Page 1 of 1

Super splitting bug

Posted: Fri Jan 19, 2018 1:53 am
by wapcaplet
I have seen a strange splitting bug a couple times in the beta (v91). It happens with phagocytes and devorocytes evolving in an experimental plate. Some cells evolve the ability to do a "super split"--where children fly apart energetically and separate by a great distance, as if viscosity were low (even when it's normal).

Here is video capture of what it looked like: https://www.dropbox.com/s/fpwulbfzva98p ... s.mp4?dl=0

Here is the saved substrate, though when I load it now, the split behavior looks normal: https://www.dropbox.com/s/9z8awin5c3x6h ... trate?dl=0

The genome editor also shows the super splits, at least at first. After loading one of the devorocyte cells from the microscope, I had to zoom way out to see its children split. But after editing something, it returned to a normal (reasonable-looking) split simulation.

Re: Super splitting bug

Posted: Fri Jan 19, 2018 2:06 am
by Megathosto14
I see that the higher cytoskeleton strength, the higher split range between two cells. But that in the video seems too much.

Re: Super splitting bug

Posted: Fri Jan 19, 2018 2:16 am
by wapcaplet
You may be able to reproduce the bug by evolving the substrate I already posted. I did this and got a modified genome that did the same huge split (and recorded more video too, just because it looks cool).

Here is the snapshot substrate when it evolved again:
https://www.dropbox.com/s/yjurh8zhntadc ... trate?dl=0

And video (44 MB): https://www.dropbox.com/s/yhi8rlo5p1c9n ... 2.mp4?dl=0

As before, if I load this substrate now, everything looks normal. The buggy splitters are just normal splitters. But evolve it awhile and I bet you'll see it appear anew.

Re: Super splitting bug

Posted: Fri Jan 19, 2018 8:27 am
by Petter
This is likely the bug i mentioned. It is fixed but not uploaded yet.