Post
by SimonOrde » 07 Jan 2008 13:00
I think I can see what the problem is. Looking at Richard's settings in the 'More Dimensions' dialog, he has changed some values to make rows thinner, and bring boxes closer together. The combined total of the (d), (b) and (e) values give the total gap between any given childbar, and the bottom of its parents' boxes above it. In your case, Richard, this comes to 0.45 cms. However, the 'Min Proximity gap below boxes' value is 0.5 cms. FH has to ensure not only that boxes don't overlap; but also that they don't get too close to each other. The min proximity gap specifies how close one box can get to another, from below. As the min proximity gap is greater than the gap down to a childbar, the childbar is effectively invading the space of spouse1. Consequently spouse1 is moved out of range of the childbar. To solve the problem, you only need to reduce the min. proximity gap, or increase other values until the combined to total of (b)+(d)+(e) are greater than it.
It ought to be the case that reducing the min proximity gap to 0.44 cms should fix the problem. However, I found I had to reduce it to 0.43 cms. This is probably a conversion/rounding issue of some kind.
Given that this is what caused the problem, you well might ask - why did the gap only appear on the left. Surely it should have appeared on the right as well. That is a valid question. Without looking in more detail, I can't say for sure why it didn't, but I would guess that it was probably a consequence of an optimising assumption (wrong in this case), designed to make the whole thing run faster. I will make a note to look into this.
Another moral of this case is perhaps that FH should warn you when your min proximity gap is too big, as in this kind of case.
