…nd BackgroundImageSrcRect don't do anything if set to the current value. BUGFIX: Widget setDirty method shouldn't have done anything if set to existing dirty region, but probably was.
…not marked dirty.
…s that have been hit with XHR requests. This mitigates logic problems caused by clock skew.
Optimized another setDirty case.
…en if the parent didn't acknowledge the child. This caused performance problems for Modals.
…ty, even if they were disowned. This caused big performance issues.
…y rooted. Minor beautification change.
Added Widget.hasChild, which says if another Widget is a child of this one. XhrManager now fires requests all serially.
…own and mouseup. Due to a bug in the scrolling code, NaNs were leaking into the dirty rectangle property, causing crashes elsewhere.
…parent doesn't link back to the child.
…e right location.
…tion that made them invisible.
…e classes that inherit from it. Layout now has a SizeToParent property. Don't try to use it with SizeToChildren. Layout now has a ConstrainChildren property, to keep children in bounds. Factored Widget's event handling code just a bit.
…dow, so I implemented it myself.
… different HoverBackgroundStyle. Removed vestigal image background code from Button. Widget does that now. Buttons now have an "Enabled" property. Disabled buttons do nothing when clicked.
BUGFIX: Hovertip fade-in/outs stacked, causing epilepsy.
…gic is split between Widget and CanvasWidget.
Added an Alert class of Dialog, to show alert messages.
… Label's drawing.
… sizes to the graphic context, which is apparently not allowed. Caused graphics to crash.
…tically. Hack; this was done by allowing negative width and height in the new BackgroundImageDestRect property, but I'm not handling the non-unit scalincg properly.