Elamur Groupantsexecution-ordered-workers
The groupant beside world/src/exec.rs in the Mova repository: what the code computes, what it never does, how to go back from its outputs, its invariants, cost and checks.
Facts
Implementation material. A Mova code model describes this code at one source revision: partial models, finite checks and their stated scope. It is not a reviewed general exposition, and its labels are shown as written, never raised.
Components
| component | label | where |
|---|---|---|
| approximant | partial | Forward |
| ast | not materialized | |
| composition | partial | Composition |
| cost | partial | Cost |
| descriptant | present | model.md |
| invariant | partial | Invariant |
| tests | coverage recorded | tests |
Obernants
The typed relations around the forward model: the behaviour complement (what the code never does), the reverse reconstruction (every input compatible with an output) and specialised views.
| id | type | label | where |
|---|---|---|---|
nonbehavior | behavior_complement | partial source derived model | nonbehavior |
antecedents | reverse_reconstruction | partial source derived model | antecedents |
chunk-envelope | partition_geometry | partial source derived model | chunk-envelope |
panic-frontier | failure_recovery | partial source derived model | panic-frontier |
Model text
descriptant
execution-ordered-workers
Forward
Compute cap-bounded chunk size; spawn slices in order, collect all joins, collect Result, then flatten.
Invariant
Positive cap gives ordered disjoint intervals covering all inputs. No successful partial result follows a joined panic.
Cost
O(n+t) vector aggregation plus job work; storage retains outputs and join results. Allocation, thread creation failure and scheduler fairness are outside this approximation.
Composition
This cross-cutting primitive serves cap 12 CPU parse and cap 16 general/GPU operations. Parent must preserve Result handling; joining does not imply transactional rollback.
Epoch memory
2026-10-29: first source-bound instance. Source SHA-256 180f88005a4cd3852f649059e420951c1fe9ad4a699c9d175358169f25e43b4d; bytes [929,2029). Production source is unchanged. Ranges are snapshot anchors, not proof of coverage. AST gap: no projection is materialized; macro expansion, type inference, semantic dependency analysis and generated proof remain absent. Catalog oracles are prospective unless explicitly recorded as executed. Counts are viewpoints, not correctness evidence.
catalog
execution-ordered-workers: typed catalog
Epoch: 2026-10-29. Subject: exec.rs, SHA-256 180f88005a4cd3852f649059e420951c1fe9ad4a699c9d175358169f25e43b4d, bytes [929,2029). Every entry is a partial source-derived specification within its stated observation domain. Fixed dependencies and successful ordinary allocation are assumptions. Transition horizon is one invocation unless a history is explicitly named. Unsampled behavior remains unknown; no semantic proof is claimed. Checks below are prospective unless evidence.md records execution.
nonbehavior
Type: behavior_complement. Domain, observation and operator: U is positive-cap input/job-completion vectors paired with returned results. R joins every worker and concatenates in spawn order. U minus R contains Ok after a joined panic and Ok with a permutation different from ordered concatenation, under normal allocation/unwind.
Approximation: source-derived conditional specification; behavioral complements are candidate inner bounds of absence under the explicit assumptions. Reverse relations retain every compatible antecedent in the stated domain and do not imply unique recovery. Specialized observations omit unmentioned state and do not establish full semantic correspondence.
Falsifiable check: Existing order and panic tests.
antecedents
Type: reverse_reconstruction. Domain, observation and operator: Given y and deterministic f, each compatible x has length len(y) and x_i in f^-1(y_i). All compatible positive caps remain possible because output omits partition boundaries.
Approximation: source-derived conditional specification; behavioral complements are candidate inner bounds of absence under the explicit assumptions. Reverse relations retain every compatible antecedent in the stated domain and do not imply unique recovery. Specialized observations omit unmentioned state and do not establish full semantic correspondence.
Falsifiable check: Enumerate f(x)=x mod 2 on short sequences.
chunk-envelope
Type: partition_geometry. Domain, observation and operator: For c=max(ceil(n/t),1), worker j covers [jc,min((j+1)c,n)). These intervals partition [0,n); n=0 spawns zero workers.
Approximation: source-derived conditional specification; behavioral complements are candidate inner bounds of absence under the explicit assumptions. Reverse relations retain every compatible antecedent in the stated domain and do not imply unique recovery. Specialized observations omit unmentioned state and do not establish full semantic correspondence.
Falsifiable check: Enumerate n=0..101 and cap=1..16.
panic-frontier
Type: failure_recovery. Domain, observation and operator: Err retains existence of at least one join panic but discards worker index, payload and successful outputs. par_map changes the observation to an expect panic; completed effects are not reversed.
Approximation: source-derived conditional specification; behavioral complements are candidate inner bounds of absence under the explicit assumptions. Reverse relations retain every compatible antecedent in the stated domain and do not imply unique recovery. Specialized observations omit unmentioned state and do not establish full semantic correspondence.
Falsifiable check: Distinct panic payloads and completion counter.
tests
Checks
Existing focused check: cargo test -p world --lib exec::tests -- --nocapture. Named tests: batches_preserve_order_and_cardinality (101 mapped values and empty input); worker_panics_are_errors_instead_of_partial_results (four panicking jobs). The file-root evidence.md records actual execution. Mode caching, exit behavior, mixed worker outcomes, GPU shape guards and GPU/CPU agreement remain untested here.
No training, live service or GPU execution; no speedup or broad semantic correctness claim.
Native correspondence is now executed: full finite pointwise fibers, mixed panic prefixes, output-drop accounting and stateful ticket projections. See ../tests.md and ../worker-laws.md for domains and actual results. The parser and arbitrary callback effects remain outside these tests.
This page in Markdown → https://md.groupants.elamur.ai/free/en/mova/world/src/exec.rs.groupant/workers.groupant/