Result: blocked
Files: none
Commit: none
Fixes: none
Pushed: none
Tests: preconditions on origin/develop (k3a1 5bf28f57, #995 0cf0b173, #997 7029fe65, #1005 7bf9dea5); probe of the move’s forms with target/fast/nikaia lower: trait method with a fn parameter, the let-bound known lambda with ?? ahead.get(name)?.clone(), set difference(..).map(clone).collect(), an enum match with a guard, map.keys().all(..), a generic recursive call all lower; only the set’s iter() is refused
Gate: none
Deviation: none
Found: BTreeSet has no iter() (NK1171), and the draft needs it in a chain: computed :376 named.iter().any(fn(name) { left_out.contains(name) || (absent.contains(name) && known(name) == null) }), the names the value function reads (names_in_block fills a set). Probe: use std::collections / fn main() { let mut s: collections::BTreeSet[String] = collections::BTreeSet(); s.insert("a"); println(s.iter().any(fn(n) { n == "a" })) } gives NK1171 collections::BTreeSet has no method called iter. std.contracts gives a set only union/intersection/difference as sequences; no ADR refuses iter (ADR-collections-1 covers extend, ADR-collections-2 covers arrays). Ways around it (a for loop, union(empty).any, the host handing back a Vec instead of the set) are not taken, per the brief
Needs: owner decision (does collections::BTreeSet get iter(), the sorted lent sequence, like ADR-collections-2 D1 for arrays); then a fresh k3a2 run. The K3a row of docs/selfhosting-plan.md has another session’s uncommitted edit (the part 1 entry); I did not touch it
Changelog: none
Owner: Does collections::BTreeSet get iter() (its elements lent, in order, as Vec::iter() gives them), so that set.iter().any(..) / .filter(..) work?
Disk freed: 0 GB; cache 3.8G, disk 4.1G free; no new cache entry over 100 MB from this run (run-bundle/b2db8b4e959f1cfc 286 MB is older, from b1005)
Rs2nika: none
