Thanks Steve for your analysis, this is exactly what the problem is.
I can not see a hole in this argument and attach a patch that implements this. I will also add a regression
test later for this. I will also think some more about it to see if I can find a flaw in the new approach.
Thanks Steve for your analysis, this is exactly what the problem is.
I can not see a hole in this argument and attach a patch that implements this. I will also add a regression
test later for this. I will also think some more about it to see if I can find a flaw in the new approach.