dvgrn wrote:It's definitely still a big open question where exactly all these Sir Robins with Minstrels To Be Removed are coming from, though. We have to copy/paste them in to make them appear in the first place, so there's no actual practical use yet for any kind of Minstrel Detector or Minstrel Remover.
Usefulness disregarded, here is Goldtiger997's Amazing 2 in 1 Reusable Multipurpose Minstrel Remover and Detector™
Beautiful! I think we can reduce the height by a few thousand rows by inserting the doubler-turner in the 2c/3 wire, so that the 2c/3 signal turns northeast instead of becoming a slow NE-directed glider. Then we should be able to fit the entire RLE into the 8192-row limit of LifeViewer.
It works for Sir Robin with any one of his three minstrels, or even with no minstrels at all. Minstrel 1 produces a SW glider, Minstrel 2 produces a NW glider, Minstrel 3 produces a SE glider, and no minstrels produces a NE glider. It is very large, mostly consisting of a very long 2c/3 signal. When brave Sir Robin turns about, and gallantly chickens out, his very brave retreat is very fast so it takes a long time for the 2c/3 signal to catch up with him.
The geometries of this pattern were very confusing to figure out! Every time the 2c/3 signal was lengthened by one unit, Sir Robin would effectively be 7 ticks later than he was before.
Yes, adding 6fd to the 2c/3 wire, 2fd to the NE glider path, and 8 rows to Sir Robin's path results in a total delay of:
(8 / (2/6)) - (6 / (2/3)) - (2 / (1/4)) = 24 - 9 - 8 = 7 ticks.
If you follow my suggestion of replacing the NE glider path with a 2c/3 double-signal wire and using the elementary turn, then it improves to:
(8 / (2/6)) - (8 / (2/3)) = 24 - 12 = 12 ticks.
So we expect the height of the pattern to be reduced by roughly five-twelfths, or 42%.
Once I added another level of adjustability I was eventually able to make the synchronization work correctly. I also had to modify dvgrn's Amazing Reusable Minstrel Remover
to make it release a glider earlier on.
calcyman wrote:Minstrel 2 can support one end of PHPBB12345's wave:...
Nice! Good luck with your search for the other end.
Thanks! It's currently reached logical width 31, and my current longest partial is this:
Code: Select all
x = 66, y = 158, rule = B3/S23
Even though I'm using a new head search, I've resumed the tail search from last time so it has a greater surface for meet-in-the-middle. (It's at 45 000 heads and 1 400 000 tails as a result of this).