I'll try to avoid exaggeration here. My commit [1] was about a Shop / utils function, maybe not exactly a bottom layer of yade (at least, not in my opinion).
As illustrated by the present bug and the launchpad question, [1] could be problematic in case of Python code (hardcoded in trunk or in users scripts) involving some other functions exposed through boost python with a specific type of arguments. It is true I did not foresee this problem in [1], in spite of the discussion [2] (see also therein for the motivations to commit [1])
As for #4 and Robert's list #1, I went through it and do not think there is now (was ?) any problem in this list due to commit [1], and in possible connection with the present bug.
I'll try to avoid exaggeration here. My commit [1] was about a Shop / utils function, maybe not exactly a bottom layer of yade (at least, not in my opinion).
As illustrated by the present bug and the launchpad question, [1] could be problematic in case of Python code (hardcoded in trunk or in users scripts) involving some other functions exposed through boost python with a specific type of arguments. It is true I did not foresee this problem in [1], in spite of the discussion [2] (see also therein for the motivations to commit [1])
As for #4 and Robert's list #1, I went through it and do not think there is now (was ?) any problem in this list due to commit [1], and in possible connection with the present bug.
I will look closely into py/pack/ _packPredicates .
[1] https:/ /github. com/yade/ trunk/commit/ 1db13fb1183b9e2 94dc9761da76cfa 4fc2791cc1 /msg13339. html, fourth item in particular...
[2] https://<email address hidden>