Issue 73

Set file size stored in header to 0 when rebuilding cache. Fixes a crash when switching back and forth between mmap- and file-modes.

Fix inverted tab navigation. Now it works as expected fully with keyboard. If no icons are provided or invalid, do not show them (we only would be getting question marks)

Follow RFC concerning ORG property.

The patch will still read the old Department values if they exist but store them correctly.

Fix plugin behavior. Please read thread on kwrite-devel mailing list with subject "Plugin on multiple views behavior"

Yay, playback over sftp:/ works again: the job would emit result and delete itself and the KioMediaStream has to recreate the job when it wants to seek back. :'(

Templates:
- Don't fail when the selection is empty and the pipe command is invoked
- Pipe the entire body through the command instead of just the selection.

This ensures that the body is correctly piped when nothing is selected.

Don't allow removal of the last identity. Bad things will happen.

Fix Czech character-to-ASCII translation for Generic Media Devices. Patch by Mate (J Laitl).

Make starting chats work.

This took entirely way too long, but starting a new chat from the model view contact list is working now. Time to merge. ;)

Fix and improve searching on keyserver

started fixing edit mode: You can now move some things, but much is still very broken. It is now at least usable though :)

Implement all missing parts of the expression-parser, which means: All types of casts, new-expressions, string-literals, and numbers are parsed quite correctly now.

Include the highlighting support, its basically working.
Commit on behalf of Piyush who has rather flaky internet connection at the moment.

Add a new handwritten lexer - only a stub until now. Real work will begin.
However everything compiles with the new lexer and thus I'm comitting to have proper tracking of the future evolution of the lexer.

Add keyword and identifier lexing, follows the scheme that Roberto used in his kwgen tool

Am now giving reasonable (but not always perfect) progress percentages when reading and re-reading stars.dat. The percentages should now end near 100% even if we are not reading in the entire stars.dat file.

The new data file starlnum.idx has the line numbers of roughly 90 different star magnitudes which enables the more reasonable percentages. If the stars.dat file changes then starlnum.idx should be regenerated via:

$ ./mag-linenum.pl stars.dat > starlnum.idx

Have stopped the one draw cycle delay before re-reading star data because the skymap is getting erased anyway.

Tracked down a several second delay after reading the star data to a faulty implementation of QTextStream::pos(). This function can take 7 or 8 seconds to complete! I have a feeling they go back and re-read the entire file. I implemented a trivial work-around by simply keeping track of the file position myself inside our read loop:

m_lastFilePos += line.length() + 1;

Also did some minor code clean up in StarComponent and KSFileReader.

implemented a level management for the verbs

added (optional) units conversion support with libqalculate (currently only for scalar fields).

removed the need for a png file for gps positioning, and did a little bit of commenting... and thats my last commit for GSoC i'll be back soon

initial implementation of wikipedia integration

Added "open in browser" toolbutton for wikibrowser
Added "open wikipedia in external browser" option
Fixed bug with broken loading of some pictures

First version of the kalgebra's console

Initial implementation for global error estimation (still only statical and only for Particle and Spring)

Initial word type management class. Rewrite of most of the common entry dialog page. Usage labels do not yet work in the common dialog.

Static and dynamic error estimation (works for Spring, Particle and EulerSolver).

Piecewise working on evaluation and simplification.
Yey! KAlgebra can calculate the fibonacci sequence (don't give him numbers too high, please!)

Here you have it :P fib:=n->piecewise { eq(n,0)?0, eq(n,1)?1, ?fib(n-1)+fib(n-2) }

Initial implementation of softbodies (currently only square softbodies).

Made examples installable.
KNewStuff support (actually that was previous commit, I just forget to mention it).

Adding batch modes for the Julian Day and Sidereal Time calculator modules.
Before testing, please compile and install the latest ExtDateTime library.

I seriously doubt I am going to have time to implement batch modes for all of the calc modules in time for 4.0 (feature freeze is in a couple of days). Should we leave in the few that have been implemented, or disable them all for the 4.0 release?

Added new tool: Tracer
Implemented statistical variance calculations for gas macroscopic quantities

Complete rewrite of the type handling.
This means the addition of a new class to keduvocdocument: KEduVocWordType.
Word types are handled by getting a pointer KEduVocWordType* KEduVocDocument::wordTypes().

The class provides functions for adding types and subtypes and also has a special type QString to prevent users from deleting types necessary for certain query types (verb for conjugation etc.). These types can be renamed still.

Within KVocTrain a few type_relation and other curious functions go away. When this is done with usages as well, a few more interdependencies will go away.
To handle types in KVocTrain a new model and treeview has been added.

Note: next step is to adapt the old reader/writer. And move the compability maps and functions away from KEduVocWordType.

At document creation time a good preset (maybe the old types) should also be created.
Also the old typeDescription QStringList has to be removed.

More gradient work, mostly to support 8bpp grayscale gradients but some other cleanups as well.

Some of you are probably wondering why I'm spending effort on the gradient code now that Qt supports them. The most important reason is the unbalanced gradients. In this case "unbalanced" refers to the line of the gradient, not it's color transition, and can't be replicated with Qt gradients. For example, if you make a diagonal gradient with a positive unbalance factor it will appear to curve to the bottom-right. This can be used for emulating a shadow or highlight and other neat stuff. You can't do this with QLinearGradient, which will always draw the gradient on a line.

The other reason is also the reason for this commit. Some applications blend two images, or an image and a color, using a gradient. That's what most of those blend() methods in KImageEffect were for. Now you can get 8bpp grayscale gradients by calling grayGradient() or grayUnbalancedGradient() and use them for an image's alpha channel. These are drawn natively and don't create a 32bpp temporary image. This is far more efficent than what you would otherwise do with Qt.

The other reasons are minor. The gradients should be slightly faster but that really matters only if your doing something like using them as a background for a resizable widget and need to recalculate them often. Another reason is the oddball gradient types like Pyramid that some applications use, but probably can be depreciated. I'm not gonna suggest that, tho... I already unintentionally started one flamewar suggesting we depreciate things >:)

Anyways, this commit adds grayGradient() and grayUnbalancedGradient(), moves the gradients to their own source file to be better organized, and has a few minor cleanups.

- Add support to retrieve the color model and bit depth of the original file (before conversion to RGB)
- allow to load only metadata fields and not the actual image data
- Add a method to retrieve a unique hash from a file. This hash is very much like what Amarok uses for file identification, and the purpose is to be fast and unique.

Similarity detection is _not_ possible with this hash - for this, we will use a Haar-like algorithm.

The implementation of the the unique hash and the selective loading, see last commit

Draft of new schema:
- add status flag to AlbumRoots
- colorDepth is an integer (8, 16)
- if you look closely, compression information is not available for JPEGs, PNGs, RAWs etc., actually only libtiff returns something here. So this is not a field suitable for the db.

Make the sidebar able to show the text for the items;
it can be shown/hidden using the popup menu of the sidebat itself.

TODO: save the state across runs.

digiKam from trunk (KDE4) : Add XMP metadata support.

- You need to use Exiv2 0.16 to handle XMP. This Exiv2 version is under development.
- A new XMP tab have been add in Metadata sidebar.
- If you use an old version of Exiv2 witch not support XMP, this tab will be hidden.
- XMP handling is read only. We only show XMP content. Advanced operations over XMP will be done when Exiv2 implementation will be more mature.

Add a preliminary implementation of a CollectionManager that supports multiple album roots and uses Solid for detecting if a particular path is available or not.

Still missing code to add and remove roots. Code is in #ifdefs, not tested nor used.

- Added 10x13.33cm into a4 (provided by Joerg Kuehne)
- Added A6 size to get the old configuration for 10x15cm page
- Changed page size to real 10x15cm instead of A6 (need to set up right page size)
- Added a new page size 13x18cm (need to set up right page size)

session management support

Add initial support for managing multiple output layouts (still unfinished)

Plasma kget data engine and plasma kget applet added

too big diff - no way to commit when not connected to the internet :-(

- VideoWidget on destruction tells xine WILL_DESTROY_DRAWABLE (it was in VideoWidgetXT before which was not the drawable)
- change Events namespace to Event class (QEvent subclass and baseclass for all events used in the backend), the Event objects have a refcount now
- let all down-/upstreamEvent reimplementations call the Source-/SinkNode implementations for event object cleanup
- add code to allow copying custom Event objects
- copy the event before posting it (postEvent takes ownership)
- in SourceNode::downstreamEvent increase refcount of events for each sink it
sends it to
- add safeguard in XT classes to check for early deletion
- add code in XineStream to send arbitrary events downstream
- keep a ref to all XT objects that were mentioned in a connectionChange for 3 seconds to give xine some time to finish the rewiring which happens at some undefined point after the rewire call.

- implicitly deinterlace video when playing a DVD
- make EVENT_SPU_BUTTON (DVD menus) work again
- remove all "<< endl"s
- stop adding a new unusable virtual device per run

the task oriented plugin skeleton

Start to add support to save/read icon and sort icons

Added a couple features to bluetooth at solid/control. They are the needed elements to handle remote device authentication (pairing) and local service usage authorization.

Please review.

RTL languages support.

First lines of the new search dialogs (don't expect much searching yet)

Add a method to render round box with a dark edge and use it for alt-tab's frame.Now the frame has a slight edge/border around it which makes it look much better IMHO.

The redland backend is now really thread-safe and properly handles iterators.
Sadly the sesame2 backend seems not to do that although it was the original trigger of the Iterato::close() method.

Replace next implemented, dialog fixes

implement contrast, hue and saturation; send middle and right mouse button events to xine as well

add plugin manager to kate part. still work in progress and some strange behaviour. hope to fix it later.

kpluginselector fork to katepluginselector. adaptation follows later, once the kte plugin unloading does not crash anymore

Replace all, shortcut letters, both dialog's width made equal for Kate, focus stuff

<Ctrl+R> working, pattern from selection for power mode, in-source todolist

Add a driverHandle() method similar to the one in AudioInterface to PortableMediaPlayer and Camera.

Allows to get the information needed to access the device with a particular driver without having to rely on backend dependent data.

This way no need to expose low level related methods.

Jeff, in the hal backend I only implemented the support for the mtp driver, that will allow you to completely remove the use of GenericInterface in your ioslave. You'll have to add the support for the other drivers though. Thanks in advance.

WARNING: Please don't commit code depending on those new methods outside of kdelibs before coming monday.

First plugin for btpm, not working.

Changing data connection parameters from hardcoded values to configurable values.

The server now attempts to read, currently depending on OS platform, the following values:
on Windows:
Connection/Port

i.e. looking for this config part:
[Connection]
Port=portnumber

Default for portnumber is the previously used 4444.
Since QTcpServer allows special value "0" for "any available" socket, it might be a better default.

on Non-Windows:
Connection/SocketDirectory

i.e. looking for this config part:
[Connection]
SocketDirectory=rel_to_home_or_abs_path

Default path is ".akonadi", i.e. $HOME/.akonadi
A starting KDE session should probably set this to the user's KDE socket directory if not explicitly set to something else.

After creating the respective server socket, the server will create a config file named akonadiconnectionrc, containing the actually used parameters.

On Windows it will basically look like this:

[Data]
Method=TCP
Address=127.0.0.1
Port=4444

On Non-Windows it will basically look like this:

[Data]
Method=UnixPath
UnixPath=/home/username/.akonadi/akonadiserver.socket

Additonally the server will add the address of its D-Bus session bus to the config:

[DBUS]
Address=unix:abstract=/tmp/dbus-IvFq3ku5bA,guid=23b1bbaca93a129ad21fcb0046c6e1db

This is currently not used on the client side, but would eventually allow clients with a different session bus.

- Implemented the dnd system to assign actions to a given device- Added kaboutdata

Remove hard-coded keyboard shortcuts

As part of the cleanup I started working on a rewrite of the viewer code.
The todo, memo and addresswidget will be rewritten so that they share code that is common for all of them. The first step in this rewrite is to get rid of the PilotComponent code. All components are now based on ComponentPageBase and the viewers will be based on a subclass of ComponentPageBase which is called ViewerPageBase.

At the moment only the TodoViewer is enabled but not functional. The other page should work as before.

And we have records in our list again. This is a first try to but i'm not sure yet if this is the best approach. The previous version had checkboxes in the list for todo's but i think it's not that bad that they are not in there anymore because it's a viewer.

add code to allow rendering of cancelations of previous key signatures when key signature changes. The cancel attribute of key signatures is never set yet, but it can be drawn properly already

add support for saving KoInlineObjects.
It means that now, it's possible to save fields inserted in the document for instance.
Currently, only the page fields are saved, the other are coming soon.

Added the "XML Viewer" python script that allows to view the ODT XML of the current document on the fly.
* This enables fast inspection of how a saved document may look like.
* content.xml, styles.xml, etc. are all handled separate.
* We even provide a "Open with KXMLEditor" button to view the XML ;)

The Painterly Framework is here!
It has the same old Painterly Mixer, and a shiny new Bidi Paint Operation. The Mixer Tool is now based on KisToolFreehand, so it's a lot smoother.
Enjoy the bugs! :-)

add support for saving KoTextAnchors.
That way, pictures are saved at their right position now :)
Thanks jaham for the comments on the mailing list.

BTW, since it's probably my last Summer of Code commit (but of course not my last commit on KWord/KOffice): thanks Sebastian for the mentoring, thanks to the other KOffice guys for their help and support...

(I'm pretty sure there isn't a lot of people reading the commit messages anyway, so thanks to the reader of this message too :) )

Make both widgets use the same code to draw the decorations. Add an extra option to configure the display, I think I've got all bases covered now Make both widgets use the same code to draw the decorations. Add an extra option to configure the display, I think I've got all bases covered now & can start implementing the various paths. Fix blitz include.

Make GridDrawer into KisGridDrawer for consistency. can start implementing the various paths. Fix blitz include. Make GridDrawer into KisGridDrawer for consistency.

Added "Compare..." functionality to the xmlviewer script that allows users to compare the current document with another ODT document using kdiff3.

* kdiff3 is used to show the diff between the content.xml, styles.xml, etc. files of the current ODT as displayed+saved by KWord and those one that got loaded (or just totaly another one).
* this should make it a bit easier to see what changed between a full load+save circle ;)

Ported the basics of the svg export filter.
Text, images and pattern fills are not exported.
Added a function to return the odf/svg path string from the KoPathShape to use it for the svg export.

NEW: Kexi introduces tabbed interface local toolbars
<a href="/issues/2007-08-26/files/mainwindow.png">http://kexi-project.org/pics/2.0/alpha3/mainwindow.png</a>;

Main Window
- tabbed toolbar: delayed tab switching on mouse hover (0.5 sec)
- standard actions work
= fixed "close window" button
- Property Editor is now visible only when there are properties to edit (i.e not in Data View)
- fixed switching between context of current tab (window), so e.g. Property Editor and Find window are properly updated
- Tabbed Toolbar gets Kexi menu and help area
- support for User Mode reintroduced
- fixed presentation of "dirty" state for designed object
- "save" button added

Table View
- fixed appearance of record navigator
- removed unnecessary Qt3 double buffering
- fixed appearance of horizontal and vertical headers
- fixed placement of context menu title, and added one for data view

Commands Chart Type.
Initial implementation.
Changes the type of the chart and should provide an undo action.
Use it in the standalone and in the chart shape.

initial work on the Wikipedia applet. svg included, etc. applet currently isn't called when the engine has new data.

add genre to the service elements. Was worried about how to handle multiple genres per track, but as the collection models always query within a specific genre, it turns out that a track will only have the qenre it has been selected under, even if it occours multiple times in different genres, so this whole issue is basically moot

add start of a service info applet. Has not been tested yet as plasma is FUBAR on this laptop, but builds and should not break anything as long as it is not activated. Also, the svg used is currently just the current track one

Make service info plasma applet attempt to show the html info that is normally shown in the small box at the bottom of services. Does not work very well at the moment

Connect the effects status to the menu and toolbar

Add support to mouse press event and a beta support for the clipboard

update svg, looks better. also, attempting to implement applet drag-n-drop but i'm having issues as the child applets get the events before the parent (the layout manager)

Export public keys to addressbook
Copyright cleanup

Make tab widget remember the current tab

Reimplemented the bookmark editor - MVC rocks!

Small improvements & fixes:
* VNC: use fallback port, if port is invalid or empty
* don't show a protocol that is not available in the protocol combo* don't ask for the username again if it is already defined before (e.g. URL or history)

apps/system-file-manager - And now, Dolphin have its icon. 16x16 and 22x22 need special version.

small versions of dolphin icon. will do 32x32 soonish :)

Feature 104815 : add a search feature to editor

Loading bar now included in tree widget
Config dialog is back (two tabs only for the moment)
Config dialog are provided by LogModeFactories
Convert FileList and LogLevelSelection to .ui filesConvert to Oxygen icon filenames

greatly improved InChI mining
this is a greedy analyzer that reads everything to find InChI chemical identifiers
it can even fix spaces in InChI's. Number of entries and version are also extractedInChI-X is removed because there is no much sense in another greedy analyzer doing same work, I have no samples of inchi-x and this shlould be implemented as a SAX parser based on InChI namespace primary filter

PNG chemical analyzer can process chemical structural information embedded into PNG images as chunks. These chunks are recognized (Molfile, InChI, CML) and substreams passed to indexer children for analysis, making use of the powerful Jstreams and Strigi architecture

This analyzer does not extract any image-related information, this task remains to pngendstreamanalyzer in Strigi.

* GUI: changes to modify task - removed 'silent' button, added 'at system restart' button, added on the fly information wizard
* GUI: changes to modify variable - added on the fly information wizard
* implements @reboot function in GUI

Bug 85425 was closed on a previous commit but I did not know about this system of notification then :)

The closure of bug 85425 sould be credited to Sune Vuorela who originally submitted the patch, I used a slightly modified version of his patch.

Revise the KGoldrunner Handbook for KDE 4.0.

A new theme for kgoldrunner. I call it: "The Treasure of Egypt". Please comment.

Added the black on white "accessible" theme -- this is meant for use on a monochrome desktop; white-on-black, blue/yellow and low contrast themes could be easyly done, if needed. But I was under the impression the plan was to recolor things on-the-fly with composite tricks or something like that.

It would be cool if somebody from the accessibility team could assess it's actually useful, or if changes are needed to make it better from the accessibility point of view.

Further reduced the HTM mesh level from 4 down to 3 which has further improved the drawing speed when zoomed out. A level-3 mesh has 512 trixels total down from 2048 for level-4 and 8192 for level-5.

Fixed a bug in deciding when to calculate the no-precess aperture which caused the equator to sometimes not get displayed.

Added timing code in SkyMapComposite::draw() to time the entire draw cycle. Uncomment line 293 to get the times to print out:

//kDebug() << QString("draw took %1 ms").arg( t.elapsed() ) << endl;

I discovered that the anti-aliased drawing can take about ten times longer than the integer drawing. It is important to turn the anti-aliasing off in order to test the timing of the rest of the system.

I can now draw a reasonable skymap, fully zoomed out, with star labels in about 50 ms. This increases to about 100 ms when I include the solar system and asteroid labels. This then jumps to about 600 ms if I turn the anti-aliasing on.

More modernizinating :) Far more efficient than it was before and from the test app looks like better results as well.

Use mmap() when possible to load the cache files into memory and then use the memory instead of opening+reading/writing+closing the file every time.
This seems to speed up reading icons from the cache by about 30% :-)

However there is a problem when switching between mmap- and file-mode, so it's disabled for now.

greatly simplify the lyrics applet, make it actually fast when moving the applet around. still has some issues with text placement

ok, this probably will be my last commit before "pencil's down" so selecting folders for searching/replacing will be possible only in svn

time for a commit.

lots and lots of changes, including:
* c++ plugins skip over #if0 blocks
* general/contractions much improved and c++-independent
* general/spelling much improved and c++-independent
* krazy2 reads a .krazy at the subdir level (i.e., the same one used by EBN)
* support the idea of "extra" checkers that aren't run by default
* code reuse with Perl modules
* the start of a .krazy POD doc (krazyrc.pod)
* better Makefiles
* and more!

removed delete button from vocabulary settings dialog, since most users dont have access to delete system-installed files anyway, also add keyboard shortcuts to ui elements

Add myself as maintainer, update about dialog.

Work properly on Windows. Never forget that Qt internally always separates paths with '/', regardless of the underlying OS. Also, using QDir we get a better impression on what we are actually doing.

The analitza module is now a shared library

Switching from magnitude spinboxes to slider widgets in the Configure window, for controlling the label density of asteroids and stars.We used to provide a magnitude limit, and all objects brighter than the limit would be labeled.

We now have a more advanced labeler that avoids overlapping text, so a UI control that controls the relative density of labels is more appropriate.I updated all code that interacted with the old controls.

The new parameters are named:
Options::starLabelDensity()
Options::asteroidLabelDensity()

Everything seems to be working as expected.I wasn't sure what range of values would be useful, so right now the range is 0 to 20. This can be changed in the code itself, or in the *.ui files.

Reduced the maximum star label density. If you think it is too small, hit the "b" key and see what the hit ratio is. If it is 50% or less then the label density is not too small. If the hit ratio is significantly less than 50% then the label density may even be too high since most labels are not getting drawn because they would overlap.

Prioritized the star labels based on magnitude. This was not exact.
I bin the labels based on the first two significant digits of the star magnitude to avoid the expense of a sort. I could not see any speed difference using this scheme. The downside is that the labels for stars that have magnitudes within 0.1 will be drawn with the same priority.

This brought the drawing of star labels outside of SkyLabeler::draw() which then enabled me to bring the drawing of the constellation names out of that routine as well which cleaned things up a little. Label priority is now set by the order things are drawn in SkyMapComposite::draw() as well as in SkyLabler::draw().

Cleaned up the starcomponent.h and SkyMapComposite::draw() a little.

lower minimum size so it fits on Classmate PC

Add local copy of libkdcraw library. (Probably temporary solution)

Remove local copy of libkdcraw.

The local copy thing didn't even work and it's a mess to setup.

Begin commits to meet the string freeze deadline today (I hope I haven't missed it due to timezone issues):

* Warn on KolourPaint startup if the screen is paletted (e.g. 8-bit) since it's not really a supported configuration (I tried to support it in KDE 3 but bugs kept appearing; in KDE 4, I haven't had time to check all the code yet e.g. KDE 4 porting regressions + Tone Enhance & HSV effects)

Usability work on the toolbar:
- shorten the toolbar labels of the Previous/Next Page actions, and put them into the toolbar by default instead of the Back/Forward [history control]
- shorten the toolbar labels of the tool buttons
- remove the Open/Open recent button from the shell toolbar

(finally!) rename the okular configuration to okularpartrc, and make okular copy the old kpdf config if found (to preserve all the user settings)

- minimal gui review (used some icons on wizard)
- added a checkbox to print out with no margins

Moved the add/modify/remove and test buttons to a better location. They now dont clash with the standard KCM buttons.

Adding the Blitz dep, as per:
<a href="http://lists.kde.org/?l=kde-core-devel&m=118735498607098&w=2">http://lists.kde.org/?l=kde-core-devel&m=118735498607098&w=2</a>;

License: LGPL v2 -> LGPL v2+ as per Aaron's comments.

mmap problems should now be fixed so lets enable it by default and see what happens ;-)

There is some interference when running apps that use mmap and those that don't together (they grow and shrink the cache files repeatedly), but that shouldn't actually happen anyway (maybe when upgrading kdelibs...) and it still works.

- Don't try to recreate cache files when the cache is disabled.
- !enabled implies !valid
- Disable the cache instead of recreating it when cache files have newer version.

I'm not sure about it, it should prevent trouble when upgrading kdelibs (old apps won't delete cache created by new apps) but OTOH it could cause problems when downgrading... (is that something we actually need to support?)

remove KMD4, Qt has equivalent replacement

!! Not-Monday BIC !!
Move KStyle to kdeui. Since no styles outside of kdelibs and playground seem to be using KStyle, this probably won't mess up anyone, but be warned :-).

no mixer in Phonon for 4.0

Set fixed string for redo/undo icon-text. This prevents e.g. resizes of undo+redo buttons within toolbars.

Cygwin fixes by Ben van Klinken

For fun I tried compiling strigi with cygwin. I got the core things working (deepfind, etc), here's a small patch of the things i had to change to make it work (against the current svn).

Mostly compile stuff (assumptions are made that if the compiler is not msvc then it is mingw). Also there are some ambiguous casts for addValue. And a small typo in the strigiconfig.h.cmake file (which wasn't actually necessary, but is probably incorrect).

Basically the last change regarding XDG base dir spec:
install the .desktop files of agents and searchproviders into share/akonadi rather than share/apps/akonadi, i.e. removing the KDE specific /apps part.

Plugin .desktop files stay in share/apps/akonadi/plugins because they are part of the KDE specific implementation

Last but not least, vanRijns beloved memoviewer. All viewers are back in now. One of the things that's still missing are the export functions some of them had.

Remove external sqlite3 dependency, use Qt's drivers for it.

Get rid of pimItemPosition() which is evil, kills performance and is not really needed anyway.

Removed the stylesheet stuff as it takes much too long with me and is buggy.
Removed the layout, too, since it does not behave as expected (a real button), the other code is commented out.

Ported the wmf import filter. I pull in the source file of some shapes they are only a plugin, not a library to link to. And the header files are installed neither.

I am sure we have there is a better way using the shapes in filters, but for now that works as well.

Waiting for a time when I will finally understand what was wrong with the latest implementation of the paint transfer algorithm, I go back for you and present the previous paint transfer algorithm + the shiny new (tm) complex op.

Too bad that the SoC deadline is over... ;-)

Rename this mathematics.cpp.20070810 to mathematics.cpp if you want yellow and blue mixed to green before Saturday ;-)

Krita now uses blitz (or it will, with my next commit)

Use Blitz for scaling: there are still artefacts, but we are rid of our own asm file now. (Phew!) Also, use Qt's non-smooth scaling if zoom level > 1.0,1.0, because you don't want any smoothing when zooming in.

Start KChart 2 development: Remove old KDChart

Some updates, but still no luck in getting green from yellow and blue (I don't know why I cannot replicate the results of some days ago!).

Working on new algorithms to obtain useful reflectance spectra.

Minor tweaks again. Time for a status update:
After all these tests, try after try, there are (as always, but now, a lot more verified), two possibilities:
- We *can't* get realistic reflectance spectra from RGB Triplets without some other information.
- We can get realistic reflectance spectra.

The latter is what I hope :-) and I discovered I'm not the only one that worked on this: there are at least other two (old) papers about calculating a Reflectance Spectrum from RGB values. This is conforting, for two reasons: I'm not the only one that think it's possible; the guys of the papers got the same problems I'm facing, so I'm on the right path :-) !

First draft of design document for KChart 2

Start the restructuring of the source code for KChart 2.0.

This commit removes the data editor and saves it in the new directory attic/. We also copy all the files that were affected by this commit to the same directory for future reference.

The idea is that once we have sucked all useful code out of the classes in attic/ we will make a garage sale, er I mean we will remove it and all its contents and be left with a new pristine KChart baby that will lead to 2.0.

remove all the backgrounds from the applet SVGs, and enable setDrawStandarnBackground. more consistent now. the default background is really dark though, so we might want to change that. also, fix padding so the whole border is visible.

The start of the massive port of all popup menus to use QActions. Start with cover manager, visualizations and collection rescanning

Update the project structure and KTabEdit is not anymore a kpart compoment

~yay illissius fixed the sidebar resizing. Its slow because QGV sucks at anything fast but it works. It turns out QSplitter will not adjust sizes if one of its children is resized, only if it is hidden.. so we do it ourselves.

Usability Point, text along the icons by default in the toolbar : <a href="http://reviews.holehan.org/KGet/MainWindow#toc3">http://reviews.holehan.org/KGet/MainWindow#toc3</a>;

Added a dbus method to export the transfers information (percents and sizes of the downloads)

move the javascript plugin to the playground.

See <a href="http://lists.kde.org/?l=kopete-devel&m=118731765911962&w=2">http://lists.kde.org/?l=kopete-devel&m=118731765911962&w=2</a>; for more info

move the motionautoaway plugin to the playground.

See <a href="http://lists.kde.org/?l=kopete-devel&m=118731765911962&w=2">http://lists.kde.org/?l=kopete-devel&m=118731765911962&w=2</a>; for more info

Move the statistics plugin to the playground.

See <a href="http://lists.kde.org/?l=kopete-devel&m=118731765911962&w=2">http://lists.kde.org/?l=kopete-devel&m=118731765911962&w=2</a>; for more info

move the smpppdcs plugin to the playground.

See <a href="http://lists.kde.org/?l=kopete-devel&m=118731765911962&w=2">http://lists.kde.org/?l=kopete-devel&m=118731765911962&w=2</a>; for more info.

USability issues fixed:
More ways to activate drop target: <a href="http://reviews.holehan.org/KGet/DropTarget#toc4">http://reviews.holehan.org/KGet/DropTarget#toc4</a>;
Popup for duplicate downloads: <a href="http://reviews.holehan.org/KGet/DropTarget#toc8">http://reviews.holehan.org/KGet/DropTarget#toc8</a>;
Error message when delete the main group: <a href="http://reviews.holehan.org/KGet/Settings#toc2">http://reviews.holehan.org/KGet/Settings#toc2</a>;

Remove old version of oxygen style in preparation of an entirely new implementation

copy oxygen style to kdelibs

New, nuno-approved code for radio buttons (which are NOT windeco buttons!). I have this merged into the style already as well, but need to do some cleanup before I can commit the style.

Added a new area in the statusbar to display the name of the current molecule. (This nice information was not displayed anymore in KDE 4 so far.)

Revised the Nostalgia and Nostalgia Blues themes.
- Shaded the bricks in the Nostalgia as in the Blues theme.
- The enemy in Nostalgia brings the gold on his back instead of in his hand.
- Nostalgia Blues now has its own actors file; the enemy graphic is identical for the enemy and the enemy-with-gold sprites. The reference game did not show which enemies had the gold, so this is for truely nostalgic players! :)

Disable the old KDE 3 graphics themes.

My try on the default Kolf theme. I did all I could... unfortunatly this time 'all I could' wasn't enough. This game either has to be thaken care of, or moved to the playground.

Remove code of KDE 3 XPM-based graphics, now obsolete.

Starting developement on kdiplomacy/kompromise