g_membed apparent bugs
Two issues with g_membed:
1. -zinit 0.3 -zend 1.0 appears not to scale the z coordinates back up correctly--z coordinates are still squished in the output frame.
2. If the topology and coordinates have molecules in the following order:
And the index group for insertion is INS_A | INS_B, rm_bonded() will not tag INS_B for removal. Apparently the first atom in INS_B fails the bINS check, but I wasn't able to pin down why. I manually checked the index groups against the coordinate indices.
Unfortunately, I don't have a good simple test system prepared for these issues (they occurred with a production system in git-current gromacs).