Fixed Issue 83; Transforming graphics context was introducing slight rounding errors in the font metrics measurements. Calculating string widths using compnent.getGraphics() would yield slightly smaller widths than calculating them using the active Graphics object. ...

The fix is to calculate the "minimum acceptable" width that erring components (JButtons and JLabels) can be given the current graphics context and checking to see if it's larger than the minimum width the JLabel is reporting (which is computed at creation time assuming an identity transform). If it's larger (ellipsis would be shown) then it changes the minimum size of the component to be the the nearest ceil(minAcceptableWidth).

The calculation for the minimum size involves measuring text, adding insets, affording for icons and icon gaps.

Also, when calculating the bounds of the component, insets were not being considered but should have been.

Only hackish element of this fix is that its only applied to JButton and JLabels right now since I think they're the only components I can think of that calculate their minimum sizes in this way.

made XYArray implement Cloneable since it had a clone method

Removing unused imports

Made PNotificationCenter matchingKeys use NotificationKey.equals as opposed to reproducing the logic.

Fixing NotificationTarget and NotificationKey equals methods so that they conform to the equals contract.

Fixed Issue 71; made resulting PStyledText take the camera's scale into account. Also, made right click call stopEditing, so that the input field was always the same size as the resulting text would be, was weird otherwise

Cleaned up PNavigationEventhandler. Removed nested ifs, refactored common code out into its own private method.

Unit tests

Reverted override of createInverse in PAffineTransform since the fact that it didn't throw anything was being overlooked in Hudson, but honoured in Eclipse

Removing some calls to printStackTrace(), replacing them with appropriate responses Issue 101

removed printStackTrace call by making addListner return false if it was not able to add the listener

Some refactoring to make PSwingRepaintManager clearer, nothing too fancy

Commenting out assert since it doesn't work on all environments

Issue 74 ; minor style and javadoc fixes

Issue 74 ; minor style and javadoc fixes

Trying to get test to pass. Seems to get GC to run

Fixed Issue 74; removed the need to track PSwingCanvas.SwingWrapper instances in the PSwingRepaintManager, making the memory leak a non-issue

Changed test name since it was incorrectly referring to PCanvas

Added PSwingCanvas Test for Memory Leak

Issue 74 ; some assertions fail when test case is run as part of a test suite

Issue 74 ; adding unit test for PSwingRepaintManager

Issue 69 ; adding render quality parameter

Issue 53. Removing Test Code that can't be run in Headless mode

Fixes Issue 97 and possibly Issue 53 as a result. ...

Moved incorrectly labeled examples TestPSwing and TestPSwingFull to the examples module. Renaming them to PSwingExample1 and PSwingExample2 for lack of a better imagination.

Removing local private members which are no longer needed since Buffering has been removed. ...

Also applied the formatting defined in the eclipse formatting file.

Fixed Issue 83; Guessed that the ellipses were caused my Mac's ability to do subpixel font rendering and MS's inability to do so. MS was probably rounding down to the next pixel and macs were not, allowing the text to be a couple pixels longer. Disabling it seems to do the trick.

Issue 82; Fixed ; Made PSwing register itself as a ComponentListener to the component. When the component's visibility changes, so does the PSwing node. Also, I've overridden setVisible so that it calls setVisible on the underlying component.

Added some tests to PSwing

Added some tests for PComboBox and added getters for canvas and pSwing

Issue 69 ; adding POffscreenCanvas to extras, with unit test and example