Refactor: Fine-Grained Reactive Data Context

[jlukic]

Implements Fine-Grained Reactive Data Context.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
semantic-next	Ready	Preview, Comment	May 6, 2026 1:00am

1 Skipped Deployment

Project	Deployment	Actions	Updated (UTC)
mcp	Ignored	Preview	May 6, 2026 1:00am

[semantic-performance-bot]

🟡 Mixed Performance (Net Positive) for f5f82a1 on Benchmark Suite 📊

Base: main · Action: #25410752474 · Raw: bench-report.json

^{Refactor: Fine-Grained Reactive Data Context}

Warning

This PR improves ✅ 7 tests while regressing on ❌ 2 tests.

✅ 7 faster · ❌ 2 slower · 🔍 19 unsure · ⚪ 36 no change · 🏆 2 new peaks · 📜 9 reopened

---

✅ Faster (7)

Metrics where this PR confidently improved performance compared to main.

metric	Improvement
clear-completed-250	-52% (29ms) 🌟
remove-row-front-20	-39% (207ms) 🌟
remove-first-10	-12% (21ms)
remove-5-front	-7% (6ms)
remove-5-back	-6% (4ms)
remove-last-10	-5% (7ms)
toggle-first-10	-3% (5ms)

❌ Slower (2)

Metrics where this PR confidently regressed performance compared to main.

metric	Regression
bulk-add-500	+10% (22ms)
each-mount-1000	+4% (2ms)

🏆 New peaks (2)

These metrics reached a new best in this iteration — current's percent-delta vs its baseline dominates the prior peak's percent-delta vs its baseline. Credit candidates are the commits between the prior peak and HEAD; nearest-to-current is usually the cause.

metric	current	prior peak	vs prior peak	credit candidates
remove-5-back	-6%	-2% @ aaeb304	improved 4%	9d8761e
toggle-10	-2%	+1% @ 9d8761e	improved 2%	—

📜 Regressions from peak (9)

These metrics were better on a prior iteration than they are now. The peak's percent-delta vs its baseline dominates current's percent-delta vs its baseline — not attributable to per-sample noise. Bisect candidates are the commits between the peak iteration and HEAD; nearest-to-peak is usually the best bet.

metric	current	peak	vs peak	bisect candidates
remove-row-middle-20	-1%	-21% @ aaeb304	regressed +20%	9d8761e
remove-row-front-20	-39%	-56% @ aaeb304	regressed +18%	9d8761e
filter-cycle-20	+2%	-11% @ aaeb304	regressed +13%	9d8761e
bulk-add-500	+10%	+1% @ aaeb304	regressed +8%	9d8761e
remove-5-front	-7%	-13% @ aaeb304	regressed +6%	9d8761e
each-mount-1000	+4%	+0% @ aaeb304	regressed +4%	9d8761e
edit-start-10	+2%	-1% @ 9d8761e	regressed +3%	—
remove-10-middle	+0%	-2% @ 9d8761e	regressed +2%	—
remove-first-10	-12%	-14% @ 9d8761e	regressed +2%	—

⚪ No Change (36)

Metrics where this PR measured within ±2% of main — no meaningful performance change detected.

metric	Change
active-indicator-200	0.0% – 0.0%
active-indicator-nested-200	-0.1% – -0.0%
add-20	-0.1% – 0.0%
edit-cycle-5	-1.8% – -1.6%
hydrate-each-100	+0.7% – +1.8%
hydrate-each-100-mount	-0.7% – +0.6%
hydrate-helper-100-state-change	-0.3% – +1.0%
micro-build-html-string-10k	-0.2% – +1.9%
micro-compiler-parse-cold-complex-200	-0.6% – +0.7%
micro-compiler-parse-cold-normal-500	-1.1% – +1.4%
micro-compiler-snippet-args-5k	-0.7% – +0.2%
micro-dom-walker-1000x15	-0.5% – +1.8%
micro-expr-js-10k	-0.8% – +1.6%
micro-expr-lisp-50k	-1.3% – +0.5%
micro-expr-simple-100k	-0.9% – +1.0%
reaction-coalesce-200x100	+0.0% – +1.8%
reaction-dep-diff-30k	-0.8% – +1.9%
remove-10-middle	-0.1% – +0.7%
remove-middle-10	-1.6% – -1.2%
remove-row-middle-20	-1.7% – -0.2%
rename-50	-0.3% – -0.2%
signal-computed-chain-10x60k	+0.3% – +1.1%
signal-reactive-fanout-500x1200	-1.1% – +0.9%
signal-reactive-list-replace-1000x1000	-0.5% – +1.0%
signal-reactive-multi-read-5x160k	-1.1% – +0.5%
signal-reactive-set-property-by-id-200	-0.5% – +1.3%
signal-set-same-10m	-2.0% – +1.5%
snippet-args-per-key-100	-0.0% – 0.0%
snippet-in-subtemplate-100	-0.0% – +0.0%
stable-ref-mutate-500	0.0% – 0.0%
subtemplate-data-blob-100	-0.6% – -0.5%
subtemplate-reactive-data-100	-1.4% – -1.0%
subtemplate-shorthand-props-100	-0.5% – -0.3%
swap-rows-20	-1.1% – +1.4%
toggle-10	-2.0% – -1.3%
toggle-all-20	-1.1% – -0.9%

🔍 Unsure (19)

Inconclusive (11)

The measured difference is small, and our sampling couldn't confidently place it above or below zero. Running more samples in a future run might settle these metrics.

metric	Change	Expected Noise
append-1k	-6.2% – -0.7%	±2%
clear-10k	+0.2% – +8.4%	±1%
create-10k	+0.8% – +2.0%	±0%
create-1k	+0.7% – +4.4%	±2%
edit-start-10	+0.6% – +2.5%	±1%
filter-cycle-20	+0.8% – +3.3%	±0%
remove-row-back-10	+0.7% – +2.9%	±1%
select-40	-2.6% – -1.3%	±0%
signal-reactive-list-filter-1000x300	-0.8% – +2.1%	±1%
signal-reactive-push-2000x20	-2.4% – +0.3%	±1%
update-10th-10	-2.1% – +1.7%	±1%

Too Fast to Measure Precisely (8)

On benches this short, system jitter (scheduling, GC, JIT) masks sub-4% changes; larger deltas still resolve cleanly.

metric	Change	Test Time	Expected Noise
hydrate-helper-100-mount	-2.3% – -0.1%	~45ms	±3%
micro-compiler-ast-walk-5k	-0.1% – +3.9%	~15ms	±10%
reaction-flush-noop-5m	-0.7% – +2.9%	~56ms	±3%
replace-1k	-1.0% – +2.1%	~79ms	±2%
signal-reactive-set-index-300	-2.1% – -0.0%	~103ms	±2%
signal-sub-unsub-100k	-0.7% – +3.6%	~21ms	±8%
toggle-last-10	-2.3% – -1.1%	~155ms	±1%
toggle-middle-10	-2.3% – -1.6%	~159ms	±1%

📖 Bench glossary (56 metrics)

metric	what it tests
add-20	Appends 20 todo items one at a time, like a user typing entries in a row.
append-1k	Appends 1000 new rows onto an existing 1000-row table.
bulk-add-500	Renders 500 todo items added at once from a single data load.
clear-10k	Clears a 10000-row table back to empty in a single operation.
clear-completed-250	Clears 250 completed items from a 500-item list in one action, like clicking clear completed.
create-10k	Renders a fresh 10000-row table into an empty parent at ten times the create-1k scale.
create-1k	Renders a fresh 1000-row table into an empty parent.
each-mount-1000	Mounts a fresh 1000-item each block with five-field items so per-record allocation cost dominates the wall clock.
edit-cycle-5	Runs 5 full edit-then-save cycles on different items, like editing a row and saving it.
edit-start-10	Enters edit mode on 10 different items in a row, like double-clicking each one.
filter-cycle-20	Cycles through active, completed, and all filters 20 times on a 100-item list.
hydrate-each-100	Reassigns the items of a hydrated 1000-item list to a fresh array with the same keys and data.
hydrate-each-100-mount	Hydrates a server-rendered 1000-item list and waits for it to become interactive without re-rendering.
hydrate-helper-100-mount	Hydrates a 1000-item list where each item calls a helper that reads state shared across the list.
hydrate-helper-100-state-change	Cycles the shared activeID through 10 different items in a hydrated 1000-item list so two items repaint per cycle.
micro-build-html-string-10k	Builds the HTML string for a realistic card AST 10000 times. Raw assembly throughput.
micro-compiler-ast-walk-5k	Walks a kitchen-sink AST through optimizeAST 5000 times. Merge, hoist, and recurse pass.
micro-compiler-parse-cold-complex-200	Compiles a feature-dense kitchen-sink template 200 times. Catches parser regressions on uncommon block paths.
micro-compiler-parse-cold-normal-500	Compiles a TodoMVC-style component template 500 times. Headline metric for normal-component compile throughput.
micro-compiler-snippet-args-5k	Parses four representative subtemplate-call shapes 5000 times each. Snippet args extraction.
micro-dom-walker-1000x15	Runs bindMarkers across a 1000-node card fragment 15 times. TreeWalker pass and binding dispatch.
micro-expr-js-10k	Evaluates one arithmetic expression and one ternary 10000 times each. JS-eval hot path.
micro-expr-lisp-50k	Evaluates one Lisp-style helper call 50000 times. Parse-cache lookup and helper dispatch.
micro-expr-simple-100k	Evaluates one simple identifier and one dotted path 100000 times each. Property-lookup hot path.
reaction-coalesce-200x100	Sets one signal 100 times then flushes once across 200 bursts so 100 subscribers wake one time per burst.
reaction-dep-diff-30k	Toggles which of two signals a subscriber reads across 30000 cycles. Per-run dep-set diffing.
reaction-flush-noop-5m	Calls Reaction.flush() 5000000 times with no pending work. Scheduler dispatch overhead.
remove-10-middle	Deletes 10 items from the middle of a 100-item list, one click at a time.
remove-5-back	Deletes 5 items from the end of a 100-item list, one click at a time.
remove-5-front	Deletes 5 items from the front of a 100-item list, one click at a time.
remove-first-10	Deletes the first item 10 times from a 100-item list, with remaining items moving up each time.
remove-last-10	Deletes the last item 10 times from a 100-item list, with no other items needing to move.
remove-middle-10	Deletes the middle item 10 times from a 100-item list, walking halfway through to find each target.
remove-row-back-10	Removes the last row 10 times from a 1000-row table, with no other rows needing to move.
remove-row-front-20	Removes the first row 20 times from a 1000-row table, with all remaining rows sliding up each time.
remove-row-middle-20	Removes the middle row 20 times from a 1000-row table, with the rows below it sliding up each time.
rename-50	Renames 50 different items in a 100-item list via single-field setProperty without editingId co-fires.
replace-1k	Replaces 1000 rows with a fresh 1000-row set, diffing the keyed list against a populated table.
select-40	Highlights one row at a time across 40 rows so only the previous and newly highlighted rows update.
signal-computed-chain-10x60k	Propagates a value change from root to leaf through a 10-deep chain of derived signals 60000 times.
signal-reactive-fanout-500x1200	Fans out one signal's value change to 500 subscribers across 1200 successive updates.
signal-reactive-list-filter-1000x300	Changes a search-term signal 300 times, re-scanning a 1000-item list on each change.
signal-reactive-list-replace-1000x1000	Replaces a 1000-item list signal with a fresh 1000-item array and rescans it 1000 times.
signal-reactive-multi-read-5x160k	Changes five signals in turn for 32000 rounds with one subscriber reading all five.
signal-reactive-push-2000x20	Appends 20 items onto an empty list signal with a subscriber, across 2000 reset cycles.
signal-reactive-set-index-300	Replaces one item by index in a 1000-item list signal across 300 updates, with a subscriber.
signal-reactive-set-property-by-id-200	Finds an item by id and updates one field in a 1000-item list signal across 200 alternating updates.
signal-set-same-10m	Sets a signal to its current value 10000000 times. Exercises the no-op fast path when nothing changes.
signal-sub-unsub-100k	Creates and tears down a subscriber on one signal across 100000 cycles. Subscription churn cost.
swap-rows-20	Swaps the second and second-to-last rows in a 1000-row table, repeated 20 times.
toggle-10	Checks off the first 10 items one by one, like a user working down a list.
toggle-all-20	Toggles all 100 items completed and back across 20 cycles via the master checkbox.
toggle-first-10	Toggles the first item in a 100-item list ten times, alternating completed on and off.
toggle-last-10	Toggles the last item in a 100-item list ten times, alternating completed on and off.
toggle-middle-10	Toggles a middle item in a 100-item list ten times, alternating completed on and off.
update-10th-10	Updates the label on every tenth row of a 1000-row table, looped ten times to lift the work above noise.

---

_{Sample size: 50 · Resolution floor: ±2% · Timeout: 3min · Wall-clock: 21m24s}