Change-Id: I86aaad9f38ad63121805fa3dc05b3a8a33428ef7

Second or later compilation uses value type returned by previous execution of code.
Use the defined type as return value of the runtime function of Basic always.

(cherry picked from commit 7470c682)

Conflicts:
	basic/inc/basic/sbxmeth.hxx
	basic/inc/basic/sbxobj.hxx
	basic/source/runtime/stdobj.cxx
	basic/source/sbx/sbxobj.cxx

Change-Id: I3064e8403286a9c1401ef658bf139bedeae11f17

Change-Id: I17936ae2e37520abaa7dd31a5bb9aec6300ea021

Change-Id: I01816988eb3c47e60826ff01c1b8402f9f5130e4

and consequently fix bug in SwPageFrm::PreparePage where it was
comparing against the wrong #define's

Change-Id: I681f7e9f3f9bbe1ddf2613814ed36cfe0955825f

Change-Id: I1c761482cb78efc26d3beda75381557268e1bc0d

Change-Id: I7aaeda37ceea080251f9d6381a78d13b8fb055d0

Change-Id: I163bd3f521473c5357c23c3ce4978f984be7b6fe

$(call gb_Library__get_name,foo) returns "libmerged" if library foo is merged.

This reverts commit ee567a63.

Change-Id: I6ab9b7f0b01262a6f9d5a6834a6cffdd6ffc6f8a

That's what the comment on it said before it was translated.

Change-Id: I8812f99253bd4a14cb1109abf72b11d650f56003

Change-Id: I4d0e1de89d0bbdbea23bc5a46bf75ae0ce4e2796

It is scary how many barely educated C++ programmers can't tell their
inclusion polymorphism from their ad-hoc polymorphism.

Change-Id: Id4df7d168ba5257eff522a7f131c841c1ddcbccd

Change-Id: If69a57aa9ff011cc670f868cb87b3a1c4d904435

This reverts commit 340ca294 as CppunitTest_basic_vba
breaks, so needs revisiting to see which is right

Change-Id: I10168d454a8c67d91eee5bb8939aff74bba42954

Change-Id: I6d83714b78c4d149dff5ea56483d7afa6d0eb677

(cherry picked from commit a2700f75)

Conflicts:
	basic/source/runtime/methods.cxx

Change-Id: If9c671a612de87d8d37bfc40704a97a6062fd150

cf. <https://gerrit.libreoffice.org/#/c/13659/3> "Signed/Unsigned mismatch on
x86_64 platform on windows"

Change-Id: Ie526ee431f59be374d61fe8c38bdf2a579a7234b

avoids the dread assert in vcl wrt. "detect overflows" in ImplLogicToPixel of
vcl/source/outdev/map.cxx but the whole thing looks a nightmare if a high
precision double is given as an arg to Fraction and the underlying boost foo
uses longs longer than 32bits to represent that.

Change-Id: Ia0fa911cf07ed613e0b1bbd97b63a291f29628cb

This reverts commit 66c7dbbf.

See https://gerrit.libreoffice.org/#/c/13554/ .

Conflicts:
	include/xmloff/shapeexport.hxx
	reportdesign/source/filter/xml/xmlExport.cxx
	sc/source/filter/xml/xmlexprt.cxx
	xmloff/source/draw/shapeexport.cxx

Change-Id: I14fed9ad043a756a9e2d390c76fa907c349c3a2e

Innovativeness is good but perhaps not for mathematical constants.

Change-Id: Icef93802ce9123120c8527f3fa07e524b6e067d0

Change-Id: Iffc8cbf108310099318e37378c4b3033ea087cee

UNROLLING_FACTOR (or UNROLLING) has been defined since November 2013. No need
to keep old code around inside ifdefs. We have this thing called version
control.

Change-Id: I1f10f205df95014f8c223bd7a8716757577eade0

USE_FMIN_FMAX has been defined a long time. There is no information in any
comment or in the commit that introduced it why one would not want to use the
OpenCL fmin() and fmax() built-in functions (intrinsics). Keeping such stuff
in the source code (and in the generated OpenCL code) is pointless and just
makes it harder to read.

Change-Id: Ibc0b950c0f16872867eb5eff089b33b00cb95028

MD5_KERNEL has been defined always for a long time, and I don't see why we
wouldn't want that functionality.

Change-Id: I7ee220ac3354fb8e5f6556259147675847aab68d

NO_FALLBACK_TO_SWINTERP has been undefined for a long time. I doubt anybody
knows any more what it means, and in what situations it might perhaps make
sense to define it. Should it be a compile-time choice anyway, in that case,
or a run-time one? What is a "TDD run"? Test-Driven Development? But we have
our unit test mechanism for that.

Change-Id: If3e21e958ebeb689a8aac8eb7393c02f7143401c

...follow-up to 4e41227d "mingw64: change Link
class to use sal_IntPtr instead of long;" should make obsolete
<https://gerrit.libreoffice.org/#/c/13659/> "Signed/Unsigned mismatch on x86_64
platform on windows"

Change-Id: I9fcc947bb96e104816c2820cb338fc34c528bae0

Change-Id: I55db1220e234b92f5d0a3794f416154495da8b69

Change-Id: Id69fb63e8f77a1f26491fc2a5d50d17a4d6421eb

Revert part of 8e417a8f that
looks unnecessary:

this is used in ScXMLDataPilotMembersContext::CreateChildContext() which
is below table:data-pilot-level which actually uses the
data-pilot-member element, whereas the data-pilot-group element occurs
below table:data-pilot-groups.

Change-Id: I7a3183bd3dbfa98b3bdf5e581b700910025af965

see ServiceType::dumpHxxFile and SingletonType::dumpHxxFile in
codemaker/source/cppumaker/cpputype.cxx

partial revert of 45ec1d9b "brute-force
find-and-remove of unused #define constants."

Change-Id: I3c3d10f2d81926ad0b86330d46b8bf52416bc361

...apparently unused since the corresponding code was removed with
2ad02390 "#87132#removed obsolete parameter in
executeProcess" in 2001

(detected after 45ec1d9b "brute-force
find-and-remove of unused #define constants" boldly removed the corresponding
osl_Process_DF* defines from osl/process.h)

Change-Id: I49c20aec9b4acb08d42bfa0371ee1149d6c4e9d4

...they had not been documented in osl/process.h and the sal/osl/unx
implementation had been stubbed out since
7b37265b "sal: remove unx implementation of
osl_sendResourcePipe/osl_receiveResourcePipe" in 2012

Change-Id: Ia3ae853d95b6f3b2d2743f06755ef8f6246501d8

as seen with fdo55736-1.docx

Change-Id: Ie1dec06282a600e600f606a3b4ddd814613b0108

see thread starting at
<http://lists.freedesktop.org/archives/libreoffice/2015-January/065706.html>
"CppunitTest_sw_ooxmlsdrexport: division by zero in
chart2/source/controller/chartapiwrapper/LegendWrapper.cxx"

Change-Id: If8fde31dff04d9a223687838df6dd9fe4e57ee8b

Change-Id: Ic63daa16e6931dd7842faae0d3c83ae0fdeb132f

Change-Id: Ifdbcbcc51c29dc14b2a3b4c845481b45300313d6

Change-Id: I39eafa22b12e62c766a182c2ebc2b115084f4cef
Reviewed-on: https://gerrit.libreoffice.org/13231Reviewed-by: David Ostrovsky <david@ostrovsky.org>
Tested-by: David Ostrovsky <david@ostrovsky.org>

Change-Id: Idae98f65bd87d08c62b27c5c604e4d98fcacbe1e
Reviewed-on: https://gerrit.libreoffice.org/13660Reviewed-by: David Ostrovsky <david@ostrovsky.org>
Tested-by: David Ostrovsky <david@ostrovsky.org>

The code is really not in an ideal state, the key-frame generation is
horrible - involving playing the video for three seconds each time etc.
While the idea is great - the impl. and libvlc API use needs some love
from a VLC expert really.

Change-Id: If1787749d9952e30c1d80846e925cf1f65231463

Change-Id: Id5dd1ee1a29c4e1c7cb2b58419d6ccb1f032bffe