Change-Id: I5e1c66e09418e534bb47233f45c67206af20cab4

Change-Id: Ib0d4f542986dc09968cad8b76da9d6e034eddd37

Tests RAND, SUM, AVERAGE, SIN, COS, TAN, ATAN, SQRT, EXP, LN at the moment.
More will be added.

Does not yet automatically add to black/whitelist.

Change-Id: I2ca24632630937ee5559fc86453e40178e89973f

Nah, will generate the test sheeet(s) programmatically.

This reverts commit 9c2cf556.

Change-Id: I8d331914736e17d5aad2ee761fc6d758c2dea1ad

Change-Id: Ifd3fca29325dc18166d041badd5664d2ad8ba7c9

Change-Id: I39cd9ef32fafa2894ffde961d87adc6aac9d4a4f

~X11SalVirtualDevice() was destroying X resources referred to by
OpenGLSalBitmap's mpContext, which were subsequently accessed by
Bitmap::AcquireReadAccess on the floating Bitmap. Better to use the
default window's GLContext for all bitmap operations.

Change-Id: I9009980e791cff1a1f36d626592d72c7a32efd39

Change-Id: I2605175e7ee66d3da5e41497baa4178074830e99

Project: translations  6453172eac4a30a37224db66ddd76c13f614959d

Change-Id: I33ffbe87b7c7f2f1fefc18f10055e47e2454ead1

Change-Id: I0c3d6d99f478b86873d108c893798687032b79bc

Change-Id: Ic91023c0a5a70072fcffaaa5e16fc2e415c2e1ee

Loading this spreadsheet and verifying that the end result in A1 is "PASS"
will be bound to the 'Test' button on the calculation details dialog.

Actually, I wonder whether it might be saner to just populate a spreadsheet
programmatically instead of having a (not huge, but still large) test file
bundled in the installation? We could make it so that the generated document
stays open and is usable normally, so that if necessary (for instance for
comparison with other spreadsheet applications) it can be saved.

Change-Id: Idea972cd0f598643e35314576f593047d49378c6

Change-Id: Ib5683580273430a61e4daff84cd03b3b559fac0f

Change-Id: I41ba46831f24b2960a1fe982b74a2b623e682e0b

Unfortunately iwyu gets quite confused by the weird cyclic dependencies
between various foo.h/foo.hxx and cppumaker generated headers, so it's
not obvious if any improvement here is realistic...

Change-Id: I0bc66f98b146712e28cabc18d56c11c08418c721

Change-Id: Ib21a3cfe4b02e421f8848a6e2cca4e2a1c04adb9

Change-Id: I0d84545527195e0f8a48d4217cb332f2e46f2d09

Change-Id: I28c8f134825e477ffe993a6895b801c9e895f613

Change-Id: Id6c096d09b984039d54a1d152cf0da594aaf203d

And call the new method StartListeners().  This also adjusts what was
previously StartAllListener() to be group-listener aware.

Change-Id: I74de45c00f5b8ef232eea9fe3b93aa44d1d8855b

See:
c454a0cd
and
46eb0952

Change-Id: I811f1c52f8a9fba28bc4f38fd1ad34826b2bd45e

Change-Id: I3f7e0ccd5697a4ca6026bec60210e24af25fb92f

Change-Id: Ifcf30f209355d4cea4daeabe06896b0c225a65a0
Reviewed-on: https://gerrit.libreoffice.org/12958Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

- not yet included in the make for windows

Change-Id: Iee31b0ed0c6545572295ce00a3bb0f909c428b5a
Reviewed-on: https://gerrit.libreoffice.org/12425Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

to this custom widget to state that is is a candidate for tabstopping to, and
that its a (old-school) container widget which has subwidgets which should be
candidates too

Change-Id: I39452f906b30c3962026ab9c4dba625ecd2d06a8

Commit d7223774 (abi#9930
DocxAttributeOutput::AddToAttrList: avoid duplicated attributes,
2014-11-02) added a quick fix to avoid writing duplicated attributes,
but the validation problem of duplicated elements remained.

Make sure that when writing paragraph marker properties, then we write
either the hints at the end of the paragraph or the character attributes
range of the paragraph properties, but not both.

The only exception is the character style that's always set as
a hint, as explained in commit 4bb872b1
(DOCX export: fix handling of paragraph mark on empty paragraphs,
2014-10-17).

Change-Id: I494e5bb9871aa535532fef32bd344d8093e9f762

Change-Id: I90c559c59b307aac6f7adac49ea38b0d9b68ad58

Implementation uses a multiplicator of ten. I guess the intention
here was that to calculate cache size based on the size of the inserted
images and not based on the real memory usage (potentionally more
instances of one image), which seems a bad idea because it covers the
real memory usage from the users (who will be suprised when LO's memory
usage increases tenfold of the set value).

It seems better to use the exact value in the implementation and increase
the default value on the user interface instead.

Change-Id: Ie3c0568d3518922d3a9931f894eebb9e7e5515ed

Change-Id: I1752857bc036f8d389b4b71d570cf83a76ef29ea

we're still listening to the old one here when we set a new one

Change-Id: Iaf64ba17a7adf2489a4f0475bba03a2500a556c0

Change-Id: Idac493dfb3239f7ae5b5e49158d79c4f3e707509

Change-Id: I83cca6a1bd3a462a277e5f5338cafbcd037d4227

Change-Id: I897f71e5ae4d2368d5d3ced60d90fb1c833fa376

which removes the guessing you have to do when trying to find the edge of a
pres_obj in impress after text has been filled in and there is no visible
border anymore

Change-Id: I470caec90cdeaa681da70f8ec3044c1d6d232a82

Does not yet do anything, but will essentially run a "unit test" of the OpenCL
functionality required by the default subset of opcodes.

Change-Id: I230f6232c90562ea56057347fb0dc088f98c215d

Change-Id: Ia80682aeb87225b9bde7398186e121b1d3bdc2ad

Attach is the new one we are going with & offers nice contrast with
'Detach...'.

Change-Id: I4578f4c7b9d989e41d433d7c8aa96a9317aa6919

Change-Id: I5b362ad374dad5fd4a79b8a7706defb749e25eb3