Ljkiernan1 found a glider reflector partial
There's also the possibility of using the glider to eliminate the beehive, and getting output from the upper left area instead. Here's a near miss on a G-to-H, using a speedtunnel derived solution for that beehive:
Code: Select all
x = 50, y = 48, rule = LifeHistory
14.2A$14.2A3$15.B13.2A$14.3D13.A$14.BDB12.A$13.3D2B11.2A11.A$13.5B9.
2A2.B8.3AB$12.6B8.A2.2AB7.A.4B$12.6B9.A.A.2B6.2A.4B$13.5B8.2A.A.AB.2B
E4B2.B2A$13.6B11.2A3BEBEB5.AB$12.6B13.3BE2BE2B.BA.AB$7.2A3.7B14.2B2E
3B.B2A$.B5.A.A3.6B14.9B$.2B6.A.8B8.A5.9B$.3B5.2A8B6.3A5.9B$.4B5.10B4.
A7.10B$2.4B4.12B2.2A5.12B$3.4B4.15B4.13B$4.4B3.13B5.14B$5.4B2.15B2.4B
.10B.B$6.4B2.19B.12B2A$7.4B.3BD28B2A$8.7BD29B$9.6BD29B$10.30B2.B$11.
28B$11.4BC23B$11.5BC22B2A$8.6B3C10B5.B.2B2.A2BA2.2A$5.2E20B4.3B4.ABAB
3.A2.A$4.B2E14B2E2B6.B2AB4.A4B.A.A.A$4.17B2E2B7.2A7.B2A.A2.A$3.21B17.
BAB.A$4.5B.14B15.A4.A$5.2B6.10B16.5A$3.A.A.A5.10B$.3A.2A.A4.6B2.B19.A
$A7.A4.6B21.A.A$2A6.A.2A.BA3B23.A$7.2A.A2.ABA.2A$11.3A.A.A.A2.A$14.A
2.A.4A$11.3A3.A$10.A7.3A$10.2A8.A!
The tricky thing is whatever you do needs to produce some debris close to the head of that glider-releasing fishhook, which then must be cleanly consumed. Here's another possibility, connecting the B to a conduit 1, but the leftover beehive isn't quite close enough to the fishhook. Maybe there's another B-to-X that works, though.
Code: Select all
x = 50, y = 61, rule = LifeHistory
15.2A$15.2A4$14.3D$15.D$15.3D10$29.2A$30.A$29.A$29.2A11.A$27.2A11.3A$
26.A2.2A8.A$16.2E8.2A.A9.2A$16.2E9.A.A.A4.E7.2A$26.A3.2A3.E.E6.A$15.D
7.A.A8.E2.E4.A.A$14.3D6.2A10.2E5.2A$14.D.2D$27.A$12.D12.3A$11.D.D10.A
$11.D.D10.2A$12.D3$44.2A$44.2A5$15.C$16.C22.2A$14.3C21.A2.A2.2A$5.2E
31.A.A4.A2.A$5.2E14.2E9.2A5.A5.A.A.A$21.2E9.2A8.2A.A2.A$42.A2.A$39.A
4.A$39.5A$3.A.A.A$.3A.2A.A32.A$A7.A31.A.A$2A6.A.2A2.A26.A$7.2A.A2.A.A
.2A$11.3A.A.A.A2.A$14.A2.A.4A$11.3A3.A$10.A7.3A$10.2A8.A!
Since this is the second near miss at a G-to-X I've produced this way, here's my process:
(1) Find a pi catalyst that has decent clearance, ie the result is a couple clumps of cells that are a decent distance from the catalyst. Ideally you want something compact (so there's more room for further catalysts and output) that sends some stuff behind the pi (to be used to restore the block). Use CatForce, take a conduit start from the ECC, or use one of the couple pi-targeted catalysts in the catalyst lists in mvr's CatForce repo (light speed branch).
(2) Use barrister to find welds or small variations on that catalyst. Find a few that are as good or better than the original catalyst.
(3) Run CatForce on those and see if the block (from G + block -> pi) can be restored. Usually you'll get 0 results, or a couple results where the block is restored super briefly, but sometimes you get lucky.
(4) Use CatForce or barrister for cleanup.
It's nothing scientific. It's rather tedious and tends to produce large things with not-great repeat times. But hey, it's worked for me. While on paper a purely bellman/barrister approach (eg how the snark was found) could produce smaller solutions, I find that in practice, it's very hard to both produces output and restore the bait, unless you have a partial where one of the two is already achieved.
EDIT: here's a Pi-to-B that's just barely not connectable.
Code: Select all
x = 23, y = 32, rule = LifeHistory
15.A$5.D7.3A$4.3D5.A$3.2DBDB4.2A$3.5B.5B$3.9B$2.11B$2.11B$3.11B$3.12B
$3.11B$3.9B$2.11B7.2A$2.12B6.A$2.12B3.BA.A$3.13B.B2A$5.B3C9B$7.BC9B$
6.3C9B$4.13B$3.11B$3.11B$4.9B$5.7B$6.5B$3.2A5B$3.A2.2B.2B2.2A$2A.A2.A
.A2.A2.A$.A2.3A.6A$.A$2.3A5.2A$4.A5.2A!