Compact child lifecycle

[SandroMaglione]

Summary

• reserve child starts and create their shared scope in one atomic runtime step
• close child scopes atomically and execute zero- or one-item cleanup directly
• specialize invoke start/stop traversal while retaining unbounded parallelism for multiple invokes and real finalizers
• cover concurrent child interruption and scoped-resource finalization explicitly

Why

The common parent-with-one-child lifecycle paid general collection construction and unbounded traversal machinery during both invoke startup and shutdown. It also split child scope creation and registry reservation across separate Effect boundaries.

Impact

The public API, type inference, ordering, interruption, and finalization contracts are unchanged. The full local three-process comparison improves parent-with-one-child lifecycle throughput from 16,271 to 17,699 families/s (+8.8%) and slightly reduces retained family heap from 17.0 to 16.8 KiB.

Changeset

• Added a patch changeset for the runtime lifecycle improvement.

Validation

• pnpm check
• pnpm perf:types
• 15-pair focused lifecycle comparison
• Three independent full base/candidate runtime benchmark processes
• Relevant example checks are covered by pnpm check
• Reviewed the independent CI runtime performance report

[github-actions]

Type performance

Measured with TypeScript 6.0.3 and skipLibCheck=true.

Scenario	Base	PR	Difference
Effect only	55	55	0 (0.0%)
Import effect-machine	55	55	0 (0.0%)
Machine.defineStates (3 states)	2,965	2,965	0 (0.0%)
Machine.make (3 states, 2 events)	8,715	8,715	0 (0.0%)
machine.handle (3 states, 2 transitions)	23,492	23,492	0 (0.0%)
machine.handle (depth 8)	117,075	117,075	0 (0.0%)
machine.handle (depth 12)	129,713	129,713	0 (0.0%)
machine.handle (depth 16)	145,055	145,055	0 (0.0%)
machine.handle (depth 24)	183,851	183,851	0 (0.0%)
machine.handle (wide depth 16)	214,189	214,189	0 (0.0%)
machine.handle (parallel/history/choice)	132,667	132,667	0 (0.0%)
machine.handle (4 successive calls)	128,674	128,674	0 (0.0%)
machine exact input/output/error/services	110,584	110,584	0 (0.0%)
execution adapter readiness	127,854	127,854	0 (0.0%)

Marginal instantiations are measured against the matching setup without that API call:

Scenario	Base	PR	Difference
Import effect-machine	0	0	0
Machine.defineStates (3 states)	2,910	2,910	0 (0.0%)
Machine.make (3 states, 2 events)	5,742	5,742	0 (0.0%)
machine.handle (3 states, 2 transitions)	14,777	14,777	0 (0.0%)
machine.handle (depth 8)	109,601	109,601	0 (0.0%)
machine.handle (depth 12)	121,039	121,039	0 (0.0%)
machine.handle (depth 16)	135,181	135,181	0 (0.0%)
machine.handle (depth 24)	171,577	171,577	0 (0.0%)
machine.handle (wide depth 16)	202,224	202,224	0 (0.0%)
machine.handle (parallel/history/choice)	114,331	114,331	0 (0.0%)
machine.handle (4 successive calls)	113,703	113,703	0 (0.0%)
machine exact input/output/error/services	100,670	100,670	0 (0.0%)
execution adapter readiness	102,092	102,092	0 (0.0%)

Check times (informational)

Scenario	Base	PR
Effect only	0.03s	0.03s
Import effect-machine	0.03s	0.03s
Machine.defineStates (3 states)	0.10s	0.11s
Machine.make (3 states, 2 events)	0.16s	0.16s
machine.handle (3 states, 2 transitions)	0.23s	0.25s
machine.handle (depth 8)	0.51s	0.51s
machine.handle (depth 12)	0.55s	0.58s
machine.handle (depth 16)	0.61s	0.62s
machine.handle (depth 24)	0.71s	0.71s
machine.handle (wide depth 16)	0.78s	0.78s
machine.handle (parallel/history/choice)	0.63s	0.63s
machine.handle (4 successive calls)	0.60s	0.61s
machine exact input/output/error/services	0.52s	0.52s
execution adapter readiness	0.58s	0.60s

Type instantiations are the comparison metric. Check time varies with runner load and is informational only.

[github-actions]

Runtime performance

Median of 5 independent benchmark processes on AMD EPYC 7763 64-Core Processor with Node v24.18.0.

Pull request baseline

Scenario	Effect Machine	XState 5	XState 6 alpha
Plan counter transitions	114,605 transitions/s	137,896 transitions/s	125,651 transitions/s
Drain burst with terminal fence	92,137 increments/s	419,052 increments/s	115,124 increments/s
Drain burst with a change observer	67,875 increments/s	414,492 increments/s	114,810 increments/s
Lookup and send to one child	85,211 increments/s	411,431 increments/s	109,587 increments/s
Start and stop a machine	18,769 machines/s	193,087 machines/s	74,377 machines/s
Start and stop a parent with one child	7,680 families/s	70,741 families/s	27,383 families/s

Effect runtime reference points

Scenario	Effect Machine	Effect runtime primitives
Start and stop a raw compiled process	30,357 processes/s	—
Start and interrupt a suspended fiber	—	167,168 fibers/s
Start and stop a queue worker	—	115,128 workers/s
Start and stop an actor shell	—	106,180 actors/s
Start and stop two actor shells	—	66,587 families/s
Update an owner-only mutable snapshot	—	317,863,954 updates/s
Update a synchronized snapshot	—	1,546,865 updates/s
Create, resolve, and await a terminal latch	—	1,435,416 latches/s

Memory profile	Effect Machine	XState 5	XState 6 alpha	Effect runtime primitives
Idle machine	7.5 KiB	3.8 KiB	6.1 KiB	—
Raw generic managed process	12.3 KiB	—	—	—
Raw compiled process	7.6 KiB	—	—	—
Two independent idle machines	14.5 KiB	7.3 KiB	11.8 KiB	—
Idle parent with one child	16.8 KiB	5.5 KiB	8.8 KiB	—
Parent with observed child registry	21.0 KiB	—	—	—
Parent with observed invoked child snapshots	17.0 KiB	—	—	—
Suspended Effect fiber	—	—	—	0.6 KiB
Effect queue with waiting fiber	—	—	—	2.8 KiB
Effect mailbox actor shell	—	—	—	3.2 KiB
Two Effect actor shells	—	—	—	6.9 KiB

Effect Machine change from base

Metric	Base	Base variability	PR	PR variability	Difference
Plan counter transitions	114,796 transitions/s	1.6% MAD	114,605 transitions/s	1.9% MAD	-0.2%
Drain burst with terminal fence	92,886 increments/s	2.2% MAD	92,137 increments/s	3.1% MAD	-0.8%
Drain burst with a change observer	67,992 increments/s	1.7% MAD	67,875 increments/s	1.3% MAD	-0.2%
Lookup and send to one child	86,421 increments/s	2.2% MAD	85,211 increments/s	3.5% MAD	-1.4%
Start and stop a machine	18,653 machines/s	2.0% MAD	18,769 machines/s	4.2% MAD	+0.6%
Start and stop a parent with one child	7,062 families/s	5.6% MAD	7,680 families/s	4.9% MAD	+8.7%
Idle machine heap per unit	7.5 KiB	0.0% MAD	7.5 KiB	0.0% MAD	-0.0%
Raw generic managed process heap per unit	12.5 KiB	0.0% MAD	12.3 KiB	0.0% MAD	-1.4%
Raw compiled process heap per unit	7.8 KiB	0.1% MAD	7.6 KiB	0.1% MAD	-1.5%
Two independent idle machines heap per unit	14.5 KiB	0.0% MAD	14.5 KiB	0.0% MAD	-0.0%
Idle parent with one child heap per unit	17.0 KiB	0.0% MAD	16.8 KiB	0.0% MAD	-1.2%
Parent with observed child registry heap per unit	21.2 KiB	0.0% MAD	21.0 KiB	0.0% MAD	-0.9%
Parent with observed invoked child snapshots heap per unit	17.2 KiB	0.0% MAD	17.0 KiB	0.0% MAD	-1.2%

Effect runtime reference change from base

Metric	Base	Base variability	PR	PR variability	Difference
Start and stop a raw compiled process	30,845 processes/s	1.4% MAD	30,357 processes/s	0.3% MAD	-1.6%
Start and interrupt a suspended fiber	168,039 fibers/s	1.2% MAD	167,168 fibers/s	0.8% MAD	-0.5%
Start and stop a queue worker	117,014 workers/s	0.0% MAD	115,128 workers/s	1.4% MAD	-1.6%
Start and stop an actor shell	106,078 actors/s	0.5% MAD	106,180 actors/s	2.1% MAD	+0.1%
Start and stop two actor shells	66,765 families/s	1.5% MAD	66,587 families/s	1.7% MAD	-0.3%
Update an owner-only mutable snapshot	317,863,954 updates/s	0.0% MAD	317,863,954 updates/s	0.0% MAD	0.0%
Update a synchronized snapshot	1,525,937 updates/s	3.0% MAD	1,546,865 updates/s	1.4% MAD	+1.4%
Create, resolve, and await a terminal latch	1,433,273 latches/s	2.1% MAD	1,435,416 latches/s	1.1% MAD	+0.1%

Versions and interpretation

• Effect Machine: 0.3.0
• XState 5: 5.32.5
• XState 6 alpha: 6.0.0-alpha.27
• Effect runtime primitives: 4.0.0-beta.105

Higher throughput is better; lower heap is better. Variability is the median absolute deviation across independent processes, relative to their median. Runtime measurements on shared GitHub-hosted hardware remain informational, so small differences should be confirmed across multiple workflow runs.