Discord drafts (BB(6) channel)
STALE (overnight 2026-10-10): #3, #11 and #12 were KILLED (check_kill PASS). Message 1 still asks about #12 and Message 2 about #3/#11. Redraft before posting: announce those kills from each machine's
submission/discord.md, and only ask about the machines still open.
2026-10-10. Not posted. Paste each block as one message. Both are under Discord's 2,000-character limit. The messages say "we"; add the disclosure line (targeting.md §5 Q8) if you want it.
Built from the 11 discord_check drafts in machines/*/observe.json. Six were dropped (bottom of this file). Five were merged into the two messages below.
Facts re-checked on 2026-10-10:
- all 21 machine strings in this file are on data/holdouts/2026-10-07_778.txt;
- none of the 12 in the two messages is on the UBRRBA (mxdys) or longitudinal (dyuan01) lists in data/discord/;
- hipparcos's renaming and the mirror twin hold (checked by brute-force state relabeling).
Message 1: Structural counter + counter, Counter Class 4
Why we ask: these are pilot leads #1 (+ #21, #22), #14 and #12. They have no sheet notes, threads or claims. Before spending proof time, we want to know whether anyone has them, or has tried them and hit a known obstacle (§5 Q4).
Is anyone working on any of these Structural counter + counter / Counter Class 4 holdouts, or has anyone tried them and got stuck? We found no sheet notes, threads or claims, and none are on mxdys's UBRRBA list or dyuan01's longitudinal list.
1. `1RB0LC_1LA0RE_0RD1LF_1LC1LB_0LD0RB_0LA---`, `1RB0LC_1LA0RE_0RD0LA_1RE1RA_0LD1RF_0RB---`, `1RB1RC_0LA1RF_1RD0LE_1LC0RB_0RA0LC_0RD---` (identical trajectories to 1e9). Period-5 background `10110` carrying a binary counter whose digits are 3-cell defects (`110` = 1, `100` = 0). Digits drift left on carries, and a bit counter at the left edge absorbs them.
2. `1RB---_1LC1RE_1LD0LC_1LE1LC_1RF0RE_1RA0RB`. Period-5 tape cut into segments by phase-shift walls. Each wall moves 3/2 times as often as the one to its left, so segment lengths run 12, 18, 27, 41, 61, 92, 138, 206 blocks.
3. `1RB1LC_1RC1LB_1LD0RB_0LE1RD_---0LF_0LA1LF` and its twin `1RB1LA_1LC0RA_0LD1RC_---0LE_0LF1LE_1RF1LB`. Run-length counter whose epochs double in length. Each epoch starts at `0^inf <A 0 1^3 0^2 1^2 0^2 1^(2k) 0^2 1^(2k+1) 0^P 1 0^Q 1 0^inf` with P+Q = 2k and P irregular. No halt in 1e11 steps; in a macro simulation every even P closes the epoch for k <= 13.
What we checked: direct simulation (1e9 steps for groups 1 and 2, 1e11 for group 3), plus the local rules above on random instances. No proofs yet. If nobody has them, we'll write Rocq proofs, starting with group 1.
Message 2: Skelet 1-like Class 2
Why we ask: these are pilot leads #3 and #11, and the same mechanism shows up in both. #11 is hipparcos's Level 1 machine started in a different state. #3 may sit on an unpublished Sk1 Class 2 list (§5 Q3). Sk1Machines.v was committed on 2026-09-30 by ccz181078, which is mxdys's GitHub handle (PLAN §2).
Is anyone working on any of these Skelet 1-like Class 2 holdouts? @hipparcos, group 2 is the table of your `1RB0RB_1LC0RA_1LE0LD_1LC1LF_1RA0LD_1LC---` (Oct 2025 thread), so that part is mainly for you.
1. `1RB0LD_1RC0RA_1RD1RF_1LA0LE_1RE1LD_0RB---`, `1RB0RE_1LC0RA_1LD0LB_1LA1LF_1LE1RA_0LC---`, `1RB1RF_1LC0LE_1RD0LB_1RA0RC_1RE1LB_0RD---` (identical trajectories to 1e9). Binary down-counter on a list of `(10)^p (01)^q` pairs; on overflow the list is re-paired.
2. `1RB0LE_1RC0RC_1LD0RB_1LA0LE_1LD1LF_1LD---` is hipparcos's machine with states renamed (A..F -> E,A,B,C,D,F), i.e. the same table started in E instead of A. `1RB0LD_1RC0RE_1LD0RE_1LA0LA_1RB1RF_1RB---` is its mirror. Binary counter over `(10)^u [00] (01)^v` blocks, also with a re-pairing overflow.
Both share SBCv6 TM21's architecture (carries plus a re-pairing overflow), but TM21's configs don't fit either one. What we checked: all are still on the 2026-10-07 holdout list; `Sk1Machines.v` only covers Class 1; the carry and overflow rules hold in simulation to 1e9. No proofs yet. If nobody has them, we'd like to try one Rocq proof per group.
Dropped: Dennis's 2026-10-10 decisions make these moot
| pilot # | machine | why dropped |
|---|---|---|
| 4 | 1RB0LA_0RC1RD_0LD1RE_1LA0RB_0RD1RF_1RC--- |
On UBRRBA_TODO (fair game) and 1 transition from SBCv7 TM4/TM37 (template neighbour: fair game). dyuan01's list is a technique filter, not a claim. Re-read that thread before submitting. |
| 5 | 1RB1RA_0LC0RF_1LD1LC_1RE1LB_---0RA_1RF0RD |
UBRRBA_TODO: treated as tried and failed. |
| 6 | 1RB1LC_1RC---_1LD1LC_1RE0LC_1RF0RE_1RA0RD (+ twin 1RB0RA_1RC0RF_1RD1LE_1RE---_1LF1LE_1RA0LE) |
UBRRBA_TODO (both). |
| 9 | 1RB0LD_0RC1RA_1RD1RF_1LA0LE_0RA0LC_0RA--- |
UBRRBA_TODO. Whether the left-end overflow comes back after 2e10 is for our own accelerated simulation to answer. |
| 13 | 1RB0LE_1LC1RD_1LA0RB_1RC1RB_1LF0LC_---0RC (+ near-twins 1RB0LC_1LC0RE_1RA1LD_1LA1LC_1RF0RA_---0LA, 1RB1LD_1RC0LA_1LA0RE_1LB1LA_1RF0RB_---0LB) |
UBRRBA_TODO (all three). |
| 18 | 1RB1RF_1RC0RA_1LD1RE_1LE0LE_0RA0LD_0RB--- |
BMO7.v neighbour (C1 only): template neighbour, fair game. |