wwei47 wrote: ↑
April 26th, 2021, 9:09 am
It might be worth not separating LWSSes on LWSSes. It's hard to find soup-based syntheses for those when they just get separated!
There's a practical problem with "not separating LWSSes on LWSSes". They're separated because the current census code separates them.
I'm not really an expert here, but from what I understand, changing the census code to not separate LWSSes on LWSSes, but continue to separate everything else that currently gets separated
, is either hard to do... or it's somewhat slower. Like, you could add some new code that runs whenever a soup census includes an LWSS count greater than one, and just look to see whether two of them happen to be in the LWSS-on-LWSS position.
Changing methodology midstream would kind of invalidate the relevant census statistics, though. It might be something to consider if a whole new census run is started at some point -- say starting from zero counts with 32x32 soups, or something like that.
Another example that comes to mind is there are a lot of cases where it would be useful to collect soups for various pseudo-still lifes -- block on eater, beehive on loaf, or whatever. Currently it's hard to dig up soups that might help synthesize these pseudo combinations more efficiently. But adjusting the census code to track these separately would probably be a highly non-trivial design effort.