A for awesome wrote: ↑December 9th, 2020, 5:31 pm
Okay, that makes a lot more sense. I've seen a lot of "can this eater be restored" types of posts recently so I guess I just assumed. Sorry about that!
Replacing a fishhook isn't completely unfeasible, but chances are that it is not worth the large size and long repeat time necessary to either synchronize two gliders for a synthesis, search for a transparent reaction, or find a reaction that creates a fishhook, clean it up, and attach the input. There are some ways to make or restore a fishhook, but their scarcity proves the general rule.
Here is a transparent fishhook reaction that I'm sure was found by a Bellman search for transparent catalysts in general, and
here is a slight modification of a reaction that I found when I was trying to do something else with a century (My best guess is that I was trying to get an I-sequence with some junk behind it because I had found an I→C that left a beehive behind (I think that it could have theoretically been removed with a stable catalyst, but that would have made inserting the I-sequence very awkward.), and wanted a way to make an I-sequence with some trailing junk that would annihilate the beehive.), wanted to remove one of the blocks, and tried a fishhook.
Code: Select all
x = 19, y = 16, rule = Conway's Life
8b2o$8b2o$o16b2o$3o14bo$3bo11bobo$2b2o11b2o6$5bo$6bo$7bo$6bo$4b2o!
wwei23 realized that that block can be removed with a dot sparker in a way that works with any relative timing and therefore can be used in stable circuitry. I never found a way to make an I-sequence with appropriate junk behind it, though.
By the way, here are some more examples of partial conduits that can use a fishhook or something else as a sacrificial object but that I'd like to see using a permanent catalyst instead.
Code: Select all
x = 128, y = 47, rule = LifeHistory
7.D31.D31.D31.D$6.3D29.3D29.3D29.3D$8.2D30.2D30.2D30.2D17$2E$.E64.E30.
2E$.E.E30.2E29.E.E29.E.E$2.2E30.2E30.2E30.2E2$20.2A30.2A30.2A30.2A$19.
A.A29.A.A29.A.A29.A.A$20.A31.A31.A31.A9$14.C31.C31.C31.C$13.3C29.3C29.
3C29.3C$12.2C30.2C30.2C30.2C2$25.2A30.2A30.2A30.2A$25.2A30.2A30.2A30.
2A$25.2A30.2A30.2A30.2A$25.A31.A31.A31.A$12.2A10.A.A17.2A10.A.A17.2A10.
A.A17.2A10.A.A$12.2A11.A.A.3A12.2A11.A.A.3A12.2A11.A.A.3A12.2A11.A.A.
3A$26.A.4A26.A.4A26.A.4A26.A.4A$27.A31.A31.A31.A!
Code: Select all
x = 37, y = 37, rule = LifeHistory
6.2E$6.E28.E$4.E.E27.E.E$4.2E28.2E2$.2C28.2C$2C28.2C$.C29.C2$5.D29.D$
5.2D28.2D$4.2D28.2D14$7.E$5.3E27.2E$4.E29.E.E$4.2E28.2E2$.2C28.2C$2C28.
2C$.C29.C2$5.D29.D$5.2D28.2D$4.2D28.2D!
Code: Select all
x = 88, y = 29, rule = LifeHistory
8.2E13.2E41.E$9.E13.E21.2E18.E.E18.2E$6.3E15.3E18.E.E17.E.E17.E.E$6.E
19.E19.E19.E19.E3$2.C19.C19.C19.C19.C$2.3C17.3C17.3C17.3C17.3C$3.C19.
C19.C19.C19.C18$.D19.D19.D19.D19.D$3D17.3D17.3D17.3D17.3D$2.D19.D19.D
19.D19.D!
The third example has two periodic stabilizations (
here and
here (although the glider-based stabilizaton is probably dependent)) and
a StateInvestigator stabilization. On the other hand, the first example
was searched with Bellman by Kazyan without any solutions, although Kazyan did mention that he could have done a more thorough search (which I didn't request at the time because I was not yet aware of our lack of ways of making a century with good output clearance). (For a while, I was mistaken and thought that the partial conduit of mine that was searched with Bellman was from my work investigating conduits accepting R−B or R−H, and I didn't check the post and realize that it was my partial C→C/I→R until recently, so if you've seen any posts where I display seemingly unrealistic confidence that the century is good for making conduits with or that it is possible to get a century into positions that seem difficult, that is probably the reason.)
Edit: Stable catalysts are available for two out of three in Symbiosis.
Code: Select all
x = 24, y = 25, rule = Symbiosis
.B$B.A$.2A$.A$.B17.B$18.A.A$19.A9$13.A$12.3A$11.2A3$22.A$23.B$22.BA$11.
2A$11.2A!
Code: Select all
x = 17, y = 12, rule = Symbiosis
10.AB$11.A3$7.A$7.3A$8.A$BA3$16.A$15.B!
For the second, while trying to find a Symbiosis catalyst, I realized that the R-sequence goes back to the same place again, so the catalyst can't recover right away. It is okay if the catalyst interacts with that region again if it waits long enough before doing so, but it definitely makes finding a permanent catalyst (in Symbosis or in ConwayLife) much less plausible.
I am tentatively considering myself back.