The colored background of the left RowSpan cell appears, and then as the page finishes loading, it disappears. Interestingly, if you remove the "colspan=2" from the final row (or remove the row), only a 1x1 pixel area switches from colored background to default background (white). The example is cut down from infospace.com
this worksforme on win32 M18 073108 build. are you still seeing this with current builds?
Still happens for me with fresh pull 8/1/00. I just noted something: when it happens, at the bottom of the screen it says "Sending request to a1040.g.akamai.net". When it works (with the minimal example), it says "Document: Done (1.776 secs)". In both cases, the stdout says that the page loaded successfully, and the image in the corner is stable. It happens more frequently with the original URL (http://inspcl.infospace.com/info/classads/iclass?cat=gene&op=search&PDBN=INSPGPRICE&NEXTF=switchv&QFDMA=0&QDMA=&QCQS=&QC=&QS=&QZ=&QO=&QREG=0&QVCAT=1&_debug=&_DEFDMA=&QVSCAT=1&QVCAT=1&QVSSCAT=1&firstsearch=1&CATFILE=/classads/gene_mstr.html&VTITLE=General+Ads&TITLE=&_ADV_ST=2&_ADV_SRC=&_ADV_USR=); with my little test case, it appears to happen only the first time after I start up - reload doesn't cause it (or doesn't cause it anymore).
over to HTML Tables
I can reproduce this every time with a debug build on WinNT with code from 8/19. Load the attached test case (7/31/00 16:21). It displays correctly. Then hit reload. The bottom half of the first cell (the rowspan'd cell) doesn't paint it's background. The top half does paint. If you replace the image URL with a non-existent image reference, it paints fine. Makes me think it has something to do with the incremental reflow generated when the image loads, maybe. Marking Future, we have bigger fish to fry.
WORKSFORME Platform: PC OS: Windows 98 Mozilla Build: 2001012205 marking as such.
Problem indeed seems to be gone