From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.1+) Gecko/20010620 BuildID: 2001062004 If table cells have borders, and the widest cell in a column contains an image's alt text because the image can't be loaded, the width of the cell with the alt text is wider than corresponding cells in other rows by the width of any borders. Reproducible: Always Steps to Reproduce: 1. Load the attached test case. 2. Look at each of the tables. Actual Results: Cells with borders and broken images are wider than the corresponding cells in other rows. Expected Results: The first cell of each row in each table should have the same width as the first cell of all other rows in the table. I'm pretty sure this is a regression. I'll try other builds once this bug is filed.
The is a regression. The test case works in 0.8.1, but not in 0.9 or later. It also fails on the Mac with 0.9.
Seeing this on Sun build 2001-06-22-10 as well.
adding depends, regression
Does this truly depend on 41924? It doesn't really matter what content is substituted for the image; the problem is that the width of that substituted content is used to adjust the width of the affected cell, but none of the other cells in the table are adjusted. I'm not an expert on the code, but it looks likely from the outside that this bug could be fixed without 41924 being addressed. (Especially since it was introduced long after that bug was filed.)
wfm, 12/7/1 win2k debug
WFM with 2001120703.
Verified WFM with build ID 20011207