside beamed notes.
Adjust the position of the 'crdc' decorations and update deco.abc.
Fix loss of the 'svg' definition used by the PostScript code on SVG output.
Fix again bad display of unisons
(reported by David Lacroix).
Apply %%topspace when %%newpage inside a tune
(reported by Johann Schatzer).

---- Version 8.4.0 - 14/12/16 ----

Remove features.txt from the tarball.
Fix "%%microscale 4" on SVG output.
Handle %%microscale.
Add '+' in %%score/%%staves.
Fix bad handling of "K:none", especially when transposition.
Fix bad note heads when %%combinevoices
(reported by Johann Schatzer).
Fix bad transposition of the latin guitar chord "La".
Fix loss of 'exp none' in K:.
Fix bad display of unisons
(reported by David Lacroix).
Fix some compilation warnings
(reported by John Chambers).
Fix loss of some clef warnings at end of previous lines.

---- Version 8.3.4 - 14/12/02 ----

Fix double clefs when new voices in %%staves
(reported by Francis Stevens).
Fix loss of 128 decorations per tune since 8.3.0
(reported by Henry Bley-Vroman).
Fix problems with decorations or guitar chords/annotations
in sequence/measure repeat.
(reported by Frédéric Boulanger).

---- Version 8.3.3 - 14/11/26 ----

Fix no display of lyrics after tune since 8.3.2
(reported by Gerhard Schacherl).
Fix bad position of grace notes in overlay voices at start of measure
(reported by Jean-Luc Zins).
Fix memory leak relative to decorations.
Set back %%autoclef
(asked by Atte Jensen).
Remove the references to 'format.txt'
(reported by Martin Tarenskeen).
Fix bad repeat font when staves/voices with different scales
(reported by Gerhard Schacherl).

---- Version 8.3.2 - 14/11/01 ----

Change the PS font scale of the accidentals according to the FontMatrix
(reported by Chuck Boody).
Remove format.txt from the tarball.
Add decoration function 'crdc'.
Fix loss of voice options when many %%voice's without "%%voice end" in %%tune.
Fix loss of lyrics when ABC version 2.1 and generation restart.
Accept '<note1><note2>' in %%transpose.
Fix bad note transposition since 8.2.0.

---- Version 8.3.1 - 14/10/14 ----

Fix crash when %%combinevoices on beamed notes since 8.1.5
(reported by David Lacroix).
Fix bad start/stop of ties since 8.1.0.
Fix abnormal error when '<none>' or '<EOL>' in %%linebreak ending with 'lock'.
Fix abnormal error when spaces after backslash in music lines
(reported by Daniel Branning).
Don't let vertical space when all staves are empty.

---- Version 8.3.0 - 14/10/05 ----

Fix program loop when no measure bar at end of line since 8.2.2.
Fix some clashes of rests when more than 2 voices per staff.
Don't reset the staff to the main voice when finding a previous
overlaid voice.
Fix the documentation about %%stretchlast and change its default
value from 0.2 to 0.25.
Handle utf-8 in the names of the notes/history/.. after tune
(reported by Hudson Flávio Meneses Lacerda).
Fix crash when line break before a note alone in a staff
and a new voice since 8.2.0.
Fix crash when %%alignbars since 8.2.0.
Fix some clashes of dots and stems when voice overlap
(reported by Hudson Flávio Meneses Lacerda a long time ago!).
Fix bad PS output when exact minimum spacing since 8.2.1.
Fix crash when tune starting with grace note in the middle voice
of a 3-voices staff.
Fix crash when many %%staves/%%score and new pages
(reported by David Lacroix).
Don't print the lyrics after tune when "%%writefields W n".
Fix loss of explicit clefs since 8.2.0.
Fix M$ compiler errors
(reported by Olivier Levon).
Simplify the ABC parser.

---- Version 8.2.2 - 14/09/12 ----

Fix crash when slur of second voice continued on next line.
Fix bad auto clef after generation restart when clefs inserted.
Split the music lines at better places.
Fix many errors due to the lack of clefs in secondary voices.
Fix abnormal 'Bad tie' errors when tie on 2 music lines
(reported by Olivier Levon).
Fix compilation problems with VS Studio 2013 Express
(reported by Olivier Levon).
Fix generation abort when lack of V: after %%staves/%%score
(reported by Olivier Levon).

---- Version 8.2.1 - 14/09/03 ----

Fix crash when forced clef after %%staves/%%score since 8.2.0
(reported by Francis Stevens and Olivier Levon).

---- Version 8.2.0 - 14/08/29 ----

Fix bad clefs when voice/staff changes in %%staves/%%score.
Remove %%autoclef which was not used.
Fix loss of indentation since 8.1.0
(reported by David Lacroix).
Change the internal treatment of %%score/%%staves for no extra space.
Add !marcato! and !^!
(asked in the ABC mailing list).
Fix bad stem direction in lowest voice when invisible and visible rests
at a same time in a measure since 8.1.0
(reported by Eric Reinbold).
Set lowest value of %%staffscale to 0.3
(asked by Mike Scott).
Accept '^' and '$' as beginning and end of line in tune select.
Fix crash when multi-rest at start of a second voice of a staff.

---- Version 8.1.5 - 14/08/09 ----

Fix crash when %%combinevoices and different beaming.
Fix lack of ledger lines when large chord.
Always remove the invisible rests at start of tune when L:auto.
Add !stemless!.
Fix abnormal error 'Bad tie' when tie at end of voice
(reported by Daniel Branning).
Fix some compilation warnings
(reported by Daniel Branning).
Fix bad handling of !()! since 8.1.4
(reported by David Lacroix).
Fix horizontal offset of the custodes
(asked by Stephen West).
Fix crash when %%continueall and no measure bar in the tune.
Fix bad horizontal offset of rests alone in a measure
(reported by Stephen West).

---- Version 8.1.4 - 14/07/17 ----

Fix lack of source last character in syntax errors.
Change the definition of the start of long decorations for correct
display when the ending decoration is not in the tune.
Set the extra accidentals of K: at their indicated pitches
(asked by Paul Morrisset).
Fix double interpretation of '.' in "!p!.(c"
(reported by David Lacroix).
Fix lack of vertical space when empty lines in %%begintext
(reported by Henry Bley-Vroman).
Fix bad PostScript output when %%begintext justify.
Output the PDF marks of all T:'s, even if the first one is empty
(asked by David Lacroix).
Fix the size of font of the page header/footer, broken in 8.1.3
(reported by Francis Stevens).

---- Version 8.1.3 - 14/07/01 ----

Add more information about the titles/subtitles in the PostScript
output for (external) index generation
(asked by Tim Macdonald).
Add Δ (delta) in the known glyphs
(asked by Chuck Boody).
Fix some bad 'Line overfull' messages when automatic line break.
Fix bad clef when L:auto and quick clef change at start of tune.
Invert the behaviour of '\' in the guitar chords
(asked by David Lacroix).
Fix bad interval between staff systems according to %%staffsep
(reported by Eric Reinbold and David Lacroix).
Change <p> to <br/> between images on SVG output.
Fix bad %%header/%%footer vertical offsets.
Don't put the last music line on a new page when page overflow on W:.

---- Version 8.1.2 - 14/06/20 ----

Fix bad vertical offset of dots sometimes when many voices per staff.
Don't get 'auto' alone as 'clef=auto'
(reported by David Lacroix).
Fix double clef at start of staff when many voices in the staff
since 8.1.0.
(reported by David Lacroix).

---- Version 8.1.1 - 14/06/19 ----

Fix crash when a voice with no note appears as second voice of a staff
since 8.1.0
(reported by Francis Stevens).

---- Version 8.1.0 - 14/06/19 ----

Split the tune into music lines at better places.
Fix clash of voice name with staff when new voice after %%staves.
Fix bad music line indentation when more staves further in the tune.
Set the ties closer to the notes.
Fix bad automatic clef changes.
Add '%%clef auto'.
Change internal treatment of clefs.
Fix clash of notes in chords when unison or second with long notes.
Fix bad horizontal offset of accidents on long notes.
Fix bad key signature after change to 'K:none'.
Fix bad handling of '|' when the first character in w:
(reported by Alex Scheutzow).
Change internal treatment of #lines and scale of the staves.
Fix bad clef/note offset when a voice with explicit clef moves
to an other staff
(reported by  reinier maliepaard).
Fix bad font for !trill!
(reported by David Lacroix).
Check the floating voices in %%score/%%staves.
Fix bad stem direction when %%combinevoices.
Fix crash when unknown decoration on one character since 8.0.4.
Fix crash when measure bar numbering and less notes in the upper staff
(reported by : J.Joao Almeida).
Add 'cue=' in K: and V:
(asked by Alex Scheutzow).
Fix too big length of syntax error messages
(reported by David Lacroix).
Fix bad horizontal offset of dots in chords since 8.0.4.
Fix bad horizontal offset of full rests in voice overlay
(reported by J.C.L.).

---- Version 8.0.4 - 14/05/23 ----

Fix too wide space between text paragraphs when not fill or justify.
Fix crash when utf-8 and latin characters in a same file
(reported by Henry Bley-Vroman).
Fix note/accidental clash when chords with unisons and seconds
(reported by David Lacroix).
Change internal fonts to 'serif' for SVG output.
Forbid rests in grace note sequences.
Fix crash when rest in grace note sequence.
Fix direction of ties in chords with odd number of notes.
Fix bad beams in SVG output
(reported by David Lacroix).
Fix bad direction of beam stub on last note when !beambr1! / !beambr2!
(reported by David Lacroix).
May have up to 128 decorations per tune
(asked by Henry Bley-Vroman).
Fix crash when only P: or Q: in a generation sequence
(reported by Henry Bley-Vroman).

---- Version 8.0.3 - 14/05/05 ----

Fix non-function of %%voice inside %%tune since some 7.5.x
(reported by Gerhard Schacherl).
Fix small internal problem with decorations.
Fix position of accent marks (!>!)
(reported by Paul Rosen).
Fix bad margins when tune inside %%multicol
(reported by A.B. Steen).
Don't draw repeat brackets when bar between 2 staves.
Fix compilation warning on MS-windows
(reported/fixed by Olivier Levon and Jean-Luc Zins).
Don't display an accidental when {#,b,=} at start of a word in guitar chords.
Fix lack of EOS in some internal strings
(fixed by Olivier Levon).
Fix bad guitar chord / annotation when '\' inside.
Fix double transposition of 2 octaves when both %%abc2pscompat
and octave= in K:/V:.
Fix compilation warnings on MAC
(reported by Chuck Boody).

---- Version 8.0.2 - 14/04/02 ----

Fix bad text justication with PostScript output
(reported by Gerhard Schacherl).
Fix loss of positions (decorations, lyrics..) when many tunes
(reported by John Taylor).
Add HAVE_MMAP in config.h for ms-windows users
(reported by Jean-Luc Zins).
Fix loss of tie at end of line
(reported by David Lacroix).
Add '--param=value' in 'configure'.
Fix bad configure when full path to the cc compiler
(reported by Olivier Levon).

---- Version 8.0.1 - 14/03/31 ----

Fix bad configure when pango is found (Makefile empty)
(reported by Martin Tarenskeen).

---- Version 8.0.0 - 14/03/29 ----

Remove some vertical space in tune header when no composer.
Remove the vertical space of the empty staves.
Remove the GNU 'autoconf' stuff (see 'INSTALL').
Remove the License.
Change the format of the error messages.
Remove tight.fmt.
Add sample8.html.
Add command line option '-z' for ABC embedded in HTML/XHTML.
Fix loss of vertical space above the staves (%%staffsep) since 7.6.10
(reported by David Lacroix).
Creation from abcm2ps version 7.8.0.

(wiz)

Updated devel/afl to 2.35b

(wiz)

Updated afl to 2.35b.

--------------
Version 2.35b:
--------------

  - Fixed a minor cmdline reporting glitch, spotted by Leo Barnes.

  - Fixed a silly bug in libdislocator. Spotted by Johannes Schultz.

(wiz)

Updated print/ghostscript-agpl to 9.20

(wiz)

Updated ghostscript-agpl to 9.20.

Remove Makefile.common that is not used by anything, despite
a comment to the contrary.

Highlights in this release include:

    The usual round of bug fixes, compatibility changes, and
    incremental improvements.

For a list of open issues, or to report problems, please visit
bugs.ghostscript.com.  Incompatible changes

    The planned device API tidy did not happen for this release,
    due to time pressures, but we still intend to undertake the
    following: We plan to somewhat tidy up the device API. We intend
    to remove deprecated device procs (methods/function pointers)
    and change the device API so every device proc takes a graphics
    state parameter (rather than the current scheme where only a
    very few procs take an imager state parameter). This should
    serve as notice to anyone maintaining a Ghostscript device
    outside the canonical source tree that you may (probably will)
    need to update your device(s) when these changes happen. Devices
    using only the non-deprecated procs should be trivial to update.

(wiz)

Updated graphics/asymptote to 2.38

(wiz)

Updated asymptote to 2.38.

Release Notes for Version 2.38

An integer division operator # was added. Control points in xasy are now
correctly parsed. Longitudinal splitting in the revolution structure of
the solids module was fixed. Portability fixes were implemented.
The ncurses library is now only required with --enable-readline.
A --disable-sigsegv configuration option was added.

Release Notes for Version 2.37

The xasy graphical user interface now runs under both Python 2.7 and Python 3.
Legacy versions (prior to 9.14) of Ghostscript can be supported by
assigning settings.epsdriver="epswrite" (or by setting the environment
variable ASYMPTOTE_EPSDRIVER to epswrite). The quiet flag suppresses
noninteractive standard output when settings.verbosity <= 1. A progress
function was added to the plain_strings module. The smoothcontour3 module
was optimized to use Bezier triangles where appropriate, along with a
built-in least-squares routine and an improved root finder based on
quadratic interpolation. If settings.sysdir is empty, preference is given
to a version of kpsewhich in the same directory as the executable for
determining the correct sysdir. The handling of degenerate normals of
Bezier triangles was fixed. Bugs in forking, integer formatting, dash
adjustment, subpaths, and guide reversion were fixed. Version 1.30 of
asymptote.sty (auto-generated) and version 0.35 (or later) of media9.sty
are now required.

Release Notes for Version 2.36

Bezier triangle patches have been implemented in place of degenerate Bezier
tensor product patches. Surface rendering was improved. The configuration
of the readline and gc libraries was fixed. The asy configuration
directory is only created if localhistory=false. Patches are now sorted by
projected distance. Animations were fixed by running LaTeX twice.
The asy-mode.el headers were updated. Intermittent segmentation faults and
floating point exceptions in the OpenGL renderer were fixed. Support for
GSL 2.0 was added. A quite nan constant was added. Straight segments are no
longer split in bezulate. Segmentation faults in tab completion were fixed.
A work around for a clang 3.7.0 compiler bug was implemented.
The smoothcontour routine was sped up. Several bugs in the file read routines
were fixed. A bug in rest argument signature equivalence was fixed.
Threads are no longer used in batch mode, except under MacOS X. A convenience
function graphicscale was added for using graphic with the conTeXt tex
engine. The splinetype detection for Spline surfaces was fixed.

Release Notes for Version 2.35

A work around was implemented for a ghostscript eps2write bug that forces
all postscript to the first page, breaking multiple 3D XeLaTeX and ConTeXt
labels.

Release Notes for Version 2.34

The readability of named pen colors was improved in the documentation.
A surface cone(path3 base, triple vertex) routine was added for
constructing an approximate cone over an arbitrary base. A test for
Ghostscript 9.14 or later was added to the build process. The documentation
was updated. A CYGWIN warning message under Windows 8 was fixed.

Release Notes for Version 2.33

A work around was implemented for the missing epswrite driver in
ghostscript-9.15. Deconstruction is now always done in the C locale.
A work around for a unit change in dvisvgm-1.5.3 was implemented.
The path arc(pair B, pair A, pair C, real r) function was fixed.
The handling of the boolean condition in parametric surfaces was fixed.
The default meshlight was changed to nolight so that mesh lines with
positive width appear consistent with the default mesh lines.
A nonsquare image dimension error was fixed. The definition of the
SimpleHead arrowhead was fixed. The zoom/menu button and play option were
fixed. An intersect(path, surface) function was implemented.
A smoothcontour3 module written by Charles Staats and leminiscate example
were added. The inline asymptote.sty option now works with xelatex.
An obsolete workaround for an Adobe Reader transparency artifact
was removed. An asylatexdir option was added to support the pdflatex
-output-directory option. An aligndir option for aligning the picture to
an arbitrary point of the page boundary was added. The garbage collector was
updated to gc-7.4.2. The documentation was updated.

Release Notes for Version 2.32

The libc++ stringstream workaround was also enabled for FreeBSD.
The segment(bool[] b) function was fixed. The side(pair,pair,pair) function
was renamed to orient(pair,pair,pair) and an error in its documentation was
corrected. New functions orient(triple,triple,triple,triple) and
insphere(triple,triple,triple,triple,triple) were implemented. A random
number generator incompatibility on some platforms was fixed. Support was
removed for the obsolete utility texi2dvi4a2ps. Compiler warnings were
suppressed.

Release Notes for Version 2.31

Hangs in 3D font generation and also in the "none" tex engine were fixed.

Release Notes for Version 2.30

Compilation issues were addressed. A workaround for the broken stringstream
container in MacOS 10.9 libc++ was implemented. The OpenGL zoom/menu button
was fixed.

Release Notes for Version 2.29

The TeX bidirectional pipe was overhauled to support the context tex engine
again. The luatex and lualatex tex engines were enabled. The inline option
used by the asymptote.sty LaTeX package and the inlineimage option used for
generating external PRC files were fixed. Portability issues were addressed.

Release Notes for Version 2.28

A locale bug that interfered with the 3D PRC camera transformation was fixed.
Minimum OpenGL window constraints were removed in favour of the viewportsize
variable. The transform(u,v,O) function, which projects onto the plane
spanned by u and v through point O, was fixed. Numerical overflow issues in
quadraticroots and cubicroots were fixed. The documentation was updated.

Release Notes for Version 2.27

Move Adobe transparency workaround to C++ code to allow use of texpreamble
again with the pdflatex tex engine.

Release Notes for Version 2.26

The xasy graphical user interface now terminates the asy process on exit.
The xasy code editor under MSWindows was fixed; the default code editor is
now winpad. Degenerate HookHead and SimpleHead arrows were fixed.
Portability issues were addressed.

Release Notes for Version 2.25

A superfluous play button in rendered 3D images embedded by recent versions of
media9 is now suppressed. The contour.asy module was reverted to a previous
stable linearized version. A numerical precision issue in makepen was fixed.
A routine for drawing braces was added. Deep recursion is now avoided in
guide flattening. A workaround for an Adobe Reader transparency artifact
was implemented for the pdflatex and xelatex tex engines. Raw PRC output
can now be generated with the "-f prc" command-line option. Vector patches
are now sorted to work around opacity artifacts in many rendering engines.
The xasy code editor now accepts command-line options. Under MSWindows,
the ghostscript library is searched for in both the 32 bit and 64 bit
registries. The FAQ and documentation were updated.

Release Notes for Version 2.24

A segmentation fault in drawSphere was fixed. Recursive calls to simpson
are now supported. The explicit libglapi dependency was removed.
A latexmkrc example file that shows how to store figures in a subdirectory
is now included.

Release Notes for Version 2.23

Compilation without the FFTW library is now supported again.

Release Notes for Version 2.22

Self-defined unary operators are now allowed. Formatted strings instead of
real values are compared in OmitFormat. The segment(bool[]) function was
rewritten to use the more efficient segmentlimits(bool[]) call.
Unnecessary buffering of surface and path3 data was removed.
Portability tweaks were made.  References to out-of-date trembling
examples were removed. Vertex-colored triangles now work again in Adobe XI.
The transformation of normal vectors was fixed.  PostScript extend
qualifiers were added for axial and radial shading. The TEXMFMAN
environment variable is now used to find the TeXLive sysdir.

Release Notes for Version 2.21

Explicitly transformed billboard labels now work correctly again.
The alignment point of OpenGL billboard labels was fixed. An extend
parameter was added to the axes (default true) and axes3 (default false)
routines. A history recall bug was fixed. A typo was corrected in the
documentation of the Jacobi elliptic functions sncndn.

Release Notes for Version 2.20

A work around was implemented for a dvipdfmx bug that prevents the
xelatex tex engine from properly embedding PRC objects. Billboard rotation
is now disabled for explicitly transformed labels.

Release Notes for Version 2.19

Numerical resolution issues with the PRC camera orientation and viewportshift
were fixed. The lighting of NURBS surfaces was fixed. The special 8192
strlen NSIS build was now correctly reinstated, with stubs, to prevent the
installer from overwriting Windows PATH environment variables > 1023 bytes.

Release Notes for Version 2.18

A compilation issue on MacOSX was addressed. Secondary axes pictures now
inherit the size of the primary picture, so that the markthin marker works
properly. The special 8192 strlen NSIS build was reinstated to prevent the
installer from overwriting extremely long Windows PATH environment variables.

Release Notes for Version 2.17

A bug with non-square pen function images was fixed.  Autoscaled
logarithmic axes were fixed. Offscreen and non-offscreen rendering are now
supported in a single binary (requiring OSMesa version 8), with
settings.offscreen defaulting to false. The media9 LaTeX style file is now
used to embed 3D PRC content instead of movie15. Local 3D coordinates are
now used. PRC Part names are nolonger generated by default. A bug in
bezulate was fixed. A settings.axes3 flag was added to control the
visibility of PRC axes. An efficient 3D routine for drawing many
triangles, with specified vertices and optional normals or vertex colors,
was implemented.

Release Notes for Version 2.16

Ticks are no longer autoscaled when the number of major intervals is
specified and autoscale is false. Manual tick scaling was fixed. A bug in
the palette range was fixed. A division by zero in constructing curved arrows
was fixed. A numerical underflow was fixed. A picture bound error was
fixed. The current value of currentpen is now always respected in default
arguments. A default viewportwidth is no longer imposed for attached
images. A routine for computing camera positions was added. The format
command is now more consistent with C++ printf formatting. Named arguments
can now appear in function calls after rest arguments. The wheel example
was improved to support PDF animations. The erase command no longer resets
the machine state. Pipes are now used for xasy communication. A new mode
parameter to input and output replaces xinput, xoutput, binput, and
boutput. The icon directory path for 64-bit MSWindows systems was fixed.
Compilation of native CYGWIN binaries is now supported.

(wiz)

+ asymptote-2.38.

(wiz)

Updated print/chktex to 1.7.6

(wiz)

Updated chktex to 1.7.6.

1.7.6 (2016-09-09), released by Ivan Andrus <darthandrus@gmail.com>

  * Properly copy and free strings.  This caused incorrect file
      reporting in some cases #49000
  * Add "verbatim" commands and environments from tipa #46734
  * Avoid infinite loop when a regex matches an empty string
  * Fix crash on invalid regular expression
  * Add warnings for $$...$$ and $...$ #47044
  * Update the banner to reflect current bug reporting contacts

1.7.5 (2015-12-07), released by Ivan Andrus <darthandrus@gmail.com>

  * Track whether the previous line ended in a comment #45980
  * Add a few source code environments to VerbEnvir
  * Don't show long line warning if the last line doesn't have a newline #46539
  * Fix deweb's handling of file names #46043,#46042
  * Use the exit status to indicate warnings were found #45979
  * Consider OS X a UNIX

1.7.4 (2015-04-11), released by Ivan Andrus <darthandrus@gmail.com>

  * Fixed undefined behavior (patch from TeXLive)

1.7.3 (2015-04-09), released by Ivan Andrus <darthandrus@gmail.com>

  * Fixed memory leaks and a crashing bug #42813
  * Added ConTeXt \start... \stop... tracking #44517
  * Added DashExcpt list for exceptions to dash rules #42237
  * Support \cite with two optional arguments (biblatex) #44413
  * Removed warning about using \dots instead of .. #40731
      This incorrectly warns about relative paths
  * Removed check for intro since it's common in file names #43013
  * Turn off warning 19 by default #44226
      Warning 19 is about Latin-1 quotes, but gives false positives
      for some utf-8 characters.

1.7.2 (2013-04-17), released by Ivan Andrus <darthandrus@gmail.com>

  * Fixed display bug on platforms where sizeof(int) != sizeof(regoff_t)
  * Warn that .. should be \dots like lacheck does
  * Fixed warnings under clang
  * Incorporated several build related patches from TeX Live

1.7.1 (2012-08-22), released by Ivan Andrus <darthandrus@gmail.com>

  * Improvements for TeX Live and CTAN
  * Fixed bug when checking for interword spacing
  * Added warning when a file has lines that are too long which can cause problems
  * Added ability to suppress user regular expression warnings by "naming" them
  * Added ability to suppress warnings for the rest of the file
  * Allow using PCRE or POSIX ERE or neither (at configure time).
      Regular expressions can be marked as PCRE or POSIX only.
  * Allow () to specify an optional argument in WipeArg to support \cmidrule[](){}
  * Allow regular expressions for Silent keyword, specified in [] for "case insensitive"
  * Added warning about \hline and vertical rules in tabular environments
  * I changed user warnings to be post WipeArg (like they were pre
      1.7.0).  I had changed them in 1.7.0 to be pre WipeArg so that I
      could match things like \label{...}, but this caused many false
      positives by matching in comments, etc.

1.7.0 (2012-05-28), released by Ivan Andrus <darthandrus@gmail.com>

  * Added support for user defined regular expression based warnings
  * Added support for macros like \ensuremath and \text
  * Added support for suppressing messages on a single line

1.6.6 (2012-04-07), released by Ivan Andrus <darthandrus@gmail.com>

  * Fixed some memory issues
  * Do not increase line number when the line was very long
  * Some fixes for TexLive

1.6.5 (??), released by Jean-Marc Lasgouttes <lasgouttes@lyx.org>

  * Fix bug where ChkTeX generates a warning about missing space before
    parenthesis when the parenthesis is at start of line.
  * Update ChkTeX.tex.in to use proper LaTeX2e markup.
  * Update the build machinery a bit.

(wiz)

Updated devel/liblangtag to 0.6.1

(wiz)

Updated liblangtag to 0.6.1.

0.6.0 -> 0.6.1
=================
Akira TAGOH (2):
      Fix the build issue regarding to 'index' (fixes: issue#15)
      Get rid of the arch-dependent code from lt-config.h

David Tardon (1):
      enum must be declared before it's used

(wiz)

Updated devel/global to 6.5.5

(wiz)

Updated global to 6.5.5.

Version 6.5.5 - Sep 21 2016

This is a bug fix release.

[FIXED BUGS]
o exuberant-ctags.la: The path name of ctags command couldn't changed.
  Now, it can be changed by configuration variable 'ctagscom'.
o Tag files got corrupted when the built-in C++ parser gave up parsing and aborted.
  Now, it is avoided.

o Some inappropriate error messages was changed.

  Error1: NULL file (file size == 0)

$ cp /dev/null GTAGS
$ global -x main

(GLOBAL-6.5.4) global: /tmp/test/GTAGS seems older format. Please remake tag files.
(GLOBAL-6.5.5) global: /tmp/test/GTAGS seems to be corrupted.

  Error2: Corrupted file

$ echo aaa >GTAGS
$ global -x main

(GLOBAL-6.5.4) global: GTAGS not found.
(GLOBAL-6.5.5) global: /tmp/test/GTAGS seems to be corrupted.

(wiz)

Updated security/py-cryptography to 1.5.1

(wiz)

Updated py-cryptography to 1.5.1.

1.5.1 - 2016-09-22
~~~~~~~~~~~~~~~~~~

* Updated Windows and OS X wheels to be compiled against OpenSSL 1.0.2i.
* Resolved a ``UserWarning`` when used with cffi 1.8.3.
* Fixed a memory leak in name creation with X.509.
* Added a workaround for old versions of setuptools.
* Fixed an issue preventing ``cryptography`` from compiling against
  OpenSSL 1.0.2i.

(wiz)

Updated textproc/p5-Text-Glob to 0.10

(wiz)

Updated p5-Text-Glob to 0.10.

0.10    Wednesday 14th September, 2016
        Added ability to alter regex seperator (patch from Mark Fowler)
        Switch distribution packaging back to ExtUtils::MakeMaker (RT#104876)

(wiz)

Updated textproc/p5-Text-BibTeX to 0.77

(wiz)

Updated p5-Text-BibTeX to 0.77.

0.77 2016-09-20
* Fixes for testing and installing on Darwin (install_name issues).
  Thanks to Nuno "smash" Carvalho for the report and debug help.

(wiz)

Updated devel/p5-Specio to 0.26

(wiz)

Updated p5-Specio to 0.26.

0.26    2016-09-24

- Require Role::Tiny 1.003003. This should fix the test failures some
  CPANTesters reported with this error:

    Can't resolve method "???" overloading "&{}" in package
    "Specio::Constraint::Simple" at Specio::Constraint::Simple->new line 35.

(wiz)

Updated devel/p5-Params-Validate to 1.25

(wiz)

Updated p5-Params-Validate to 1.25.

1.25    2016-09-25

- Switched to Params::Validate.

(wiz)

Updated net/p5-Net-Domain-TLD to 1.75

(wiz)

Updated p5-Net-Domain-TLD to 1.75.

1.75  Sat Sep 17 20:00 2016
        - gtld_new update as per latest IANA db entries

(wiz)

Updated mail/p5-Email-Valid to 1.202

(wiz)

Updated p5-Email-Valid to 1.202.

1.202    2016-09-25 21:48:09-04:00 America/New_York
        - avoid using a new Net::DNS feature, which should restore function
          with old versions of Net::DNS

1.201    2016-09-22 19:21:01-04:00 America/New_York
        - mx check once again allows for A records; a future check may
          allow mx-only checks on purpose

(wiz)

Updated time/p5-DateTime-TimeZone to 2.04

(wiz)

Updated p5-DateTime-TimeZone to 2.04.

2.04    2016-09-26

- Automatically add the latest version of DateTime::Time::Local::Win32 as a
  prereqs on Windows system automatically when building the release.

(wiz)

Updated time/p5-DateTime-Locale to 1.08

(wiz)

Updated p5-DateTime-Locale to 1.08.

1.08    2016-09-23

- Require namespace::autoclean 0.19+.

(wiz)

Updated security/p5-Crypt-ECB to 2.21

(wiz)

Updated p5-Crypt-ECB to 2.21.

v2.21, 23.09.2016
- no code changes to ECB.pm
- ecb.pl -l now prints module versions
- ignoring Serpent in test suite as it is broken on many platforms

(wiz)

Updated devel/p5-Canary-Stability to 2012

(wiz)

Updated p5-Canary-Stability to 2012.

2012  Sun Sep 25 01:45:56 CEST 2016
- another year where the UNFCCC had to adjust their predictions
          in the worse direction.
- do not print the big intro blob in yellow.

(wiz)

Updated misc/p5-Business-ISBN to 3.003

(wiz)

Updated p5-Business-ISBN to 3.003.
Change LICENSE per changelog.

3.003 2016-09-19T22:03:24Z
* Clarify everywhere that I'm using Artistic License 2.0

(wiz)

Updated net/tor to 0.2.8.8

(wiz)

Updated tor to 0.2.8.8.

Changes in version 0.2.8.8 - 2016-09-23
  Tor 0.2.8.8 fixes two crash bugs present in previous versions of the
  0.2.8.x series. Relays running 0.2.8.x should upgrade, as should users
  who select public relays as their bridges.

  o Major bugfixes (crash):
    - Fix a complicated crash bug that could affect Tor clients
      configured to use bridges when replacing a networkstatus consensus
      in which one of their bridges was mentioned. OpenBSD users saw
      more crashes here, but all platforms were potentially affected.
      Fixes bug 20103; bugfix on 0.2.8.2-alpha.

  o Major bugfixes (relay, OOM handler):
    - Fix a timing-dependent assertion failure that could occur when we
      tried to flush from a circuit after having freed its cells because
      of an out-of-memory condition. Fixes bug 20203; bugfix on
      0.2.8.1-alpha. Thanks to "cypherpunks" for help diagnosing
      this one.

  o Minor feature (fallback directories):
    - Remove broken fallbacks from the hard-coded fallback directory
      list. Closes ticket 20190; patch by teor.

  o Minor features (geoip):
    - Update geoip and geoip6 to the September 6 2016 Maxmind GeoLite2
      Country database.

(wiz)

Updated devel/py-flake8 to 3.0.4

(wiz)

Updated py-flake8 to 3.0.4.

2.3.0 - 2015-01-04
------------------

- **Feature**: Add ``--output-file`` option to specify a file to write to
  instead of ``stdout``.

- **Bug** Fix interleaving of output while using multiprocessing
  (`GitLab#17`_)

.. _GitLab#17: https://gitlab.com/pycqa/flake8/issues/17

2.4.0 - 2015-03-07
------------------

- **Bug** Print filenames when using multiprocessing and ``-q`` option.
  (`GitLab#31`_)

- **Bug** Put upper cap on dependencies. The caps for 2.4.0 are:

  - ``pep8 < 1.6`` (Related to `GitLab#35`_)

  - ``mccabe < 0.4``

  - ``pyflakes < 0.9``

  See also `GitLab#32`_

- **Bug** Files excluded in a config file were not being excluded when flake8
  was run from a git hook. (`GitHub#2`_)

- **Improvement** Print warnings for users who are providing mutually
  exclusive options to flake8. (`GitLab#8`_, `GitLab!18`_)

- **Feature** Allow git hook configuration to live in ``.git/config``.
  See the updated `VCS hooks docs`_ for more details. (`GitLab!20`_)

2.4.1 - 2015-05-18
------------------

- **Bug** Do not raise a ``SystemError`` unless there were errors in the
  setuptools command. (`GitLab#39`_, `GitLab!23`_)

- **Bug** Do not verify dependencies of extensions loaded via entry-points.

- **Improvement** Blacklist versions of pep8 we know are broken

2.5.0 - 2015-10-26
------------------

- **Improvement** Raise cap on PyFlakes for Python 3.5 support

- **Improvement** Avoid deprecation warnings when loading extensions
  (`GitLab#59`_, `GitLab#90`_)

- **Improvement** Separate logic to enable "off-by-default" extensions
  (`GitLab#67`_)

- **Bug** Properly parse options to setuptools Flake8 command (`GitLab!41`_)

- **Bug** Fix exceptions when output on stdout is truncated before Flake8
  finishes writing the output (`GitLab#69`_)

- **Bug** Fix error on OS X where Flake8 can no longer acquire or create new
  semaphores (`GitLab#74`_)

2.5.1 - 2015-12-08
------------------

- **Bug** Properly look for ``.flake8`` in current working directory
  (`GitLab#103`_)

- **Bug** Monkey-patch ``pep8.stdin_get_value`` to cache the actual value in
  stdin. This helps plugins relying on the function when run with
  multiprocessing. (`GitLab#105`_, `GitLab#107`_)

2.5.2 - 2016-01-30
------------------

- **Bug** Parse ``output_file`` and ``enable_extensions`` from config files

- **Improvement** Raise upper bound on mccabe plugin to allow for version
  0.4.0

2.5.3 - 2016-02-11
------------------

- **Bug** Actually parse ``output_file`` and ``enable_extensions`` from config
  files

2.5.4 - 2016-02-11
------------------

- **Bug** Missed an attribute rename during the v2.5.3 release.

2.5.5 - 2016-06-14
------------------

- **Bug** Fix setuptools integration when parsing config files

- **Bug** Don't pass the user's config path as the config_file when creating a
  StyleGuide

2.6.0 - 2016-06-15
------------------

- **Requirements Change** Switch to pycodestyle as all future pep8 releases
  will use that package name

- **Improvement** Allow for Windows users on *select* versions of Python to
  use ``--jobs`` and multiprocessing

- **Improvement** Update bounds on McCabe

- **Improvement** Update bounds on PyFlakes and blacklist known broken
  versions

- **Improvement** Handle new PyFlakes warning with a new error code: F405

2.6.1 - 2016-06-25
------------------

- **Bug** Update the config files to search for to include ``setup.cfg`` and
  ``tox.ini``. This was broken in 2.5.5 when we stopped passing
  ``config_file`` to our Style Guide

2.6.2 - 2016-06-25
------------------

- **Bug** Fix packaging error during release process.
3.0.0 -- 2016-07-25
-------------------

- Rewrite our documentation from scratch! (http://flake8.pycqa.org)

- Drop explicit support for Pythons 2.6, 3.2, and 3.3.

- Remove dependence on pep8/pycodestyle for file processing, plugin
  dispatching, and more. We now control all of this while keeping backwards
  compatibility.

- ``--select`` and ``--ignore`` can now both be specified and try to find the
  most specific rule from each. For example, if you do ``--select E --ignore
  E123`` then we will report everything that starts with ``E`` except for
  ``E123``. Previously, you would have had to do ``--ignore E123,F,W`` which
  will also still work, but the former should be far more intuitive.

- Add support for in-line ``# noqa`` comments to specify **only** the error
  codes to be ignored, e.g., ``# noqa: E123,W503``

- Add entry-point for formatters as well as a base class that new formatters
  can inherit from. See the documentation for more details.

- Add detailed verbose output using the standard library logging module.

- Enhance our usage of optparse for plugin developers by adding new parameters
  to the ``add_option`` that plugins use to register new options.

- Update ``--install-hook`` to require the name of version control system hook
  you wish to install a Flake8.

- Stop checking sub-directories more than once via the setuptools command

- When passing a file on standard-in, allow the caller to specify
  ``--stdin-display-name`` so the output is properly formatted

- The Git hook now uses ``sys.executable`` to format the shebang line.
  This allows Flake8 to install a hook script from a virtualenv that points to
  that virtualenv's Flake8 as opposed to a global one (without the virtualenv
  being sourced).

- Print results in a deterministic and consistent ordering when used with
  multiprocessing

- When using ``--count``, the output is no longer written to stderr.

- AST plugins can either be functions or classes and all plugins can now
  register options so long as there are callable attributes named as we
  expect.
3.0.1 -- 2016-07-25
-------------------

- Fix regression in handling of ``# noqa`` for multiline strings.
  (See also `GitLab#177`_)

- Fix regression in handling of ``--output-file`` when not also using
  ``--verbose``. (See also `GitLab#180`_)

- Fix regression in handling of ``--quiet``. (See also `GitLab#180`_)

- Fix regression in handling of ``--statistics``. (See also `GitLab#180`_)

3.0.2 -- 2016-07-26
-------------------

- Fix local config file discovery.  (See also `GitLab#181`_)

- Fix indexing of column numbers. We accidentally were starting column indices
  at 0 instead of 1.

- Fix regression in handling of errors like E402 that rely on a combination of
  attributes. (See also `GitLab#186`_)

3.0.3 -- 2016-07-30
-------------------

- Disable ``--jobs`` for any version of Python on Windows.
  (See also `this Python bug report`_)

- Raise exception when entry_point in plugin not callable.
  This raises an informative error when a plugin fails to load because its
  entry_point is not callable, which can happen with a plugin which is buggy or
  not updated for the current version of flake8. This is nicer than raising a
  `PicklingError` about failing to pickle a module (See also `GitLab#164`_)

- Fix ``# noqa`` comments followed by a ``:`` and explanation broken by
  3.0.0 (See also `GitLab#178`_)

- Always open our output file in append mode so we do not overwrite log
  messages. (See also `GitLab#193`_)

- When normalizing path values read from configuration, keep in context the
  directory where the configuration was found so that relative paths work.
  (See also `GitLab#194`_)

- Fix issue where users were unable to ignore plugin errors that were on
  by default. (See also `GitLab#195`_)

- Fix our legacy API StyleGuide's ``init_report`` method to actually override
  the previous formatter. (See also `GitLab#200`_)

3.0.4 -- 2016-08-08
-------------------

- Side-step a Pickling Error when using Flake8 with multiprocessing on Unix
  systems. (See also `GitLab#164`_)

- Fix an Attribute Error raised when dealing with Invalid Syntax. (See also
  `GitLab#203`_)

- Fix an unhandled Syntax Error when tokenizing files. (See also
  `GitLab#205`_)

(wiz)

Added devel/py-codestyle version 2.0.0

(wiz)

+ py-codestyle

(wiz)

Mention that this package is obsolete.

(wiz)

Import py-codestyle-2.0.0 as devel/py-codestyle.

This is the successor of py-pep8.

pycodestyle is a tool to check your Python code against some of
the style conventions in PEP 8.

(wiz)

Updated x11/xkeyboard-config to 2.19

(wiz)

Updated xkeyboard-config to 2.19.

Another regular release:

----------------------------------

Baurzhan Muftakhidinov (1):
      Add extended Kazakh layout type

Benno Schulenberg (4):
      symbols/rs: Do not specify the key.type where it isn't needed.
      symbols/rs: Remove pointless trailing "any" and "NoSymbol".
      symbols/rs: Specify some keys explicitly as four-level alphabetic.
      symbols/de: Do not specify the key.type where it isn't needed.

Mart Lubbers (2):
      add phonetic russian dvorak layout
      add contact and author

Mike FABIAN (1):
      Move Indian keyboard layouts from base.xml.in to base.extras.xml.in

Peter Hutterer (1):
      Add Australian as separate entry

Sergey Udaltsov (14):
      Added Carpalx intl layout
      de(legacy) to de(deadtilde)
      Nonexisting option removed
      Added parens:swap_brackets
      Fixed carpalx layout
      Fixed carpalx layout
      Updated cm mmuock
      ru sometimes need ralt as ralt, not as level3
      cm(mmuock), not ma
      The quote should be normal quotedbl
      2.18.99, sent to translators
      Use digits literally
      Prerelease sync with TP
      Release 2.19

Tomi Lepp辰nen (1):
      rules: Move Finnish DAS layout to extras

(wiz)

Updated x11/xf86-video-nouveau to 1.0.13

(wiz)

Updated xf86-video-nouveau to 1.0.13.

Adam Jackson (1):
      Adapt Block/WakeupHandler signature for ABI 23

Ben Skeggs (2):
      fix use of out-of-scope data
      exa/nv50-: fix some potential incomplete pushes

Hans de Goede (1):
      Properly cleanup fb for reverse-prime-offload

Ilia Mirkin (2):
      HAS_DIRTYTRACKING_ROTATION also supports multiple CRTCs
      Bump version to 1.0.13

Keith Packard (1):
      Use NotifyFd for drm and udev fds

(wiz)

Updated x11/xproto to 7.0.31

(wiz)

Updated xproto to 7.0.31.

7.0.31

A bit of a brownbag release; sorry I didn't catch the typo in review.

Keith Packard (2):
      Fix typo __has_extenstion -> __has_extension
      xproto 7.0.31

7.0.30

Jeremy Huddleston Sequoia (5):
      Xfuncproto.h: Define __has_feature() if it isn't already
      Xfuncproto.h: Define __has_extension() if it isn't already
      Xfuncproto: Add support for _X_DEPRECATED_MSG() macro
      Xfuncproto: Add support for _X_NOTSAN macro
      xproto 7.0.30

(wiz)

Fix parallel build with BSD make.
Sent upstream.

(wiz)

Use gmake to allow "make -j" build.

(wiz)

Updated net/libcares to 1.12.0

(wiz)

Updated libcares to 1.12.0. Security fix.

Version 1.12.0 (29 Sep 2016)

Daniel Stenberg (29 Sep 2016)
- RELEASE-NOTES: 1.12.0

- [David Drysdale brought this change]

  ares-test-misc: test ares_create_query with escaped trailing dot

- ares_create_query: avoid single-byte buffer overwrite

  ... when the name ends with an escaped dot.

  CVE-2016-5180

  Bug: https://c-ares.haxx.se/adv_20160929.html

- ares_library_initialized.3: added

- make: bump CARES_VERSION_INFO for release

David Drysdale (29 Sep 2016)
- man: update ares_init_options.3

Daniel Stenberg (29 Sep 2016)
- ares_library_init.3: corrected the ares_library_init_mem proto

- README.md: remove space from link

- README: link to the correct c-ares badge!

  Reported-by: David Hotham

  Fixes #63

- docs: minor formatting edits

- ares_destroy.3: formatting polish

- ares_init.3: split the init docs into two separate man pages

- SECURITY: point to the vulnerabilities page now

- RELEASE-NOTES: synced with daa7235b1a5

- ares_create_query.3: edit language

  Tried to make the man page more readable.

David Drysdale (26 Sep 2016)
- test: fix gMock to work with gcc >= 6.x

  Taken from:
  https://github.com/google/googletest/issues/705#issuecomment-235067917

Daniel Stenberg (26 Sep 2016)
- [Brad House brought this change]

  headers: remove checks for and defines of variable sizes

  ... they're not really used and by avoiding them in the ares_build.h
  output we make the public header less dependent on data sizes.

David Drysdale (24 Sep 2016)
- api: add ARES_OPT_NOROTATE optmask value

  Fix up a couple of problems with configuring whether c-ares rotates
  between different name servers between requests.

  Firstly, ares_save_options() returns (in *optmask) the value of
  (channel->optmask & ARES_OPT_ROTATE), which doesn't necessarily
  indicate whether the channel is or is not actually doing rotation.
  This can be confusing/incorrect if:
  - the channel was originally configured without ARES_OPT_ROTATE
    (so it appears that the channel is not rotating)
  - the /etc/resolv.conf file includes the 'rotate' option
    (so the channel is actually performing rotation).

  Secondly, it is not possible to reliably configure a channel
  to not-rotate; leaving off ARES_OPT_ROTATE is not enough, since
  a 'rotate' option in /etc/resolv.conf will turn it on again.

  Therefore:
  - add an ARES_OPT_NOROTATE optmask value to allow explicit
    configuration of no-rotate behaviour
  - in ares_save_options(), report the value of channel->rotate
    as exactly one of (optmask & ARES_OPT_ROTATE) or
    (optmask & ARES_OPT_NOROTATE).

  In terms of back-compatibility:
  - existing apps that set ARES_OPT_ROTATE will continue to rotate,
    and to have ARES_OPT_ROTATE reported back from ares_save_options()
  - existing apps that don't set ARES_OPT_ROTATE will continue to
    use local config/defaults to decide whether to rotate, and will
    now get ARES_OPT_ROTATE or ARES_OPT_NOROTATE reported back from
    ares_save_options() rather than 0.

- ares_init_options: only propagate init failures from options

  Commit 46bb820be3a8 ("ares_init_options: don't lose init failure")
  changed init behaviour so that earlier errors in initialization
  weren't lost.  In particular, if the user passes in specific
  options but they are not applied (e.g. because of an allocation
  failure), that failure needs to be reported back to the user; this
  also applies when duplicating a channel with ares_dup().

  However, other initialization failures can be ignored and
  overridden -- in particular, if init_by_resolv_conf() or
  init_by_environment() fail, then falling back to default values
  is OK.

  So only preserve failures from the init_by_options() stage, not
  from all initialization stages.

  Fixes issue 60.

- test: Force reinstall of libtool on OSX

  Travis build environment appears to have changed.

- test: Add valgrind build variant

- test: Add null pointer to gtest args

  GoogleTest assumes that there is a null pointer in argv[argc],
  so make it look like that. Without this change, tests run with
  command-line arguments get memory errors under valgrind/ASAN.

Daniel Stenberg (21 Aug 2016)
- AUTHOR: maybe gitgub isn't really an author =)

- AUTHORS: added contributors from the git log

- LICENSE.md: add a stand-alone license file

  Just the MIT license used in the top the source files moved out to a
  stand-alone file for easier reference and discovery.

- README: added "CII best practices" badge

- SECURITY.md: suggested "security process" for the project

David Drysdale (17 Aug 2016)
- test: Add Clang static analysis build to Travis

  Run scan-build over the library source code, but skip the
  tests.  Needs a later Clang install in Travis

- test: more info on how to run fuzz testing

- test: make fuzzer driver code C not C++

- test: fuzzer mode for AFL's persistent mode

  When fuzzing with AFL, if the LLVM-based instrumentation is
  used (via the afl-clang-fast wrapper), then it is possible to
  have a single execution of the fuzzer program iterate multiple
  times over the fuzzing entrypoint (similar to libFuzzer's normal
  mode of execution) with different data.  This is much (e.g. 10x)
  faster.

  Add code to support this, by checking whether __AFL_LOOP is
  defined at compile-time.

  Also, shift the code to effectively be C rather than C++.

- test: simplify deps for fuzzer entrypoint

  No need to depend on the rest of the test code (ares-test.h) for
  the fuzzer entrypoint; this makes the entrypoint slightly simpler
  to build with LLVM's libFuzzer.

  Also shift the code to effectively be C rather than C++

- test: disable MinGW tests

  The test binary built in the MinGW build is failing for some
  reason.  It works for me when I build locally, so I'm guessing
  it's down to some sort of AppVeyor environment issue.

  Disable for now.

Daniel Stenberg (16 Aug 2016)
- read_tcp_data: remove superfluous NULL check

  CID 56884 by Coverity. The pointer is already derefenced before this
  point so it can't be NULL here anyway.

- web: http => https

GitHub (20 Jul 2016)
- [David Drysdale brought this change]

  Merge pull request #59 from fuze/master

  Update msvc_ver.inc for VS2015 Update 3

- [Chris Araman brought this change]

  Update msvc_ver.inc

  support Visual Studio 2015 Update 3

David Drysdale (2 May 2016)
- Fix trailing comment for #endif

Daniel Stenberg (30 Apr 2016)
- email: use Gisle's "new" address

David Drysdale (18 Apr 2016)
- test: drop superfluous fuzz inputs

  Where there are multiple fuzz input files that only differ in
  the first two bytes (the query ID), just keep the first such
  file.

svante karlsson (15 Apr 2016)
- Update msvc_ver.inc

  support Visual Studio 2015 Update 2

David Drysdale (31 Mar 2016)
- test: Run fuzzcheck.sh in Travis build

- test: add fuzzing check script to tests

  Add a test script that runs the fuzzing command over the
  corpus of DNS packets.  This doesn't actually do any fuzzing
  (it just runs them as inputs without generating any variations)
  but it does ensure that the fuzzing entrypoint is still working.

- test: allow multiple files in aresfuzz command line

  If no arguments are specified, use stdin as input.
  Otherwise treat each argument as a filename and feed
  its contents to the fuzz entrypoint.

- test: Add corpus of DNS packets

  For fuzz testing it is useful to start from a corpus of valid
  packets, so fill out the test/fuzzinput/ directory with a bunch
  of inputs.

  These packets were generated by temporarily modifying the c-ares
  process_answer() function to save off any incoming response messages.

- test: Add utility to show DNS packet from file

- [nordsturm brought this change]

  Fix nsort initialization

  Author: Alexander Drachevskiy
  http://c-ares.haxx.se/mail/c-ares-archive-2014-07/0004.shtml
  http://c-ares.haxx.se/mail/c-ares-archive-2014-07/0014.shtml

- test: Check setting nsort=0 option is respected

- test: Update fuzzing function prototype

  libFuzzer changed expected return type from void to int
  in LLVM 3.8.

- Explicitly clear struct servent before use

  On a build where MSAN has been manually set up (which involves
  using an MSAN-instrumented version of the standard C++ library, see
  https://github.com/google/sanitizers/wiki/MemorySanitizerLibcxxHowTo)
  there's a warning about use of uninitialized memory here.  It
  might be a false positive, but the fix is trivial so include it.

- test: for AF_UNSPEC, return CNAME only for AAAA, but valid A record

  Also shuffle expected responses rsp6/rsp4 into the order they will occur.

- [Chris Araman brought this change]

  msvc_ver.inc: support Visual Studio 2015 Update 1

- build: commonize MSVC version detection

  Remove the need to copy/paste version number mapping between
  Makefile.msvc and test/Makefile.msvc.

- test: Use different name in live test

- test: Only pass unused args to GoogleTest

- ahost.c: add cast to fix C++ compile

  If ahost.c is force-compiled as C++ the missing cast from
  (void *) to (char **) is problematic.

- ares_library_cleanup: reset ares_realloc too

  Otherwise a subsequent use of the library might use a previous
  incarnation's realloc() implementation.

Daniel Stenberg (9 Mar 2016)
- [Brad House brought this change]

  configure: check if tests can get built before enabled

  The current approach for disabling tests is not a good solution because
  it forces you to pass --disable-tests, rather than auto-detect if your
  system can support the tests in the first place.  Many (most?) systems
  do not have C++11.  This also causes issues when chain-building c-ares,
  the hosting system needs to be updated to support passing this
  additional flag if necessary, it doesn't seem reasonable to add this
  requirement which breaks compatibility.

  This change auto-detects if the system can build the tests and
  automatically disable them if it cannot.  If you pass --enable-tests to
  configure and the system cannot build them either due to lack of system
  support, or because cross-compilation is being used, it will throw an
  appropriate error since the user indicated they really did want the
  tests.

David Drysdale (3 Mar 2016)
- [Viktor Szakats brought this change]

  Makefile.m32: add support for CROSSPREFIX

- [Viktor Szakats brought this change]

  Makefile.m32: add support for extra flags

  Allow specification of CARES_{LD,C}FLAG_EXTRAS envvars
  for mingw

- test: Build with MinGW on AppVeyor

- test: avoid in6addr_* constants

  These aren't available on MinGW, so use explicit addresses instead.

- test: add missing #includes for dns-proto.cc

- [Gregor Jasny brought this change]

  Fix man page typos detected by Lintian

Daniel Stenberg (19 Feb 2016)
- configure: acknowledge --disable-tests

  Fixes #44

- AUTHORS: added contributors from the 1.11.0 release

- bump: start working on the next version

Version 1.11.0 (19 Feb 2016)

Daniel Stenberg (19 Feb 2016)
- RELEASE-NOTES: final edits for 1.11.0

David Drysdale (15 Feb 2016)
- ares_dup.3: remove mention of nonexistent function

  ares_dup_options() doesn't exist, so don't document it.

- test: skip repeated build steps

  Top-level buildconf/configure now triggers for the
  test/ subdir too, so don't need to do explicitly.

- test: namespaces unavailable when cross-compiling

Daniel Stenberg (13 Feb 2016)
- configure: only run configure in test when NOT cross-compiling

  ... as the tests won't run cross-compiled anyway

David Drysdale (13 Feb 2016)
- test: prefer ON_CALL to EXPECT_CALL to reduce flakes

  For UDP tests, there's a chance of a retry.  EXPECT_CALL only
  expects a single request to arrive at the server; ON_CALL allows
  for a UDP retry and repeats the same answer.

  Note that ON_CALL and EXPECT_CALL can't be mixed in the same
  test, and that tests that have a varied sequence of responses
  for the same repeated request still have to use EXPECT_CALL.

Daniel Stenberg (13 Feb 2016)
- configure: run configure in 'test' too

  Having the test dir completely stand-alone causes too many issues for
  users and devs. It still needs to be built specifically.

- configure: build silently by default

- buildconf: run test/buildconf too if present

- test/configure: build silently by default

- [Gregor Jasny brought this change]

  dist: Distribute README.md

  Closes #42

Version 1.11.0 (11 Feb 2016)

Daniel Stenberg (11 Feb 2016)
- Makefile.am: distribute the test dir too

- RELEASE-NOTES: synced with 385582bd14b68a

- [Nicolas \"Pixel\" Noble brought this change]

  ares_win32_init: make LoadLibrary work when using UNICODE too

  Closes #17

David Drysdale (11 Feb 2016)
- Use "resolve" as synonym of "dns" in nsswitch.conf

  Modern Linux systems may have libnss_resolve from systemd as the
  resolver, which is then configured in /etc/nsswitch.conf with
  the "resolve" keyword rather than "dns".

  Fixes #33

- ares_set_socket_callback: make manpage match code

  The code in ares_process.c that invokes the socket creation/connection
  callback only checks for rc < 0, not for standard ares error codes.

- Merge pull request #36 from AGWA-forks/master

  Add ares_set_socket_configure_callback()

- test: Update init tests to match behaviour

  Unreadable config files are now treated the same way
  as absent config files.

- [Fedor Indutny brought this change]

  Ignore `fopen` errors to use default values

  After 46bb820be3a83520e70e6c5f0c5133253fcd69cd `init_by_resolv_conf`
  errors are no longer swallowed in `ares_init_options`. This has exposed
  a previously unknown bug in `lookups` initialization code.

  If there is no lookup configuration in `resolv.conf`,
  `init_by_resolv_conf` will attempt to read it from other files available
  on the system. However, some of these files may have restricted
  permissions (like `600`), which will lead to `EACCESS` errno, which in
  turn is handled like a fatal error by `init_by_resolv_conf`.

  However, it sounds illogical that this error should be handled as a
  fatal. There is a `init_by_defaults` call that overrides `lookups` with
  default value, and certainly possible absence of lookup information is
  the reason why this function exists in a first place!

  I suggest handling any `fopen` errors as non-fatal ones, allowing to
  pick up the `lookups` value from different config files, or to pick up
  default value.

Andrew Ayer (9 Feb 2016)
- Document callback type in man page for ares_set_socket_callback

- Add ares_set_socket_configure_callback()

  This function sets a callback that is invoked after the socket is
  created, but before the connection is established.  This is an ideal
  time to customize various socket options.

David Drysdale (9 Feb 2016)
- test: ares_set_socket_callback failure behaviour

- test: Check ares_parse_txt_reply_ext() entrypoint

- [Fedor Indutny brought this change]

  txt: introduce `ares_parse_txt_reply_ext`

  Introduce `ares_txt_ext` structure with an extra `record_start`
  field, which indicates a start of a new TXT record, thus allowing to
  differentiate the chunks in the same record, from a chunks in a
  different record.

  Introduce a new API method: `ares_parse_txt_reply_ext` that works with
  this kind of struct.

- doc: Update missed repo references

- doc: Update docs on contributing

- test: Run command line tools in Travis

  Do a quick execution of each of the command line tools
  in the continuous integration build, so that any (say)
  sanitizer failures show up.

- acountry: drop inert test

  If ver_1 is true, then z0 and z1 must both be 'z', and so
  (z0 != 'z' && z1 != 'z') can never be true.

  CID 56879, pointed out by Coverity.

- doc: update badge locations to master repo

- test: Enable maintainer mode + debug in Travis

- test: Add an iOS build target

- test: Ignore SIGPIPE in tests

- test: More initialization tests

- test: Improve containerized test mechanism

  Aim is to ensure that code coverage information can escape the
  container.  To do this:
  - Enter a new mount namespace too, so that we can...
  - Bind mount the expected source directory into the container
  - Share memory with the sub-process so coverage information is
    shared too.

- test: Make contained tests easier to write

- test: Add framework for containerized testing

  On Linux we can potentially use user and UTS namespaces to run  a test
  in a pseudo-container with:
  - arbitrary filesystem (e.g. /etc/resolv.conf, /etc/nsswitch.conf, /etc/hosts)
  - arbitrary hostname/domainname.

  Include a first pass at the framework code to allow this, along with a
  first test case that uses the container.

- test: Use a longer timeout for less flakiness

  Having occasional test failures from timeout before multiple
  queries can complete, so up the default timeout for the test
  from 100ms to 1500ms.

- test: Make failure tests more robust

  Different platforms will do different numbers of allocations
  in the processing of a given API call; just check that the
  return code is either success or ENOMEM, and free off any
  returned state in the former case.

  Also cope with ECONNREFUSED as well as ENOTFOUND.

- test: Get test code building under Windows

  - Initial nmake file based off library nmake file
  - Cast socket call arguments to (char *)
  - Use wrapper sclose() that maps to closesocket() or close()
  - Build a config.h indicating presence of headers
  - Conditionally include netdb.h
  - Remove unnecessary include of sys/socket.h
  - Force longer bitmask for allocation failure tracking
  - Call WSAStartup() / WSACleanup() in main()
  - Set TCP_NODELAY for mock server
  - Turn on tests in AppVeyor build

- test: Disable tests that manipulate env on Windows

- test: Move file lists into Makefile.inc

  In preparation for a Win32 build of the test suite.

- test: Add a simple multi-server test

  Check rotate option does something

- test: Allow for multiple mock servers

  - Update the MockServer to allow separate specification of
    UDP and TCP ports
  - Have an array of mock servers listening on consecutive
    sets of ports.
  - Rename Process(fd) to ProcessFD(fd) to avoid confusion.
  - Initialize channel by using the new ares_set_servers_ports()
    entrypoint, so multiple ports on the same loopback address
    can be used.

- test: Update test for set/get_servers variants

  Ports are significant in the _ports_ variant functions, so update test to cope.

- test: Make GetNameServers() utility function port-aware

  Also make it generally available.

- test: more testing, including of internal static functions

- test: more tests, especially fallback processing

  - Make mock server listen on UDP + TCP in parallel.
  - Test UDP->TCP fallback on truncation
  - Test EDNS->no-EDNS fallback
  - Test some environment init options
  - Test nonsense reply

  test: short response

- test: more tests, particularly of initialization

- test: Run mock tests over both TCP and UDP

  With the exception of a few tests that make use of the timed
  retry aspect of UDP.

- test: Run mock tests over both IPv4 and IPv6

- test: Add more tests for edge cases

- test: more nooks and crannies of pton functions

- test: More tests for PTR parsing

- test: Use of HOSTALIAS environment variable

- test: Add RAII utility classes for testing

  - TempFile holds specific contents
  - EnvValue sets an environment variable

- test: More search domain scenarios

- test: Remove duplicate flags from Makefile.am

- test: Make test code leak-free

- test: More tests

  - test use of sortlist
  - test gethostbyname(AF_UNSPEC)

- test: Test ares_gethostbyname_file()

- test: Add more tests of ares_getnameinfo()

- test: Tweak tests, add alloc failure test

- test: Test init with options

- test: More tests

  - ares_inet_net_pton() variants
  - ares_getsock() variants

- test: Expose ProcessWork() function

- test: More parsing tests

  Including:
  - Split each parse function test set out into separate files.
  - Add an allocation failure test for each parsing function.
  - Add error check test for each parsing function.

- test: Add various additional tests

- test: More tests

  Include tests of internal functions, based on the value of the
  CARES_SYMBOL_HIDING macro; need to configure the library with
  --disable-symbol-hiding to enable these tests.

- test: Allow command line override of mock server port

- test: Add README.md documentation

- test: Temporarily avoid latest Python requests package

  Currently get error from Travis on this install step, and downgrading one
  version appears to fix the problem.

  "Could not find any downloads that satisfy the requirement pyOpenSSL>=0.13
  (from requests[security])"

- test: Add AppVeyor config file for Windows build

- test: Add configuration for a Travis build

  Cover Linux & OSX on the container infrastructure, but install
  a later G++ to satisfy the tests' need for C++11.

  Use a build matrix to include a variety of build variants:
  - ASAN
  - UBSAN
  - LSAN
  - Coverage via coveralls.io

  test: invoke ASAN and coverage in Travis build

  Also shift to use explicit build matrix

  test: Use coveralls.io for coverage tracking

  test: Add a build with UBSAN

  Also expand and re-order the setting of environment variables
  for easier modification.

  test: Add LSAN build to Travis config

- test: Add initial unit tests for c-ares library

  The tests are written in C++11, using the GoogleTest and GoogleMock
  frameworks.  They have their own independent autoconf setup, so that
  users of the library need not have a C++ compiler just to get c-ares
  working (however, the test/configure.ac file does assume the use of
  a shared top-level m4/ directory).  However, this autoconf setup has
  only been tested on Linux and OSX so far.

  Run with "./arestest", or "./arestest -v" to see extra debug info.
  The GoogleTest options for running specific tests are also
  available (e.g. "./arestest --gtest_filter=*Live*").

  The tests are nowhere near complete yet (currently hitting around
  60% coverage as reported by gcov), but they do include examples
  of a few different styles of testing:

  - There are live tests (ares-test-live.cc), which assume that the
    current machine has a valid DNS setup and connection to the
    internet; these tests issue queries for real domains but don't
    particularly check what gets returned.  The tests will fail on
    an offline machine.

  - There a few mock tests (ares-test-mock.cc) that set up a fake DNS
    server and inject its port into the c-ares library configuration.
    These tests allow specific response messages to be crafted and
    injected, and so are likely to be used for many more tests in
    future.

      - To make this generation/injection easier, the dns-proto.h file
        includes C++ helper classes for building DNS packets.

  - Other library entrypoints that don't require network activity
    (e.g. ares_parse_*_reply) are tested directly.

  - There are few tests of library-internal functions that are not
    normally visible to API users (in ares-test-internal.cc).

  - A couple of the tests use a helper method of the test fixture to
    inject memory allocation failures, using the earlier change to the
    library to allow override of malloc/realloc/free.

  - There is also an entrypoint to allow Clang's libfuzzer to drive
    the packet parsing code in ares_parse_*_reply, together with a
    standalone wrapper for it (./aresfuzz) to allow use of afl-fuzz
    for further fuzz testing.

- test: Add local copy of GoogleMock/GoogleTest 1.7.0

  Don't check in gtest/m4 files, as they are unused and interfere
  with the top-level configure process.

- doc: Show build badges in README.md

  Note that these URLs will need to be updated if/when the test branch
  gets pulled into the master repo/branch.

- doc: Convert README to README.md

  Gives better display on GitHub

- doc: Update in preparation for next release

  Assume 1.11.0 is next (as there are various API additions).
  Also add myself to AUTHORS.

- build: Allow header compilation by Windows C++ compiler

- build: Expose whether symbol hiding is on

  Adding the CARES_SYMBOL_HIDING definition allows the test suite to
  detect whether internal symbols are available or not.

- build: Add autoconf macros for C++11 code using pthreads

  Pull in testing macros from the GNU autoconf archive to allow
  configure scripts to test for and setup use of a C++11 compiler
  (AX_CXX_COMPILE_STDCXX_11) and the pthreads library (AX_PTHREAD).

  Note that these macros are not used by the main library autoconf,
  just by the tests (which share the same m4/ directory).

- build: Add a code coverage option

  Configure with:
    ./configure --enable-code-coverage
  Show coverage output with:
    make code-coverage-capture

  Built on m4/ax_code_coverage.m4 from the GNU autoconf archive
  to provide the macros to check for presence of gcov + lcov;
  upstream macro modified to:
  - Remove use of $(AM_DEFAULT_VERBOSITY) , as earlier versions of
    autoconf (such as the one used by default on Travis) do not have this.
  - Rather than automatically defining CODE_COVERAGE_RULES to be a set
    of makefile rules that use ifeq/endif (which is GNU make-specific),
    instead only define CODE_COVERAGE_RULES if coverages is turned on,
    and in that case don't use conditionals in the makefile.

- api: Add entrypoints to allow use of per-server ports

  Add user-visible entrypoints ares_{get,set}_servers_ports(3), which
  take struct ares_addr_port_node rather than struct ares_addr_node.
  This structure includes a UDP and TCP port number; if this is set
  to zero, the channel-wide port values are used as before.

  Similarly, add a new ares_set_servers_ports_csv(3) entrypoint, which
  is analogous to ares_set_servers(3) except it doesn't ignore any
  specified port information; instead, any per-server specified port
  is used as both the UDP and TCP port for that server.

  The internal struct ares_addr is extended to hold the UDP/TCP ports,
  stored in network order, with the convention that a value of zero
  indicates that the channel-wide UDP/TCP port should be used.

  For the internal implementation of ares_dup(3), shift to use the
  _ports() version of the get/set functions, so port information is
  transferred correctly to the new channel.

  Update manpages, and add missing ares_set_servers_csv to the lists
  while we're at it

- api: Add ares_set_sortlist(3) entrypoint

  Allow explicit configuration of the channel's sortlist, by
  specifying a string in the same format as the equivalent
  /etc/resolv.conf option.

  This allows library users to perform the same configuration
  that is available via /etc/resolv.conf, but without needing
  to change that file.

- api: Allow injection of user-specified malloc/free functions

  Add a new ares_library_init_mem() initialization function for the
  library which allows the library user to specify their own malloc,
  realloc & free equivalents for use library-wide.

  Store these function pointers in library-wide global variables,
  defaulting to libc's malloc(), realloc() and free().

  Change all calls to malloc, realloc and free to use the function pointer
  instead.  Also ensure that ares_strdup() is always available
  (even if the local environment includes strdup(3)), and change the
  library code to always use it.

  Convert calls to calloc() to use ares_malloc() + memset

- api: Add option to expose some internal functions

  Purely for testing, add --enable-expose-statics option to configure
  which converts some static internal functions to be externally visible.

- api: Expose the ares_library_initialized() function

- ahost: Allow repeated -s <domain> options

  This also removes a potential leak where later -s options would
  replace earlier ones without freeing the relevant string.

- Mark unhittable lines

  Add comments for the benefit of the lcov tool, marking
  lines that cannot be hit.  Typically these are fall-back
  protection arms that are already covered by earlier checks,
  and so it's not worth taking out the unhittable code (in case
  someone changes the code between the two places in future).

- ares_set_servers_csv.3: make return type match code

- bitncmp: update comment to match code behaviour

- ares_striendstr: fix so non-NULL return can happen

  This looks to have been broken since it was first introduced in 2005 in
  commit aba0b775ea30 ("Added ares_getnameinfo which mimics the
  getnameinfo API")

- config_sortlist: free any existing sortlist on (re)alloc failure

  If we get an allocation failure on 2nd or later entry in the sortlist, the
  code would return ENOMEM but still leave the initial entries allocated.
  Ensure that *sortlist is set to NULL whenever ENOMEM is returned.

- ares_dup: clear new channel on failure

  If the attempt to transfer IPv6 servers from the old to the new channel
  fails, the previous code would still return a channel to the user even though
  an error return code was generated.  This makes it likely that users would
  leak the channel, so explicitly clear the channel in this case.

- ares_init_options: don't lose init failure

  If (say) init_by_options() fails, the subsequent call to
  init_by_defaults() was overwriting the return code with
  success.  Still call init_by_defaults() regardless, but track
  its return value separately

- ares_gethostbyname: don't leak valid-but-empty hostent

  If an AF_UNSPEC query gets a valid response to its AAAA query,
  but which has no IPv6 addresses in it, then the code chains on to
  a A record query.  However, the hostent from the AAAA response
  was being leaked along the way (because it gets replaced before
  the follow-on end_hquery() invocation).

- ares_parse_txt_reply: propagate errors from per-substring loop

  If we get an allocation failure when processing a particular substring in a
  TXT record, that failure is silently lost; fix that by propagating errors from
  the inner loop to the outer loop.

- process_answer: fix things up correctly when removing EDNS option

  When a server rejects an EDNS-equipped request, we retry without
  the EDNS option.  However, in TCP mode, the 2-byte length prefix was
  being calculated wrong -- it was built from the answer length rather than
  the length of the original request.

  Also, it is theoretically possible that the call to realloc() might change
  the data pointed to; to allow for this, qbuf also needs updating.

  (Both these fixes were actually included in a patchset sent on the mailing
  list in Oct 2012, but were included with other functional changes that
  didn't get merged:
  http://c-ares.haxx.se/mail/c-ares-archive-2012-10/0004.shtml)

- ares__read_line: clear buf pointer on realloc failure

- ares_expand_name: check for valid bits in label length

  The top two bits of the label length indicate whether this is a
  label length (00) or an index to a name elsewhere in the message
  (11).  RFC1035 4.1.4 says that the other possible values for the
  top two bits (01, 10) are reserved for future use.

Daniel Stenberg (23 Jan 2016)
- [Gregor Jasny brought this change]

  Fix typos detected by lintian

  Closes #32

- [Gregor Jasny brought this change]

  Distribute all man pages

- README.cares: s/I/Daniel

  ... and add a pointer to an existing version of the original area 1.1.1
  package.a

- read_tcp_data: don't try to use NULL pointer after malloc failure

  CID 56884, pointed out by Coverity. We really should make this function
  return an error code so that a malloc() failure can return back a major
  failure.

- configure_socket: explicitly ignore return code

  CID 56889 in Coverity pointed out the return code from setsocknonblock()
  is ignored, and this added typecast to (void) makes it explicit.

- ahost: check the select() return code

  Fixes CID 137189, pointed out by Coverity

David Drysdale (18 Jan 2016)
- Fix buildconf on platforms using glibtoolize

  Commit c49a87eea538 changed buildconf to only check for
  libtoolize, but missed a line

- Don't exit loop early leaving uninitialized entries

  Update for commit affc63cba875d.

  The original patch from Gregor Jasny did not have the break
  statement; I incorrectly added it to prevent continuing the loop.
  However, the later entries in the array would then be left
  uninitialized, causing problems for later cleanup.

  So fix to match Gregor's original patch, with apologies.

Daniel Stenberg (18 Jan 2016)
- buildconf: remove check for libtool, it only requires libtoolize

David Drysdale (17 Jan 2016)
- [Gregor Jasny brought this change]

  Use libresolv to initialize cares on iPhone targets

  On iPhone targets like iOS, watchOS or tvOS the file
  /etc/resolv.conf cannot be used to configure cares.

  Instead the resolver library is queried for configuration
  values.

  CC: Yury Kirpichev <ykirpichev@yandex-team.ru>

Daniel Stenberg (17 Jan 2016)
- README: updated to new repo URL

David Drysdale (14 Jan 2016)
- [Lei Shi brought this change]

  Fixing slow DNS lookup issue

  This patch is fixing the dns lookup issue due to dummy dns information
  of a disconnected adapter(in my case is a bluetooth adapter). I changed
  the dns lookup policy to try GetNetworkParams first because the
  GetNetworkParams provides the most reliable dns information (lots of
  checks were done by system). I also filter out inoperable adapter in
  DNS_AdaptersAddresses in case GetNetworkParams fail.

- Merge pull request #30 from p-push/vs-2015

  Support Visual Studio 2015

Oleg Pudeyev (3 Jan 2016)
- [Gisle Vanem brought this change]

  Support Visual Studio 2015

David Drysdale (11 Nov 2015)
- [Andrew Andkjar brought this change]

  added another version case to Makefile.msvc

  nmake version 11.00.61030.0 resolves to CC_VERS_NUM = 110

- Merge pull request #26 from bitbouncer/vs-2013

  added define for visual studio 2013

svante karlsson (25 Jun 2015)
- added define for visual studio 2013

Jakub Hrozek (6 Nov 2014)
- ares__read_line: free buf on realloc failure

- Destroy options if ares_save_options fails

  It's possible that, if ares_save_options failed, the opts structure
  would contain some allocated memory. Calling ares_destroy_options in
  this case is safe, because ares_save_options zeroes out the memory
  initially.

- [David Drysdale brought this change]

  Continue loop if space for hostname not large enough

  When attempting to build a search domain from the local hostname
  (used as a fallback when no other methods have given a search
  domain), the code doubles the buffer size on each loop iteration.

  However, the loop previously had a WHILE_FALSE terminator so the continue
  statement exited the loop rather than going round again.

Daniel Stenberg (30 Oct 2014)
- ares_getnameinfo.3: there is no ares_getaddrinfo

David Drysdale (30 Sep 2014)
- [Gregor Jasny brought this change]

  Prevent tmpbuf from overrunning

  Fix Coverity error CID 56886.

  Signed-off-by: Gregor Jasny <gjasny@googlemail.com>

- [Gregor Jasny brought this change]

  Re-start loop if select fails

  Fix Coverity error CID 56882

  Signed-off-by: Gregor Jasny <gjasny@googlemail.com>

- [Gregor Jasny brought this change]

  Free temporary variable in error path

  Fix Coverity CID 56890

  Signed-off-by: Gregor Jasny <gjasny@googlemail.com>

- [Gregor Jasny brought this change]

  Fix integer shift overflow if both tcp_socket and udp_socket are set

  The problem occurs if at the start of the loop the sockindex is at the
  last valid ARES_GETSOCK_MAXNUM position. If then both udp_socket and
  tcp_socket are valid, sockindex gets incremented for UDP first and
  points one entry behind the array for the tcp block.
  So the fix is to check after every increment of sockindex if it is still
  valid.

  Fix Coverity error CID 56878

  Signed-off-by: Gregor Jasny <gjasny@googlemail.com>

- [Gregor Jasny brought this change]

  Null check before dereference

  Fix Coverity error CID 56880

  Signed-off-by: Gregor Jasny <gjasny@googlemail.com>

Jakub Hrozek (28 Jul 2014)
- [Gisle Vanem brought this change]

  Comment in ares_ipv6.h

David Drysdale (25 Jul 2014)
- CONTRIBUTING: add file to indicate mailing list is preferred

- Add -t u option to ahost

  Add an option to allow specification of the AF_UNSPEC
  address family.

Jakub Hrozek (24 Jul 2014)
- host_callback: Fall back to AF_INET on searching with AF_UNSPEC

  Previously, when an ares_gethostbyname() searched with AF_UNSPEC and the
  first AF_INET6 call only returned CNAMEs, the host_callback never
  retried AF_INET.

  This patch makes sure than on ARES_SUCCESS, the result of AF_INET6 is
  taken as authoritative only if the result contains some addresses.

- [David Drysdale brought this change]

  Move memset call below platform-specific declarations

  A GitHub commenter [1] says that my recent change to ahost.c has
  problems compiling on Windows + C89 platforms.

  [1]  https://github.com/bagder/c-ares/commit/ee22246507c9#commitcomment-6587616

- [David Drysdale brought this change]

  Update ahost man page to describe -s option.

  Commit ee22246507c9 added the -s <domain> option to the
  ahost command, but neglected to update the man page to
  describe it.

  Also fix typo in description of -t option.

- ares_parse_soa_reply: Do not leak rr_name on allocation failure

  If ares_malloc_data failed, already allocated rr_name would go out of
  scope.

- [David Drysdale brought this change]

  Don't override explicitly specified search domains

  Only set search domains from /etc/resolv.conf if there isn't a value
  already present in the channel.

- [David Drysdale brought this change]

  Allow specification of search domain in ahost

  Add the "-s domain" command line option to override the search
  domains.

Daniel Stenberg (12 May 2014)
- Revert "ares_parse_aaaa_reply: fix leak when reply contains 1 alias and no address"

  This reverts commit 440110b303fdbfadb3ad53d30eeb98cc45d70451.

- [Frederic Germain brought this change]

  ares_parse_aaaa_reply: fix leak when reply contains 1 alias and no address

- [Doug Kwan brought this change]

  ares_build.h: fix building on 64-bit powerpc

  There are two issues.

  1. gcc actually does not use __ppc__ and __ppc64__ but __PPC__ and
  __PPC64__.  The tests of __ILP32__ and __LP64__ are sufficient for gcc.

  2. clang defines __GNU__ and defines both __ppc64__ and __ppc__ when
  targeting ppc64.  This makes CARES_SIZEOF_LONG to be 4 on a ppc64 system
  when building with clang.

  My patch is two change the order of the checks so that we check the
  64-bit case first.

- refresh: updated now with automake 1.14

- [David Drysdale brought this change]

  single_domain: Invalid memory access for empty string input

  We noticed a small buglet in ares_search() when it gets an empty string
  as input -- the single_domain() utility function in ares_search.c
  accesses invalid memory (before the start of the string).

Guenter Knauf (31 Aug 2013)
- Fixed warning 'type specifier missing'.

Daniel Stenberg (30 Aug 2013)
- [Tor Arntsen brought this change]

  ares_rules.h: CARES_SIZEOF_LONG doesn't exist anymore, don't test for it

  It was removed in f19387dd72432

- nowarn: use <limits.h> instead of configure for size of long

  This makes the header file much more multi-arch friendly and can be used
  as-is with both 32 bit and 64 bit builds.

- timeoffset: made static and private

  ares__timeoffset() was only used once within this single source file

- timeadd: make static

  ares__timeadd() was only ever used from within the same source

Yang Tse (18 Jul 2013)
- xc-am-iface.m4: comments refinement

- configure: fix 'subdir-objects' distclean related issue

  See XC_AMEND_DISTCLEAN comments for details.

- configure: automake 1.14 compatibility tweak (use XC_AUTOMAKE)

- xc-am-iface.m4: provide XC_AUTOMAKE macro

Daniel Stenberg (12 May 2013)
- gitignore: ignore all ares_*pdf but also CHANGES.dist

- bump: start working towards 1.10.1

(wiz)

+ 3proxy-0.8.7, 9menu-1.9, 9wm-1.3.9, abcm2ps-8.12.4, abcmidi-20160505,
  abook-0.6.1, accerciser-3.22.0, ack-2.14, admesh-0.98.2, adms-2.3.5,
  adplug-2.2.1, advancecomp-1.20, adwaita-icon-theme-3.22.0,
  aegis-4.25, aewm++-1.1.5, afbinit-1.0.1, aide-0.16, akonadi-16.08.1,
  alephone-20150620, alevt-1.6.2, fira-4.106, hdf5-1.10.0pl1,
  p5-Crypt-ECB-2.21, p5-DateTime-TimeZone-2.04, p5-Email-Valid-1.202,
  p5-Params-Validate-1.25, py-django-1.10.1, py-mezzanine-4.2.2,
  the_silver_searcher-0.33.0, xf86-video-nouveau-1.0.13 [wiz],
  xproto-7.0.31 [wiz].

(wiz)

Fix HOMEPAGE.

(wiz)

Switch slang option to slang2 so it builds.

>From gcw@primenet in PR 51513.

(wiz)

Updated pkgtools/url2pkg to 2.28

(wiz)

Fix HOMEPAGE when packaging github releases. Bump version.

(wiz)

Updated multimedia/minidlna to 1.1.5nb2

(wiz)

Install man pages. Requested by Ben Gergely in PR 51506.

Bump PKGREVISION.

(wiz)

Fix build of abiword-plugins with gnutls-3.5.
>From upstream via David H. Gutteridge in PR 51501.

Bump PKGREVISION.

(wiz)

+ ImageMagick-7.0.3.1, MesaLib-12.0.3, afl-2.35b, cups-filters-1.11.4,
  global-6.5.5, liblangtag-0.6.1, orcus0.12-0.12.1,
  p5-Canary-Stability-2012, p5-DateTime-Locale-1.08, p5-Email-Valid-1.201,
  p5-Specio-0.26, p5-Text-BibTeX-0.77, py-cryptography-1.5.1,
  py-html2text-2016.9.19, py-hypothesis-3.5.1, tor-0.2.8.8,
  x264-devel-20160924, xf86-video-nouveau-1.0.13, xproto-7.0.31.

(wiz)

Remove nonworking master site.

(wiz)

Use more up-to-date homepage.

(wiz)

Remove restrictions, this now has standard licensing.

(wiz)

+ gnome-3.22, gnuchess-6.2.3, libjpeg-turbo-1.5.1, p5-Business-ISBN-3.003,
  p5-Net-Domain-TLD-1.75, p5-Text-Glob-0.10.

(wiz)

Updated devel/py-cffi to 1.8.3nb1

(wiz)

Add patch to distinfo and bump PKGREVISION for it.

('make test' still fails with mprotect enabled.)

(wiz)

Fix PLIST on non-NetBSD.

(wiz)

+ gssdp-1.0.0, gupnp-1.0, opensmtpd-6.0.0, pan-0.140.

(wiz)

Updated print/dvidvi to 1.1

(wiz)

Update dvidvi to 1.1, using the texlive distfile.

Diff provided by markd.

The 1.0 distfile server has not been available for a couple years
now, there is no copy on nftp, and this package is part of one of
the texlive meta packages, so it would be good if it was fetchable.

Changes since 1.0:

2015-07-07  Peter Breitenlohner  <peb@mppmu.mpg.de>

* Makefile.am: Better dependencies for 'make check'.

2015-06-16  Peter Breitenlohner  <peb@mppmu.mpg.de>

* dvidvi.c: Drop <kpathsea/config.h> (isascii via KPSE_COMMON).

2015-04-17  Peter Breitenlohner  <peb@mppmu.mpg.de>

* dvidvi.c: #include <kpathsea/config.h> (for isascii).

2014-06-16  Peter Breitenlohner  <peb@mppmu.mpg.de>

* Makefile.am: Drop the obsolete ACLOCAL_AMFLAGS.

2013-06-28  Peter Breitenlohner  <peb@mppmu.mpg.de>

* dvidvi.c: First include <config.h>.

2012-11-19  Peter Breitenlohner  <peb@mppmu.mpg.de>

* Makefile.am: Avoid use of deprecated INCLUDES.

2012-05-29  Peter Breitenlohner  <peb@mppmu.mpg.de>

* dvidvi.c: Drop the problematic prototype for isatty(),
already declared in <unistd.h> or <kpathsea/c-fopen.h>.
Cast arguments to avoid MSVC compiler warnings.
From Akira Kakuto <kakuto@fuk.kindai.ac.jp> (W32TeX).

2011-09-25  Peter Breitenlohner  <peb@mppmu.mpg.de>

* dvidvi.test: New shell script for tests.
* tests/{play.dvi,playout.2}: Test input and expected output.
* Makefile.am: Add the test and its data.

2011-09-23  Peter Breitenlohner  <peb@mppmu.mpg.de>

* dvidvi.c (writedvifile): Write newline at end of stdout.

2011-04-18  Peter Breitenlohner  <peb@mppmu.mpg.de>

* dvidvi.c (writepreamble): Cast string literal to 'unsigned
const char *' to avoid 'differ in signedness' warning.

2011-02-07  Peter Breitenlohner  <peb@mppmu.mpg.de>

* Makefile.am (EXTRA_DIST): read.me has been renamed README.

2010-05-10  Peter Breitenlohner  <peb@mppmu.mpg.de>

* dvidvi.c (SET_BINARY): Remove casts to void.

2010-02-12  Peter Breitenlohner  <peb@mppmu.mpg.de>

* Makefile.am (AM_CPPFLAGS): Remove '-DNO_DEBUG' ...
* configure.ac: ... now added via KPSE_KPATHSEA_FLAGS.

2009-08-21  Peter Breitenlohner  <peb@mppmu.mpg.de>

Avoid maximal compiler warnings.
* dvidvi.c (oname): Constify global string.
(error, putstr): Constify string arg.

2009-06-22  Peter Breitenlohner  <peb@mppmu.mpg.de>

* Makefile.am (AM_CPPFLAGS): remove -DNeedFunctionPrototypes.

2009-06-09  Peter Breitenlohner  <peb@mppmu.mpg.de>

* Makefile.am (AM_CFLAGS): enable compiler warnings.
* dvidvi.c: declare most function as static and define them
with ANSI C prototypes. #include <unistd.h> to declare isatty.
Remove unused variables.
Remove SHORTINT (to select '%ld' or '%d' format specifiers),
all relevant variables are integer (typedef'ed as long).

2009-06-02  Peter Breitenlohner  <peb@mppmu.mpg.de>

* dvidvi.c: always include <stdlib.h>, to avoid implicit
declaration of calloc, exit, free, and malloc.

2009-05-27  Peter Breitenlohner  <peb@mppmu.mpg.de>

Adapt to TL2009 build system.

2008-02-29  Peter Breitenlohner  <peb@mppmu.mpg.de>

* configure.in: added AC_CONFIG_HEADERS (moved
from ../kpathsea/common.ac to here).

2008-01-29  Peter Breitenlohner  <peb@mppmu.mpg.de>

Convert from autoconf-2.13 to autoconf-2.59+:
* Makefile.in: Add @configure_input@ line.
* aclocal.m4 (new): Generated (aclocal -I ../m4).
* configure: Regenerated (autoconf).

2004-12-30  Peter Breitenlohner  <peb@mppmu.mpg.de>
* dvidvi.c: Changed return type void of main() into int.

(wiz)

Update python3 state comment.

(wiz)

Add upstream bug report URL.

(wiz)

Fix build with readline-7.0 and depend on it.

Bump PKGREVISION.

(wiz)

Remove url2pkg marker.

(wiz)

+ cups-2.2, libprelude-3.1.0, libpreludedb-3.1.0, prelude-correlator-3.1.0,
  prelude-lml-3.1.0, prelude-manager-3.1.0, tor-browser-6.0.5,
  vim-8.0, wine-devel-1.9.19.

(wiz)

Update HOMEPAGE.

(wiz)

Recursive bump for cups openssl -> gnutls change.

(wiz)

Updated print/cups to 2.1.4nb2

(wiz)

cups removed openssl support in 2.0b1.

Switch to gnutls.

Reported by Stephan <stephanwib@googlemail.com> on tech-pkg.

Bump PKGREVISION.

(wiz)

Updated converters/orcus0.12 to 0.12.0nb1

(wiz)

Add libixion dependency. Add missing includes to bl3.mk.
Bump PKGREVISION.

(wiz)

Add bl3.mk file for lbixion.

(wiz)

Added math/libixion version 0.12.1

(wiz)

+ libixion.

(wiz)

Import libixion-0.12.1 as math/libixion.

Ixion is a general purpose formula parser & interpreter that can
calculate multiple named targets, or "cells".

The goal of this project is to create a library for calculating
the results of formula expressions stored in multiple named targets,
or "cells". The cells can be referenced from each other, and the
library takes care of resolving their dependencies automatically
upon calculation. The caller can run the calculation routine either
in a single-threaded mode, or a multi-threaded mode. The library
also supports re-calculations where the contents of one or more
cells have been modified since the last calculation, and a partial
calculation of only the affected cells need to be calculated.

(wiz)

Updated security/gnutls to 3.5.4nb1

(wiz)

Use libopts from autoopts package instead of local copy.
(Only changes bin/*, not lib). Fixes build when autoopts already is installed

Disable valgrind explicitly.

Addresses issues reported by Ricard Palo.

Bump PKGREVISION.

(wiz)

Updated net/librsync to 2.0.0nb1

(wiz)

librsync has a crash in rs_search_for_block documented here:
https://github.com/librsync/librsync/issues/50

Forwarded by Andy Carrel in PR 51841.

Bump PKGREVISION.

(wiz)

Updated net/tigervnc to 1.7.0nb3

(wiz)

Updated tigervnc to 1.7.0nb3.

Add missing mcookie dependency.
Switch back to xorg-server-1.17 since 1.18 made Xnvc dump core
on startup.

(wiz)

Added x11/mcookie version 2.28

(wiz)

+ mcookie.

(wiz)

Import mcookie-2.28 as x11/mcookie.

mcookie generates a 128-bit random hexadecimal number for use with
the X authority system.

(wiz)

evert previous.

(wiz)

Revert previous.

(wiz)

Add upstream patch so one test passes.
Replace bash binary path in more shell scripts so more tests work.

Result: no failing tests. Yay!

(wiz)

Updated meta-pkgs/texlive-collection-latexextra to 2016pre3

(wiz)

Add two. Bump version.

(wiz)

Add two new tex packages.

(wiz)

Rename package to match PKGNAME, and adapt bl3.mk.

(wiz)

Added converters/orcus0.12 version 0.12.0

(wiz)

+ orcus0.12.

(wiz)

Mark as 0.10 branch to make difference to orcus0.12 clearer.

(wiz)

Import orcus-0.12.0 as converters/orcus0.12.

Standalone file import filter library for spreadsheet documents.

This package contains the 0.12 branch of the library.

(wiz)

Updated net/filezilla to 3.21.0

(wiz)

Updated filezilla to 3.21.0.

3.21.0 (2016-08-23)

! Fixed a string format vulnerability introduced in 3.20.0-rc1 when listing directories using SFTP
+ SFTP: Added support for AES-GCM ciphers as implemented in OpenSSH
+ OS X: Ctrl+Tab and Ctrl+Shift+Tab can now be used to switch forward and backward between opened tabs
- Reduced time between TCP keepalive packets to 15 minutes on supporting platforms
- Fixed saving of directory listing filters with an attribute condition

3.20.1 (2016-08-03)

- Fixed rename file exists action on downloads
- Fixed possible crash if disabling log abbreviation
- Official binaries now link against a patched version of GnuTLS so that a better error message can be printed on broken servers that send malformed certificate chains

3.20.0 (2016-07-27)

+ Display error message if entering a non-existing path on local file search
+ Building and running FileZilla now depends on libfilezilla >= 0.6.1 (https://lib.filezilla-project.org/).

3.20.0-rc1 (2016-07-20)

+ Added compatibility for filenames with leading or trailing whitespace if using SFTP
+ Building and running FileZilla now depends on libfilezilla >= 0.6.0 (https://lib.filezilla-project.org/).
+ Building and running FileZilla now depends on GnuTLS 3.4.0 or higher
- Fixed regression introduced in 3.19.0-rc1, reconnecting again uses the last used remote directory instead of the initial default remote directory
- Fixed crash if creating a new site via the bookmarks dialog
- Queuing remote directories for transfer no longer exits comparison mode
- Fixed a rare crash using FTP over TLS if the control connection fails at the same time the data connection gets established
- Entering invalid regular expressions in filter and search conditions now shows an error message
- Fixed title of search dialog
- Stricter certificate chain validation to supplement the Tofu model
- *nix: Fix initial size of Site Manager dialog with some GTK versions

(wiz)

Recursive PKGREVISION bump for gnutls shlib major bump.

(wiz)

Remove another obsolete patch.

(wiz)

Remove another obsolete patch.

(wiz)

Revert previous.

(wiz)

Remove two obsolete patches.

(wiz)