note that the autodetect HC option is somewhat broken anyway:
it resets the icon theme hard, so there are not simple roundtrips
enableing/disabling it for that, but those havent been there before
either.
Change-Id: Ia35a41717224dfb7437054bb885c61d7e0b189d7
It can thus be used in a pkg-config replacement wrapper that makes pkg-config
look in an appropriate place, which typically is needed when not using a
proper cross-compilation toolchain. (Such would come with an own copy of
pkg-config.)
Change-Id: I4c746206cc67ed7a428833f03b9d3aad250d31e1
This also prevents a weird problem I see when cross-compiling in one tree (but
not in other trees for other cross-compilation targets).
Change-Id: I102e652c259a66ea124b344bd229e8988801ac58
- dark theme default to high contrast
- as per fdo#35365, having a dark document background is inconvenient
for non-a11y endusers
- a11y standard require the (rather ugly) background for a11y
- thus, when "automatically detect high contract mode of operating
system"
in Tools->Options->a11y is enabled, use the dark document
background
by default, otherwise use a white default
Change-Id: Id8ad1eb3d57b3708ac5a241092208e3a8d98ade0
This hopefully is a model which we can build in further
to remove the rest of the direct SID... stuff and the fragile
mapping from the known .uno: commands to SIDS to dispatch, and
instead just dispatch the .uno: command directly
Change-Id: Iec0f92123fa4c2dae15e0ac716d5ef687e67da79
the sidebar widget-layout enabled panels could use frame::XDispatch::dispatch
to send their commands directly rather than mapping back to SIDS and Executing
those, which would allow removing piles of weird-ass stuff
Change-Id: Ibbff56d4fb96820d3bdbf4b1cb582d25337fe48b
Only adjust tokens for non-shared formula cells and shared formuls cells
that are the top of their respective shared groups.
Change-Id: I9cd57653602c97504a802001947a28c4d03368a5
This reverts commit ef37781683, now that
constructor function support for singleton implementations is in place and the
proper singleton theGlobalEventBroadcaster has been introduced (and drops
sfx_component_getFactory completely, which had become unused except for
SfxGlobalEvents_Impl in the meantime).
Change-Id: I2d587989dd0a1fbea6fe75f7dc2af954ac81c4ed
The service manager now keeps track of instances of singleton implementations
(i.e., implementations whose XML description lists at least one
<singleton ...>). These instances will be disposed either when the service
manager is disposed, or, for instances that have been instantiated into the
component context's /singleton/* map, when the component context is disposed.
This change allows to use constructor functions for such singleton
implementations, too.
Change-Id: I220c9ddc9824e4d7f7556daefb599e2ec36b0e6c
add minimal padding around slide preview (to not touch actionbar and
bottom of screen - horizontal whitespace will be larger becasue height
is limiting factor anyway)
avoid adding two left-paddings for the notes - the Text is "indented"
compared to the notes area already, so margin implied by the layout
weight distribution and the text indent is enough to separate it from
the preview.
avoid adding two right paddings for the note-text, as the text is not
justified and the word-break itself will account for a "padding" at the
edge, and the container already has a right-padding (that was reduced,
since the word-padding as well as centering in the unused space will
increase the effective margin)
Change-Id: I2a4605ba5a98eeed93a01db16d3d86c90df56470
Reviewed-on: https://gerrit.libreoffice.org/7230
Reviewed-by: Andrzej Hunt <andrzej.hunt@collabora.com>
Tested-by: Andrzej Hunt <andrzej.hunt@collabora.com>
by replacing the onClickListener with an onTouchListener with the simple
variant of the GestureDetector.
Also drop the check that prevents going to the "end-of-presentation,
click to exit" slide and exiting the presentation that way.
Change-Id: I54b49bf11929ad9415b8c85581fe16998ab3a7a7
Reviewed-on: https://gerrit.libreoffice.org/7107
Reviewed-by: Andrzej Hunt <andrzej.hunt@collabora.com>
Tested-by: Andrzej Hunt <andrzej.hunt@collabora.com>
it is just another obfuscating layer of one-command-function calles that
can as well be called directly.
If you really want to get the path to aquire the bluetooth adapter
differently, based on a runtime check, a single method to retrieve the
adapter is enough in any of the classes. No need to wrap the whole
adapter's api in your own class.
Change-Id: I2c631321dcf8ef143fe58a0a8246e010169409ac
and don't loop autodiscovery, provide a manual trigger button instead.
Bluetooth discovery is a hefty process and thus should not be done
lightheartedly. Moreover discovery won't even list devices that
are already bonded, but not set to visible. As you will be more likely
using the remote with the same devices, it makes sense to list the
known devices and only do discovery on explicit request/when there are
no bonded devices yet.
Fix a lifecycle problem (as the service would be quit on
screen-rotation as the only bound client is destroyed/restarted - start
the service instead to let it keep running, and only stop/release it
when finishing)
icons from Android's ActionBar Icon Pack, shrinked using optipng
Change-Id: Ie8467f942df1aab2d64b337fc7a6f816b9d658a6
Reviewed-on: https://gerrit.libreoffice.org/7091
Reviewed-by: Andrzej Hunt <andrzej.hunt@collabora.com>
Tested-by: Andrzej Hunt <andrzej.hunt@collabora.com>
it is more straightforward to only have one viewpager that can be
flipped through. This makes it easier to restore the user's default way
of connecting (bluetooth via wifi) and also simplifies the setup and
the what-tab-am-I-on checks.
* Remeber what tab (wifi/bluetooth) the user last used and restore that
on next launch
* respect Android's guidelines and ask the user whether Bluetooth should
be enabled when the user switches to the BT tab and BT is disabled.
→ if the user declines, select wifi tab instead
Fix a lifecycle problem (bt connection would be cut if bt was not
enabled before launching the remote, only restore disbled state if
really finishing, not on configuration change like rotating the screen)
Change-Id: Ice3a5c877a2a4810a80a0f76edea713700fe9c8c
Reviewed-on: https://gerrit.libreoffice.org/7090
Reviewed-by: Andrzej Hunt <andrzej.hunt@collabora.com>
Tested-by: Andrzej Hunt <andrzej.hunt@collabora.com>
there's no point in adding a layer of indirection if all that the called
function does is running one single command, and is only called in one
place. Getting rid of that indirection makes the code easier to read and
understand.
Change-Id: Ie2f3e03fe2870d1d4a84df738ebb3d0f34a2713c
Reviewed-on: https://gerrit.libreoffice.org/7089
Reviewed-by: Andrzej Hunt <andrzej.hunt@collabora.com>
Tested-by: Andrzej Hunt <andrzej.hunt@collabora.com>
This will be superceded by the following sdremote improvments, which
would otherwise have path conflicts due to this patch.
This reverts commit 3cc31f8978.
Change-Id: I49f004d068fdf852f5690e365a17168b001b9136
UI wise would makes the app compatible with android api level 7,
but other utility functions require higher level (8 for Base64 and 9 for
TimeUnit)
explicitly set the allowBackup flag and raise tested/targeted version to
api level 17
also add tool-annotations to please android-lint
Change-Id: I528e34acdeeecea6d20e8bea21b1d5c203e17c95
and only touch the start of the given range.
(cherry picked from commit ae295f7d009842cdceb50c4daffe948ede2b4b88)
Conflicts:
sw/source/core/doc/docbm.cxx
Change-Id: Id7317eeb8e9c063c9d8b30bca97ed0afee3ec8c5
SvTabListBox::GetEntryText(pEntry, 0xffff) will append 10 tabs to the
returned string, so it will no longer be equal to a style name.
(regression from 77173d8954)
Change-Id: I7d3bee47eabec9180d6352fb6cf1b2e948153d37
This is in preparation for more conversion of SvStream::operator<< calls
to use more explicit method names.
This converts the subclasses that have their own convenience overloads
of operator<< to use normal methods.
Change-Id: I5efd5d9a24c264cb86d2471303dd5849bf91ba80
