http://stackoverflow.com/questions/7406946/why-is-apple-deprecating-openssl-in-macos-10-7-lion
http://stackoverflow.com/questions/7475914/libcrypto-deprecated-on-mac-os-x-10-7-lion

Apple apparently got fed up with the unstable ABI of OpenSSL, or
something, and recommend you either move to their "Common Crypto
Architecture", or, if you have to stay with OpenSSL, just build it
yourself statically and bundle with your application. So let's do
that.

Change-Id: I5f19fcce9b4e00ff2ce3089a2398b3f08f76a9a0

Change-Id: I7e69466f189b845a807edbf3ca4422b0fb2aec94

Patch-by: orw
Tested-by: orw
(cherry picked from commit 7a8c7f6a)

Conflicts:
	svtools/source/filter/jpeg/jpeg.cxx

Change-Id: Icda84dad6b5a8395f8592fdba479b194d4cc7ff7

Change-Id: Ibc60b487fc769f152d231d962d984e3c8f6dce50

(cherry picked from commit e4a73933)

Conflicts:
	comphelper/inc/comphelper/stl_types.hxx

Change-Id: I18587958e7458cdd6847f48e8ea41ca92afd7815

Change-Id: I66de69f3ca4ff92abb0583cd330087d59a037a0f

for symmetry also the other constructors in the same class
and mark their destructors as virtual if they are anyway
(cherry picked from commit 870349ea)

Conflicts:
	vcl/inc/vcl/button.hxx
	vcl/inc/vcl/combobox.hxx
	vcl/inc/vcl/ctrl.hxx
	vcl/inc/vcl/dialog.hxx
	vcl/inc/vcl/dockingarea.hxx
	vcl/inc/vcl/event.hxx
	vcl/inc/vcl/field.hxx
	vcl/inc/vcl/fixed.hxx
	vcl/inc/vcl/floatwin.hxx
	vcl/inc/vcl/fontmanager.hxx
	vcl/inc/vcl/group.hxx
	vcl/inc/vcl/lstbox.hxx
	vcl/inc/vcl/menu.hxx
	vcl/inc/vcl/menubtn.hxx
	vcl/inc/vcl/metaact.hxx
	vcl/inc/vcl/morebtn.hxx
	vcl/inc/vcl/pngread.hxx
	vcl/inc/vcl/scrbar.hxx
	vcl/inc/vcl/spin.hxx
	vcl/inc/vcl/spinfld.hxx
	vcl/inc/vcl/split.hxx
	vcl/inc/vcl/syschild.hxx
	vcl/inc/vcl/tabpage.hxx
	vcl/inc/vcl/virdev.hxx
	vcl/inc/vcl/wrkwin.hxx

Change-Id: I8c29f74208cf382a9baa01c2e2d0757688cb4641

Project: dictionaries  61644291cba7f6f8dd7e194f5abbe0b0e146a321

solve fdo#63033 update Dutch dictionary OpenTaal to version 2.10g

Change-Id: Ia4ccb06b0db4cdd976d2ec4491e14b82eefa558b
Reviewed-on: https://gerrit.libreoffice.org/3844Reviewed-by: Andras Timar <atimar@suse.com>
Tested-by: Andras Timar <atimar@suse.com>

Enhanced EditEngine to be able to hold more than 64k paragraphs. Used
also in RTF import Calc and Writer, so that could benefit as well.

* changed all EditEngine,Outliner,... related paragraph index/count
  variables from sal_uInt16 to sal_Int32
  * sal_Int32 instead of sal_uInt32 to match accessibility API
  * matched some Outliner methods' paragraph parameters from sal_uLong
    to sal_Int32
  * containers capable to hold size_t nevertheless are limited to a
    maximum of sal_Int32
* changed definition of EE_PARA_NOT_FOUND and EE_PARA_ALL to
  SAL_MAX_INT32
+ added EE_PARA_MAX_COUNT and EE_TEXTPOS_MAX_COUNT to initialize
  ESelection with what previously were hard coded 0xFFFF all over the
  place
+ for similar reason added EE_TEXTPOS_ALL corresponding to EE_PARA_ALL
  to initialize an ESelection spanning all available text like
  aSel(0,0,EE_PARA_ALL,EE_TEXTPOS_ALL)

Change-Id: I62d1b9c399cd709a4f93dbac31b219a61c46ec00
Reviewed-on: https://gerrit.libreoffice.org/3838Tested-by: LibreOffice gerrit bot <gerrit@libreoffice.org>
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>

on cygwin, configure forced TMPDIR to /tmp, regardless whether it is set
or not. And while it respects an existing value of TMP, TMP is only used
in a few places. Changes the ones that explicitly check for TMP to
prefer TMPDIR, and don't duplicate configure's check in other scripts.

Change-Id: Ia8ac3f8d92e5f971c5237b85f66f7c0a58932930
Reviewed-on: https://gerrit.libreoffice.org/3701Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: David Tardon <dtardon@redhat.com>

Project: help  9c6ee8419cc4e7ff0d5b3cfbd9fc0e538e78f8ab

Change-Id: I276fa2fa24b5553a748e6e628129d54419e69c9a
Reviewed-on: https://gerrit.libreoffice.org/3843Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: David Tardon <dtardon@redhat.com>

The basefield flag of a stream is persistent, which I didn't realise before.
This fixes commit 256e4264

Change-Id: Ic0076c63ae6236ae75dcd79d108add27391b2125
Reviewed-on: https://gerrit.libreoffice.org/3842Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: David Tardon <dtardon@redhat.com>

Patch-by: hdu, orw
(cherry picked from commit 7b699076)

Conflicts:
	vcl/win/source/gdi/salbmp.cxx

Change-Id: I106efb3960cb99367d2ecefb8bcae006f389e271

Corrected static_cast which does not work on Win compiler

reinterpret cas needed (base classes are not related)

(cherry picked from commit 99e4d3a5)

Change-Id: I02fc89d1ab346231985c68d63d9710c036d2ab2a

Change-Id: Ia3e12c18a30e5aa0f37db5035cd308b06b2b5655

Capitalize them consistently with the reference being their UNO names defined
in main/helpcontent2/helpers/uno-commands.csv

(cherry picked from commit 6a2ea0d8)

Conflicts:
	svx/source/tbxctrls/fontworkgallery.cxx

Change-Id: I3400c68bf121f4a95cf1068c62be9864257c8ac2

Change-Id: I1aa89a8d6f712abb4f768a413ebac29e14612484

* Remove EMFP_DEBUG with OSL_DEBUG conditional defines and SAL_INFO
* While we are about it, to help with troubleshooting metafile
  issues, add another vcl.emf area to log-areas.dox
* Improve error handling when processing an EMF header:
    + replace variable name nsal_uInt32 (!!) with sane variables
      to make the code more readable
    + check to ensure that type field is 0x1, which is all it can be
      for metafiles
    + check that signature field is set to ASCII-encoded value "FME"
    + loose check of version field to see if it is 0x00010000
    + warn if record count is zero - that really shouldn't be possible
    + check bytes field in header to make sure it correlates to the
      actual size of the metafile
    + additional check - [MS-EMF] states that reserved field should be
      zero, and ignored... but we can use this to check for file
      corruption - not 0, not correct!
* Quite a few more comments in the code to clarify the intended
  structure of a metafile, per [MS-EMF] documentation

Change-Id: Id4ed486b2dd0c6e7bdee67cb344aaaf8e8d98f84
Reviewed-on: https://gerrit.libreoffice.org/3839Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: David Tardon <dtardon@redhat.com>

Change-Id: I5af86e15bcb8958a680e7309f13d7a865f29d7a9

Change-Id: I3812b60f097b1daf554bf2d337b126e8a8067573
Reviewed-on: https://gerrit.libreoffice.org/3837Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: David Tardon <dtardon@redhat.com>

unify the various c-string compares and hashes.

(cherry picked from commit b7e3470a)

Conflicts:
	oox/inc/oox/export/shapes.hxx
	oox/source/export/shapes.cxx
	sal/inc/rtl/string.hxx
	sdext/source/minimizer/pppoptimizertoken.cxx
	svx/source/customshapes/EnhancedCustomShapeTypeNames.cxx
	vcl/source/glyphs/gcach_ftyp.cxx
	writerfilter/source/resourcemodel/TagLogger.cxx
	xmloff/source/draw/EnhancedCustomShapeToken.cxx

Change-Id: Ib742744077bfb4d38a462d88b44bdef45601b4ae

use simple text hilight frame when captions are in TextEdit mode

(cherry picked from commit bdcd01c3)

Conflicts:
	svx/source/svdraw/svdmrkv.cxx
	svx/source/svdraw/svdorect.cxx

Change-Id: Id570a220873d1bbd0f9caed4db44f90ca2f67d9e

Otherwise, the FILELIST scp2 items would create a program directory even on Mac
OS X, before gid_Brand_Unixlink_Program creates the program -> MacOS symlink.

(gb_Pyuno_PROGRAMDIRNAME shoudl be renamed to something more general than Pyuno,
and maybe more */Package_*.mk would benefit from using it.)

Change-Id: Ib6fc8d1abaf3b1a47a0f026a62bf0abdb14fc521

Change-Id: I42f54c87384419f5f605c2a5a8146477238f4dfb

It seems not to be a too good idea anyway - the image has to have some
specifics, so the value of setting any random image as the background of
Writer / Calc / Impress is questionable.

Either way, if anybody wants to finish that, easy to recover from the
history ;-) - better than keeping the unused code around.

Change-Id: I5079210bd91b5e70ce20681e5477b860bb28ff07

Project: dictionaries  c178908939c5e5c727f5c7c78b9cc32356f6ea46

Change-Id: Iddcf995b079867cf14454d28c363efacc0323ae8

Project: dictionaries  7c495c0a113c52ba9f2b15a5ab38d06b02e1c8df

(cherry picked from commit feb1be19)

Change-Id: I9e6ddcf04047bdc905f00b709db84459112122db

before converting to SpinbuttonValue or ScrollbarValue

Else problems like #i112873# or even crashes can and do happen.

(cherry picked from commit d2fde1d3)

Conflicts:
	vcl/unx/kde/salnativewidgets-kde.cxx

Change-Id: I6429205ab41b3194577bc50aa02384c920e11ed2

Change-Id: Id3cc26b6fbb1ec0ddacf7228e1c147085db415e1

Change-Id: Ie5168a9e67ab9479cf7b6d0e350ec0b3c4f0616f

In a case we use harfbuzz on Windows once, it is using in the
uniscribe backend some OpenType shaping that is available only
in Vista and later. So call this functions over function pointers
if available and if not, fallback to the itemizing/shaping/placing
that is available in Windows XP uniscribe.

Change-Id: I5b1a22d812d57194d1147effa87033cdf6af78d3

Change-Id: I1167d2ebeba97380b65c6e7789ba7fc3ad6b0a1e

Change-Id: I1c68a1fd425d54e2d01ff63580e62dc0fb6416e2

Change-Id: I31b95844234135d14f117ae74124350eedf7d20a

It should have been nExtDescent + pMetric->mnDescent, but then the whole
thing does not make any sense; why add the descent to the leading and
then calculate the leading by subtracting the descent! (well, the ATSUI
code was doing this but it makes no sense either).

Just use CTFontGetLeading() directly.

Change-Id: Ia54648f6c02c11359865f4aa6476adf40b27f906

Change-Id: Icba7557140d65b769bef17df57e2cdc1bdba3193

Change-Id: I544760548866d55807a3291a31b4ce948a751946