Change-Id: I43a78b0a24d92362100cf9fab956120f01843222
Reviewed-on: https://gerrit.libreoffice.org/7516Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Michael Stahl <mstahl@redhat.com>

Change-Id: I1ff7d4bad068a8bc8cdc7f27dbfaee199d0f2e5f
Reviewed-on: https://gerrit.libreoffice.org/7513Reviewed-by: Michael Stahl <mstahl@redhat.com>
Tested-by: Michael Stahl <mstahl@redhat.com>

Change-Id: I44c4ee0ca9c9986b1479eb2a365d50319771b304

Change-Id: I07cbdf3745ac8b2522397b8f013d9dcec8a894df

Change-Id: I46ad2bee7be304e84622af88bdd975963dcaf7e7

What I am doing today?! - sorry for the incomplete push.

Change-Id: Idca25b390820527481b1c672a0ad62f2d83a0e13

Double lines are always drawn as 2 parallel hair lines that are 1 pixel
apart, at any zoom level.

Change-Id: I2796477d0ea45c9880aa8057bd1a10104df96673

Without this, some dashed lines may not get drawn at all at some zoom levels.

Change-Id: I273c1548325d14f56618df8ca4166aac58a3ff3f

Change-Id: Icfbb295abb19cf58477f4f14f4a7294a540151c2

Change-Id: I0437409b6a5d6163fadf777df5c028950727e786

Change-Id: I53d5f8b0278d1228cd941221a07cd360943c5ce6

Change-Id: Idb352dc2afeb2ee7b48620c972c2a186209228ea

Change-Id: I93d3613547b74a37cebbb6e791e0d23ea405401b

Now the dashed lines are evenly placed on screen.  For now, horizontal lines
only.  I'll work on vertical lines later.

Change-Id: I474e9c8214e5f079ea2cfca12b35381d8fcf2ae1

This makes slightly skinnier solid lines which look better on screen.

Change-Id: Ia7764be4a53d1dd6bb60ecb3ba5c8966403e4e6c

Change-Id: I6dccc887b400352cc47678d912b5e4d391c96a3b

Change-Id: I87a6c1c1c6dc92670dccee3f56302e7f234b281d

Change-Id: I5087320137dc17f27a382d564a59cec68f1a4f95

Change-Id: I9ad8798b205e697a588804de3de758be7203f1a5

There needs to be a dependency from the header on cppumaker, because
headers do get rebuilt in this case and it's possible that the content
chnaged; unfortunately it's not possible to tell whether the content
of the headers actually did change, all of them will be touched which
will result in a large rebuild.

Change-Id: I78ae0631ba9197ca3d3edbd010319b77a6f01074

There used to be a dependency between .urd file and .hpp/.hdl files but
that was removed; introduce a dependency between .idl files and
.hpp/.hdl files to replace it so headers and stuff depending on them get
rebuilt.

Change-Id: Ibf1fe1db08c36ce42c392d27b794c2bc2a33738c

ScDocument::GetString() already returns the formatted string that then
was formatted through number formatter. Format raw string instead.

Change-Id: I806b4fc904d04169a6c1dbd4a3d00e2556eda7bd

Change-Id: I759d119e0ad6c5ae2881284c82292d4d9cb7ae87

It does not make a real sense to use constructor for implementations that act
as a trampoline; instead, we should do a constructor for the real
implementations that hide behind it.

This reverts commit 2fbb4715.
This reverts (part of) commit 58ea2712.
This reverts (part of) commit 194bdbde.
This reverts (part of) commit df002e39.
This reverts (part of) commit eb89c6f7.
This reverts (part of) commit bdeb57c2.
This reverts (part of) commit 4337a066.

Change-Id: Ibd9b1066f3b6ea215d8da9f491d5f026cd75b029

Most of the constructors are supposed to be only a call of

  new TheInstance(arguments)

or an equivalent; so let's just change the constructor caller accordingly, to
accept unacquired new instance.

If there are exceptions that need to do more heavy lifting, they do not have
to use the constructor feature, or there can be a wrapper for the real
implementation, doing the additional work in their (C++) constructor.

Change-Id: I035c378778aeda60d15af4e56ca3761c586d5ded

...now that ModuleUIConfigurationManager no longer implements XInitialization
since 72b5343b "fwk: Use constructor feature for
ModuleUIConfigurationManager."

Change-Id: I479c6c93e9982d80318e53b794bac99e949a5e46

cf. the list of platform tokens at <https://wiki.openoffice.org/wiki/
Documentation/DevGuide/Extensions/Target_Platform> and in
desktop/source/deployment/misc/dp_platform.cxx.

Change-Id: Ibfc103bed1f6756654225e2c456ad73ab56c0d5a

...rather, pass them from the build to the SDK installation via dk.mk.  (The SDK
installation is already platform-specific anyway, see the binary executables in
its bin/ directory, so there is no point in not fixing this at build time.  In
the future, the shipped settings.mk could of course be stripped to just the
parts relevant for a given platform, of course.)

Along the way, fixes some apparent "ppc" vs. "powerpc" typos in settings.mk.

This is a prerequisite to create a working settings.mk for macosx_x86_64, but
does not yet do that.

Change-Id: Ib8e0c03ad6dfd610aae58a61e32bdc61c785584f

Change-Id: Ia3152b7cc5edd89f37c78841e10bb5feacc6fc12

Project: help  934de4f8116eb69e621b7eff487e743a8a1866e2

Change-Id: I6066d42abddcefc9d7fe4bd19477a11294b138b2

Change-Id: I56e161d42f7a58a7931191ab3d0af79f2af9ee5c

This took a bit of code archaelogy for me to track down. It turns out that
as part of an effort to optimize startup, the line of code in this function
was commented out, presumably to be rewritten later. This happened in
commit ee3351d7 in July 2001 (!).

About three years later, in February 2004, the function was rewritten in
commit 189c2388. As it turns out, the only two functions that used
the Window parameter were vcl/source/window/window.cxx (which passed
itself in, but of course this did nothing) and
vcl/source/window/winproc.cxx. Furthermore, winproc only ever passed in
the first frame, so it didn't really do anything either.

Consequently, the function as it stands *now* only notifies the
application that system settings have been changed. It doesn't care
which window it tells.

Therefore, I have excised this parameter from the function. I don't
think it made sense when it was implemented anyway, so there is no net
loss.

After removing the unneeded parameter, I was also able to remove the
Window parameter from winproc.cxx's ImplHandleSalSettings function as
it was only ever used to set the top level window, which is now
irrelevant.

Change-Id: I84f2c5c5ff8969387da3af81e4a9c7f9ac6237e1
Reviewed-on: https://gerrit.libreoffice.org/7541Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Michael Meeks <michael.meeks@collabora.com>

I'm not exactly sure why existing code didn't do this implicitly, but at
least this improves the situation.

Change-Id: Id2bb169c513827b7ef48640dc88fad90a83d2bee

Change-Id: If15c208f23d09dbd55ad0625448f864a7725c553

Change-Id: Iee8d757f29c03e71e48c211966daaa89548710b4

(cherry picked from commit 5c6ef989)

Change-Id: Ida372f0558fd7e5cdb9c98384c21544e197f88a9

Also note that OpenSSL is incompatible with -fvisibility=hidden on OS X

Change-Id: I1fc3cf4d515ef210f996cb04ba9c1b3216109137
Reviewed-on: https://gerrit.libreoffice.org/7233Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

Change-Id: I4b744cf10165383153d2a71c05df0c0ed327c641

Change-Id: Ibac028120286c9deb2411fc547877eebac93f36f