# Questions about "Registry of Notable Universe Models"

Posted 5 months ago
690 Views
|
|
0 Total Likes
|
 Firstly, I'm having a ton of fun observing this process and digging into materials so huge thanks to everyone involved.I've been perusing the "Registry of Notable Universes" and have a couple questions: What are the heuristics used to select a universe as "notable"? And who/what operates this filtration process? What process is generating the population of universes that is being filtered? How is the universe / model # derived? ( Order of discovery / generation? Determined by rules or initial conditions? ) How is the generation count determined? ( Most are 198, others are 49, 17.. etc. Is this because of model complexity ? Or some termination condition?) And finally, a humble request: for the "Download notebook" link in the "view universe" page, please generate unique filenames e.g. "wm.db" . I've ended up with a lot of "base(x).nb" files in my download folder with no way of deriving the model number / resource location from the file.
 All of the universes in the registry are currently selected by hand. We basically selected the ones we thought looked interesting during the various searches we have done. We are going to have a more rigorous process in the future, where we will have automatic tests for whether a given universe is consistent with known physics. See (1). See (1). It's a truncated hash of a canonicalized rule with the number of digits chosen based on the complexity of the rule and to avoid duplicates (initial condition will be a separate number if non-trivial). It's an Automatic value for the number of steps argument of WolframModel. You can check the source code here. Good point about the notebook naming. I have forwarded your suggestion to the registry developers.