Commit by jlvergara :: r1176406 krita/trunk/koffice/krita/plugins/paintops/defaultpaintops/smudge/kis_smudgeop.cpp:

Solve a bug introduced by the last commit, it was caused by misaligned borders in m_wholeTempData. My apologies for the carelessness.

Commit by jlvergara :: r1176363 krita/trunk/koffice/krita/plugins/paintops/defaultpaintops/smudge/ (kis_smudgeop.cpp kis_smudgeop.h):

Fix a design flaw in the new smudgeOp behavior that caused it to be too slow at small sizes.
o This was caused by using m_tempDev->extent() to decide which parts of the device containing the dab needed to be iterated over.
o Given that m_tempDev->extent() uses 64x64 tiles, at small sizes (i.e.: 5x5) it caused the brush to be 100 times slower.
o Using m_tempDev->exactBounds() was not an option becase it is slow.
o Therefore the maximum size the dab was cached and updated as relevant (when a new step of the brush brings a larger rectangle in), then used.

Commit by jlvergara :: r1175649 krita/trunk/koffice/krita/plugins/paintops/defaultpaintops/smudge/ (kis_smudgeop.cpp kis_smudgeop.h):

Last facelift to smudge op:
o Definite algorithm to solve bug 229012 decided, alternatives removed from comments.
o Comments extended and beautified (includes description of the algorithm at the top).
BUG: 229012

Commit by jlvergara :: r1175627 krita/trunk/koffice/krita/plugins/paintops/defaultpaintops/smudge/ (kis_smudgeop.cpp kis_smudgeop.h):

Make QPoint m_extractionTopLeft a simple variable, it doesn't need to be a class member

Commit by jlvergara :: r1175603 krita/trunk/koffice/krita/plugins/paintops/defaultpaintops/smudge/ (kis_smudgeop.cpp kis_smudgeop.h):

Commit new version of the smudge op paintop that preserves the old behavior at constant brush size, but attempts to fix problems when the brush is resized mid-stroke (example: using stylus).
o Includes cap to opacity to prevent smudging from turning into copying.
o Includes commented-out alternative algorithms for testing (each with their own drawback, like the default one).
o Includes some code style changes and extra comments.

• copyright notice added to kis_smudge_op.h CCBUG: 229012

Commit by jlvergara :: r1175590 krita/trunk/koffice/krita/image/kis_transform_worker.cc:

Add fallback safety line at KisTransformWorker after Q_ASSERTS in case Krita is compiled without them

Commit by jlvergara :: r1175584 krita/trunk/koffice/krita/image/ (kis_painter.h kis_transform_worker.cc):

o Fix parameter positions in some bit blt functions in KisPainter.h.
o Add missing apidoc there.
o Add Q_ASSERT in KisTransformWorker when someone tries to run it with xscale or yscale of 0.

Commit by jlvergara :: r1175259 krita/trunk/koffice/krita/ui/kis_aboutdata.h:

No log message

Commit by jlvergara :: r1175112 krita/trunk/koffice/krita/ (7 files in 6 dirs):

Shipping KisPainter facelift.
Includes renaming of bitBltFixedSelection to bitBltWithFixedSelection
And bltFixed to bltFixedWithFixedSelection

Commit by jlvergara :: r1173222 krita/trunk/koffice/krita/image/kis_painter.cc:

Added comments to the new kis_painter.cc fix for clarity:
Magic number 0 in mergedSelectionBytes[i] = 0 was labelled UNSELECTED in the comments and its use (initialize data) clarified.
The selectedness variable was eliminated and instead a comment block explaining the reasoning was added. (I have a small concern that this piece of code may not be portable, do all compilers and architectures evaluate that IF statement without crashing when d->selection is Null?)
totalPixels was renamed to totalBytes which is a more proper name. It is also a cosmetic variable (used only once) but I think it's worth keeping because it's self-documenting.

Commit by jlvergara :: r1173136 krita/trunk/koffice/krita/plugins/paintops/defaultpaintops/smudge/kis_smudgeop.cpp:

Erase selection bug workaround from the source, it is no longer needed after svn commit 1173130

Commit by jlvergara :: r1173135 krita/trunk/koffice/krita/plugins/paintops/hatching/kis_hatching_paintop.cpp:

Erase selection bug workaround from the source, it is no longer needed after svn commit 1173130

Commit by jlvergara :: r1173130 krita/trunk/koffice/krita/image/kis_painter.cc:

Commiting fix for bug 246639 where bitBltFixedSelection didn't respect selections, but all the paintops using it expected it to respect them. The patch had 2 revisions (thank you Lukas and Dmitry).
Patches to Hatching Brush and Smudge Op reverting to their old behavior (using bitBltFixedSelection without bug workarounds) will follow.

Commit by jlvergara :: r1171519 krita/trunk/koffice/krita/plugins/paintops/complexop/kis_complexop.cpp:

Fixed bug 245143 (Mixing Brush crashed painting outside of canvas). Also removed helpful debug lines placed probably by Dmitry as they're no longer needed.

Commit by jlvergara :: r1170555 krita/trunk/koffice/krita/plugins/paintops/waterymixbrush/ (16 files):

Cloned and butchered the hatching brush to form the basis for a mixing brush with dilution effect. Algorithm to be implemented is described in the source. This clone currently doesn't do anything interesting, but it doesn't crash.

Commit by jlvergara :: r1164411 krita/trunk/koffice/krita/plugins/filters/phongbumpmap/ (6 files):

Completely linked the GUI and created a new method to process all the new information to create a bumpmap
Many very beautiful color patterns can be achieved and the versatility of the filter increased greatly.
However, this newfound beauty came at the cost of severe mental retardation, the filter is now ~50-100% slower when processing the same information as before. Optimization is pending (the only way out I see is overloading the function down to oblivion, increasing speed and making the class very very verbose). Despite the slowdown the filter still feels 'real time' in most of the cases.
Found a very strange, annoying and offensive bug that crashes the filter when a light source is unchecked in Effect Mask mode. I found that this behavior repeats in my Krita install for other filters used in Effect Mask mode, and could be due to a damaged system. System reinstall pending. Bugfix pending. (The crash doesn't affect the filter when applied as such). Will report bug if the crash persists after reinstall.

Commit by jlvergara :: r1161123 krita/trunk/koffice/krita/plugins/filters/phongbumpmap/wdgphongbumpmap.ui:

Renamed -illuminant(s)- to -light source(s)- by user request. Added What's this information to Spin Boxes to complement the tooltips. Added 'Azimuth' tool tip to the Dial to make it very clear that it's just another way to control the Azimuth angle.

Commit by jlvergara :: r1159872 krita/trunk/koffice/krita/plugins/filters/phongbumpmap/ (6 files):

Linked most of the GUI. The last 2 colors need a properly implemented function to make use of them. Controls work whether enabled or disabled at the moment. \n Fixed glitch in the GUI that duplicated the controls. \n Changed defaults to give a nice molten gold effect, you better check it (use the deform brush with an adjustment layer of this filter for the most dramatic results). \n Added new file containing constants to reuse in my files (trying to improve readability). \n Solved the memory leak in phongbumpmap.cpp. \n Added honorary copyright notice to a certain person who has done much to assist me on the filter until it reached its current stage.

Commit by jlvergara :: r1159790 krita/trunk/koffice/krita/plugins/filters/phongbumpmap/ (5 files):

Made phongbumpmap able to be used with an adjustment layer without visual artifacts and at usable speed (sometimes the calculation is faster than the paintop modifying the color channel). Added unlinked and slightly buggy GUI. Completely converted the phongbumpmap algorithm to be tiled, with a tile size of 100, this remake made the code more readable too. Use of QImage still present.

Commit by jlvergara :: r1158825 krita/trunk/koffice/krita/plugins/paintops/hatching/kis_hatching_paintop.cpp:

Small repositioning of code, editing comments.