Change-Id: I9743fd98691695f831fa282a64482b23178f0026
Reviewed-on: https://gerrit.libreoffice.org/3444Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: David Tardon <dtardon@redhat.com>

Change-Id: I99660704931c7341a72fe303fae47f984f968ecb
Reviewed-on: https://gerrit.libreoffice.org/3443Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: David Tardon <dtardon@redhat.com>

Change-Id: Iab9cb3704385c115e1d25189a3d8833bd6721e24
Reviewed-on: https://gerrit.libreoffice.org/3440Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: David Tardon <dtardon@redhat.com>

 - replaced osl_trace with sal_info
 - replaced dbg_* with sal_*

Change-Id: Ie5d3ae7d2c5dbaaed30d0b39740748845c9f1641
Reviewed-on: https://gerrit.libreoffice.org/3423Reviewed-by: Thomas Arnhold <thomas@arnhold.org>
Tested-by: Thomas Arnhold <thomas@arnhold.org>

Change-Id: Ic7fa577fbdb527a75643c4e2919e0527432d5536
Reviewed-on: https://gerrit.libreoffice.org/3439Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: David Tardon <dtardon@redhat.com>

Change-Id: I76ed89c4a46e5a7b6a639e2c2717bc53d9b68bc9
Reviewed-on: https://gerrit.libreoffice.org/3438Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: David Tardon <dtardon@redhat.com>

Change-Id: I6c071f00eef88d3f56e76efddf6f0c71aa56b089
Reviewed-on: https://gerrit.libreoffice.org/3437Reviewed-by: David Tardon <dtardon@redhat.com>
Tested-by: David Tardon <dtardon@redhat.com>

There's so little to be gained by storing pointers to simple
or refcounted types instead of using an instance. At least
SwTxtNode::GetMinMaxSize() and SwTxtNode::GetScalingOfSelectedText()
pass string temporaries to SwDrawTextInfo, so keeping a pointer
is clearly broken, and MSVC manages to optimize the code enough
to crash because of this. As all the pointers point to const references
and are accesses using const references as well, there shouldn't
be any harm in making a copy.
This fixes smoketest crash on Windows.

Change-Id: I7d7cb42e175a2e64fe9c84c98c0d0204a0c25c13

Change-Id: I1912c9550c12a971fcc7fdbc8bd623f18ccc40b8

Just the outmost dialog itself, not the individual pages.

We have to go top down in this case to incrementally convert the pages because
of the non-standard additional widgets outside the tabcontrol area.

So this dialog is set to non-resizable until its contents are
converted over.

Change-Id: I92014a44b4cab8d345666295b052635d4b57a496

Project: help  0483b91db6d66dbe320ce8a129988a6ed8940832

Change-Id: I23cf85f5fbd2dcc333094e9003ce8f5ce4f0b68d

Project: help  0065ae78b41544cda55fddddb0cf4e776da56e4d

Change-Id: I2a464208081b42a577aafbe321d9662007208d77

Change-Id: I8268ac340f62a84bfae774194e8a6a50c8a707f3

Change-Id: Icabdaab665ac969c9ba6240d2e825d1e62776249

Change-Id: I6c983ec5f0f4597f7f86c1c9ff256881c360bb1d

Temporary directory name upon temporary directory name. Just in
case. When in doubt, add one more temporary directory name layer.

workdir/unxmacxx.pro/installation/LibreOffice_Dev/dmg/install/en-US_inprogress_temp.62979/./LibreOfficeDev_4.1.0.0.alpha0_MacOS_x86-64/LibreOfficeDev.app

Couldn't quickly figure out how to construct the
"LibreOfficeDev_4.1.0.0.alpha0_MacOS_x86-64" part so just use a
wildcard * for that for now;) There won't be any other directories
there anyway, as far as I know.

Change-Id: If950d4cdfe3f7b0663f1b6fbd62185ca8284dd7f

Change-Id: Ief5e5c89b7fb69fb9849cf2d6efe2b4c5b7f5391
Reviewed-on: https://gerrit.libreoffice.org/3375Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Eike Rathke <erack@redhat.com>

... although i'm not sure if the stub implementation of
pthread_cond_timedwait makes that much sense.

Change-Id: If1926e760c0e1b84fd2ac48431c06c6f1ba4dd1d

The vertical scrollbar remains a bit whacky in how it scales and operates when
scrolled, this is unchanged.

Change-Id: I468a87c75f1c1927d0387c4af64c7ad87f4e1922

Change-Id: I490c5bbddaefbdf06f33fbcfc86e862441b05a00

Change-Id: I8413f2ee3bcd700409f6bfa3886ba91c2fa2b921

... which has the same stupid mistake.

Change-Id: I128af77f0b75d2b60f5bb4b33e55c2bc0c94ae98

Change-Id: I8fce7d3b3da2a6dad373409f9c9ba3b9cd06aa4f

...otherwise, it will fail with "c:\program files\microsoft
sdks\windows\v7.1\include\GdiplusTypes.h(470) : error C3861: 'min': identifier
not found" errors.

Change-Id: Ib81ab7f5357b1f980202b63b144296fa252b71b1

This can apparently happen when interrupting a build; reproducible with
"touch workdir/*/UnpackedTarget/*openssl* && make openssl".

The -f parameter would apparently silently ignore some mal-formed
patches but at least it rejects potentially reversed ones,
which causes the "error handler" in gb_UnpackedTarball__command
to touch the "prepare" target so the next make is successful.

Change-Id: I7bbd7d9385d990a69214a3a2d9bb20b5a7173748

... clearly David should have reverted my commit since it didn't
actually work :-/

Change-Id: Ib34e540e247f1082bc9d155cd011eb31cd33048e

...otherwise, it will fail with "c:\program files\microsoft
sdks\windows\v7.1\include\GdiplusTypes.h(470) : error C3861: 'min': identifier
not found" errors.

Change-Id: Ie58a5e85e6712e4244ce4799c203b31a717a91e6

"exception not thrown in catch block"

Change-Id: I13f8be9913f655ab1633b29e9fe686a41c2c0fcb

Change-Id: I0f693266b7c3621f885acbe450f4faead3d8284f
Reviewed-on: https://gerrit.libreoffice.org/3429Reviewed-by: Thomas Arnhold <thomas@arnhold.org>
Tested-by: Thomas Arnhold <thomas@arnhold.org>

Nowadays in a dbgutil build with MSVC, the debugging runtime (C++
library) is used. Just like with the GNU libstdc++, it is not binary
compatible with the non-debugging variant.

All sources that go into one DLL or EXE must be compiled with the same
_DEBUG settings (unless one is extremely careful and knows exactly
what one is doing, which presumably is not the case with this code).

With previous compilers this has gone undetected, but with MSVC2012 at
least, you get link-time errors, which is a good thing:

nssinitializer.o : error LNK2038: mismatch detected for '_ITERATOR_DEBUG_LEVEL': value '0' doesn't match value '2' in biginteger.o

nssinitializer.o : error LNK2038: mismatch detected for 'RuntimeLibrary': value 'MD_DynamicRelease' doesn't match value 'MDd_DynamicDebug' in biginteger.o

Whether this actually has caused run-time erroneous behaviour or
not in a dbgutil build is another thing...

Change-Id: I0c42161f00a0e37b438e3b911e3c0058edadd800

Change-Id: I98d04d7da4c2b7ea0b769df9c2bfa8c1ad86bf2d
Reviewed-on: https://gerrit.libreoffice.org/3422Reviewed-by: Thomas Arnhold <thomas@arnhold.org>
Tested-by: Thomas Arnhold <thomas@arnhold.org>

Change-Id: Ice80350184f7a514d5beab0a5e1da5b98d5733e4
Reviewed-on: https://gerrit.libreoffice.org/3427Reviewed-by: Thomas Arnhold <thomas@arnhold.org>
Tested-by: Thomas Arnhold <thomas@arnhold.org>

Change-Id: I19a346a35eef403a7f831299e42655b9761f2c43

Change-Id: Ic2f743643f8a825a8f485b68025717e09f3d8c2f

Change-Id: Ie32b2b1283b54719f4c9525105d534fce70219af

Change-Id: I16db60c861ef7ac0a2a9076fde1996a7b7b0e991

Change-Id: I43c2ad4a601ffe20f471d81ed7697b89c496fa6f