XulRunner display problems

[jarekevi]

Environment
SmartGWT version 3.1p - commercial
Browser: XulRunner 10 with FF10 engine & XulRunner 20 with FF 20 engine
System: Win 7, Win 8

Problem description:
The problems occur on XulRunner browser. FF 20 seams to be working fine.
1) All ButtonItem are cut by a couple of pixels at the bottom.
Images: buttons 1 buttons2 buttons3
2) showPrintPreview() shows small grid with scroll bar on the side despite much larger view - when used version 1
version 2:
print preview has right size
but for both versions if scroll of original grid used shows empty print previw but the scroll bar of the print preview indicates that there is some content
Images: print preview & print preview after effect

 // version 1
				Canvas.showPrintPreview(getView().getUserListGrid());
				getView().getUserListGrid().markForRedraw();	
				
				
			//version 2	
				final List<Canvas> components = new ArrayList<Canvas>();
				Window printWindow = new Window();
				components.add(getView().getUserListGrid());

				printWindow.setWidth100();
				printWindow.setHeight100();
				printWindow.disable();			

				Canvas.showPrintPreview(components.toArray(new Canvas[components.size()]), null, getView().getUserListGrid().getTitle(), new PrintPreviewCallback() {
					
					@Override
					public void execute(PrintCanvas printCanvas, PrintWindow printWindow) {
						getView().getUserListGrid().markForRedraw();
						printCanvas.markForRedraw();
					}
				} , printWindow, "Print");

[Isomorphic]

Please test with a normal browser release - if this is specific to XulRunner and doesn't happen with Firefox, sorry, but it's not something we'd look into.

[jarekevi]

Problem 2 still persists in FF 20 and Chrome 23.0.1271.97

[Isomorphic]

We're not seeing a general problem with PrintPreview and ListGrids.

It looks like you have a getUserListGrid() method returning a ListGrid that is configured in a way we can't see. We would need to see what's special about this ListGrid to help troubleshoot why it isn't behaving normally.

As always, if you think there's a framework bug here, please create a ready-to-run, minimal test case that reproduces the issue, and we can take a look at that.