Hello,

a few days ago I submitted a sample (http://www.princexml.com/bb/viewtopic.php?f=2&t=51351) that is rendered quite differently in Prince than in current browsers (5 different browsers totaling 4 layout engines).

Up to now the post didn't elicit reactions

I would find it helpful if you posted examples of what you're seeing, and what the expected behavior is. I myself was a bit puzzled by column widths that add up to 100% plus 50mm

Dave

Sorry for the delay, table issues tend to be complicated and take time to investigate. We will take a look at this issue.

To dauwhe.

That's why I attached the code, so that everybody gets a chance to try it with the preferred (recent) browser and play by resizing the window's width.

The fact that "column widths [...] add up to 100% plus 50mm" is exactly the point. The behavior of the tested browsers is to subtract the fixed columns widths (50mm) from the available width and distribute the remaining space proportionally among the "percentage" columns, in this case in ratios 2:1:1 corresponding to 50%:25%:25%.

Gecko, Presto, Trident and WebKit do that, Prince should do that.

1. flexible-row.png‎ 12.5 kB

Just for information, today I tried the sample with the new Prince 8.1r4 release for Windows. It still renders differently when compared to the 4 major engines (5 tested browsers) mentioned in previous posts.

Hello,

I retried with Prince 9, the results are similar to version 8 and quite different from all the other layout engines around which all render in the same way and differently from Prince.

Probably part of the problem derives from the following (http://www.princexml.com/doc/8.1/tables/)
"For columns that do not have specified widths, the remaining space (specified table width - sum of all the specified column widths) is evenly divided among them as their minimum widths."

The other layout engines divide the remaining space proportionally to the 'percentage' columns, not evenly.

Right, we did make some improvements to table cell height calculation in Prince 9, but the width calculation has to wait a little longer, until we have time.