If you have a cell that spans to multiple rows, and try to declare row heights for different rows, Mozilla brakes the layout in most of the situations. I've created a few pages to illustrate that. (i) Transitional rendering in "quirks mode" http://www.geocities.com/mvanzin/mozilla/quirks_1.html This should be the correct behaviour for all cases. The spanned row has height of 68, and the first row has a 28 height. In all cases, the height for the second row should be 40, even when not declared (this row has only empty cells). (ii) Transitional rendering in "quirks mode" - part 2 http://www.geocities.com/mvanzin/mozilla/quirks_2.html Here I ommit the row height for the second row, and in this case Mozilla ignores the row height setting for the first row, breaking the layout. (iii) Transitional rendering in "standards mode" http://www.geocities.com/mvanzin/mozilla/stds_1.html Same as case (i) above, but rendering in standards mode (4.01 Transitional). Layout is broken (row height for the first row is wrong). (iv) Transitional rendering in "standards mode" - part 2 http://www.geocities.com/mvanzin/mozilla/stds_2.html Same as case (ii) above. Deleting the spanned row (with the logo image) fixes case (ii) and makes case (iv) looks just like case (iii) (being both standards mode rendering wrong).
Related to bug 58402?
Hmm, yes, it seems. Maybe marking as duplicate? Anyway, the test cases I created show problems in quirks mode also, which are not listed on bug 58402. I'm marking as duplicated and will post my comments there. *** This bug has been marked as a duplicate of 58402 ***
QA contact update