you get to this page by going to http://www.netscape.com and clicking "travel" code: NS_ASSERTION(NS_FRAME_IS_NOT_COMPLETE(aStatus), "unexpected status"); stack: nsTableRowGroupFrame::SplitRowGroup() line 875 + 38 bytes nsTableRowGroupFrame::Reflow() line 996 nsContainerFrame::ReflowChild() line 392 + 28 bytes nsTableFrame::ReflowMappedChildren() line 2496 + 31 bytes nsTableFrame::ResizeReflowPass2() line 1801 + 31 bytes nsTableFrame::Reflow() line 1610 + 35 bytes nsContainerFrame::ReflowChild() line 392 + 28 bytes nsTableOuterFrame::Reflow() line 990 + 37 bytes nsBlockReflowContext::ReflowBlock() line 153 + 33 bytes nsBaseIBFrame::ReflowBlockFrame() line 2190 + 41 bytes nsBaseIBFrame::ReflowLine() line 1651 + 20 bytes nsBaseIBFrame::ReflowDirtyLines() line 1380 + 26 bytes nsBlockFrame::ReflowDirtyLines() line 4626 + 12 bytes nsBaseIBFrame::Reflow() line 676 + 25 bytes nsBlockFrame::Reflow() line 4253 + 25 bytes nsBodyFrame::Reflow() line 259 + 25 bytes nsContainerFrame::ReflowChild() line 392 + 28 bytes RootFrame::Reflow() line 209 nsHTMLReflowCommand::Dispatch() line 163 PresShell::ProcessReflowCommands() line 861 PresShell::ExitReflowLock() line 522 PresShell::ContentAppended() line 1012 nsDocument::ContentAppended() line 890 nsHTMLDocument::ContentAppended() line 426 + 17 bytes HTMLContentSink::WillInterrupt() line 1445 CNavDTD::WillInterruptParse() line 3595 + 18 bytes nsParser::ResumeParse() line 726 nsParser::OnDataAvailable() line 968 + 15 bytes nsDocumentBindInfo::OnDataAvailable() line 1553 + 24 bytes OnDataAvailableProxyEvent::HandleEvent() line 606 + 45 bytes StreamListenerProxyEvent::HandlePLEvent() line 452 + 12 bytes

Removed the assert and changed the code to deal with that case. The problem was that we currently can end up with a constrained height even when we're not paginated. This case was because there was an IFRAME that wasn't scrollable and so there was no scroll frame and the table was getting reflowed with a constrained height. The fact that the height is constrained is bad, but fixing that is a much bigger problem...

qa contact set to eli@netscape.com

rubber-stamping as verified.