In devprev, running a version of Mutate No Wiggle, the shadow copy of the monomer becomes distorted on puzzle 1524. The distortion changes as the recipes progresses.
I'm guessing the shadow will resume its normal shape in the end, but it seems like something unexpected is happening here.
| [img_assist | nid=2005322 | title=Distorted shadow copy of monomer during mutate. | desc= | link=node | align=left | width=640 | height=232] |
The shadow was still messed up when the recipe finished, but unfortunately I didn't get a picture. The client crashed with "asymmetry introduced by ActionGlobalMinimize" when I wiggled.
I had two open cutpoints. The same thing happened on the previous h-bond puzzle, and I shared with scientists, but forgot to open a feedback.
For this puzzle, I again shared with scientists. To reproduce, wiggle the shared solution with low wiggle power. The client crashes when you stop the wiggle. Log file attached.
The shadow was back to normal when I restarted. So it's not really clear if the distorted shadow and the crash are related.
For what it's worth, I was able to close the cutpoints, which allows the protein to be wiggled without a crash.
I first selected the loop segments on either side of a cutpoint, did wiggle-shake-wiggle, closed the cutpoint, and repeated wiggle-shake-wiggle. No crash. After closing the second cutpoint this way, I could wiggle all with no crash.
Tested again using puzzle 1524 on 20180725-d14dae34c3-win_x86-devprev, and I'm not seeing the same distortion with MNW, and no crashes when wiggling with cutpoints. Closing.