🇨🇿 PragueJS 2023 #4

Slide 1

Slide 1 text

Hello, PragueJS! 👋 🇨🇿 DEEP DIVING ON CONCURRENT REACT • THE 27TH OF APRIL, 2023.

Slide 2

Slide 2 text

DEEP DIVING ON CONCURRENT REACT MATHEUS ALBUQUERQUE

Slide 3

Slide 3 text

DEEP DIVING ON CONCURRENT REACT ↑ ALL THE LINKS! 🤓 🧑🏫 @techlabs 🐦 @ythecombinator 👨💻 @medallia ⚡ Perf GDE

Slide 4

Slide 4 text

DEEP DIVING ON CONCURRENT REACT ↑ ALL THE LINKS! 🤓 🍺 LET’S DISCUSS MORE AFTER!

Slide 5

Slide 5 text

DISCLAIMER

Slide 6

Slide 6 text

#QUESTION 🤔 If you were to summarize Concurrent React in one word/expression, what’d be your pick?

Slide 7

Slide 7 text

#QUESTION 🤔 If you were to summarize Concurrent React in one word/expression, what’d be your pick? e.g. fibers = units of work Concurrent React = ???

Slide 8

Slide 8 text

DEEP DIVING ON CONCURRENT REACT If you were to summarize Concurrent React in one word/expression, what’d be your pick?

Slide 9

Slide 9 text

The Main Thread DEEP DIVING ON CONCURRENT REACT

Slide 10

Slide 10 text

TASKS JAVASCRIPT STYLES LAYOUT PAINT COMPOSITE A UNIT OF WORK THAT THE BROWSER DOES TO RENDER A FRAME

Slide 11

Slide 11 text

TASKS JAVASCRIPT STYLES LAYOUT PAINT COMPOSITE

Slide 12

Slide 12 text

LONG TASKS ↝ IF A TASK TAKES MORE THAN 50 MS, USER INPUT FEELS DELAYED ↝ BASED ON THE USER-CENTRIC PERFORMANCE MODEL CALLED RAIL ↝ THEY TAKE TOO LONG AND BLOCK OTHER TASKS

Slide 13

Slide 13 text

LONG TASKS

Slide 14

Slide 14 text

No content

Slide 15

Slide 15 text

LONG TASKS

Slide 16

Slide 16 text

LONG TASKS

Slide 17

Slide 17 text

#RESEARCH 📚 Phone users experience slow First Input Delay on 7x more websites. — Web Almanac By HTTP Archive, 2021

Slide 18

Slide 18 text

RESEARCH RESULTS: ↝ LONG TASKS DELAYED TTI ↝ MOBILE HAD UP TO ˜12X LONGER LONG TASKS ↝ OLDER DEVICES COULD BE SPENDING HALF OF THEIR LOAD-TIME ON LONG TASKS — AKAMAI AND CHROME RESEARCH, 2017 LONG TASKS

Slide 19

Slide 19 text

— AKAMAI AND CHROME RESEARCH, 2017 BUSINESS OUTCOMES

Slide 20

Slide 20 text

#QUESTION 🤔 How to avoid blocking the main thread?

Slide 21

Slide 21 text

TASK RUNNING STRATEGIES A B C D

Slide 22

Slide 22 text

TASK RUNNING STRATEGIES PARALLELISM CONCURRENCY SCHEDULING

Slide 23

Slide 23 text

TASK RUNNING STRATEGIES PARALLELISM CONCURRENCY SCHEDULING

Slide 24

Slide 24 text

WORKERS ↝ DATA EXCHANGE IS THROUGH MESSAGE-PASSING ↝ NO ACCESS TO ANY VARIABLES/CODE FROM THE PAGE THAT CREATED THEM OR VICE VERSA ↝ NO ACCESS TO THE DOM, MAKING UI UPDATES FROM A WORKER BARELY IMPOSSIBLE ↝ TWO MODELS: ACTORS & SHARED MEMORY 🤿

Slide 25

Slide 25 text

ACTORS ↝ EACH ACTOR MAY OR MAY NOT RUN ON A SEPARATE THREAD ↝ EACH ACTOR FULLY OWNS THE DATA IT IS OPERATING ON ↝ ACTORS CAN ONLY SEND/REACT TO MESSAGES ↝ MAIN THREAD = ACTOR THAT OWNS THE DOM/UI 🤿

Slide 26

Slide 26 text

ACTORS ↝ EVERY MESSAGE WE SEND NEEDS TO BE COPIED ↝ BALANCE: MOVING CODE TO A WORKER VS COMMUNICATION OVERHEAD/WORKER BEING BUSY ↝ postMessage IS A FIRE-AND-FORGET MESSAGING MECHANISM WITH NO BUILT-IN UNDERSTANDING OF REQUEST AND RESPONSE 🤿

Slide 27

Slide 27 text

SHARED MEMORY ↝ ONE DEDICATED TYPE: SharedArrayBuffer ↝ A LINEAR CHUNK OF MEMORY THAT CAN BE MANIPULATED USING TypedArrays OR DataViews ↝ IF SENT VIA postMessage, THE OTHER END GETS A HANDLE TO THE EXACT SAME MEMORY CHUNK 🤿

Slide 28

Slide 28 text

SHARED MEMORY ↝ MOST OF THE APIS ARE BUILT NO CONCURRENT ACCESS TO OBJECTS IN MIND ↝ YOU BUILD YOUR OWN MUTEXES AND OTHER CONCURRENT DATA STRUCTURES ↝ NO DIRECT WAY OF WORKING ON FAMILIAR OBJECTS/ARRAYS; JUST A SERIES OF BYTES 🤿

Slide 29

Slide 29 text

WEB ASSEMBLY ↝ WORKERS + SharedArrayBuffers TO SUPPORT THE THREADING MODEL OF C++ AND OTHERS ↝ BEST EXPERIENCE FOR SHARED-MEMORY MODEL ↝ FASTER THAN JS WHEN YOU STAY WITHIN WASM, BUT THE MORE YOU HAVE TO CROSS OVER TO JS APIS THE SLOWER IT IS

Slide 30

Slide 30 text

WEB ASSEMBLY ↝ JAVASCRIPT IS OFTEN FASTER AT DOING DOM RENDERING ↝ HIGH-LEVEL LIBRARIES CAN BE MORE PERFORMANT THAN LOW-LEVEL WASM IMPLEMENTATIONS ↝ DOESN’T OFFER LOT OF THE BENEFITS (AND COMFORT) OF JAVASCRIPT

Slide 31

Slide 31 text

↝ Atomics ↝ BuffferBackedObject ↝ Comlink ↝ WorkerDOM ↝ AND MUCH MORE!

Slide 32

Slide 32 text

WORKERS ↝ GOOD FOR DATA PROCESSING AND CRUNCHING NUMBERS ↝ HARD TO USE FOR UI-RELATED STUFF ↝ HARDER THAN ADJUSTING WORK FOR A SCHEDULER

Slide 33

Slide 33 text

TASK RUNNING STRATEGIES PARALLELISM CONCURRENCY SCHEDULING

Slide 34

Slide 34 text

#QUESTION 🤔 If you were to summarize Concurrent React in one word/expression, what’d be your pick?

Slide 35

Slide 35 text

Scheduling in React DEEP DIVING ON CONCURRENT REACT

Slide 36

Slide 36 text

HEURISTICS COOPERATIVE MULTITASKING WITH A SINGLE INTERRUPTIBLE RENDERING THREAD PRIORITY LEVELS REGISTER CALLBACKS WITH DIFFERENT PRIORITY LEVELS IN THE BROWSER RENDER LANES ABSTRACTIONS AROUND A BITMASK; BRING GRANULARITY, AVOID OVERHEAD & ALLOW BATCHING SCHEDULING IN REACT

Slide 37

Slide 37 text

{result}

; }

Slide 38

Slide 38 text

DEEP DIVING ON CONCURRENT REACT

Slide 39

Slide 39 text

#QUESTION 🤔 How could we improve that?

Slide 40

Slide 40 text

DEEP DIVING ON CONCURRENT REACT

Slide 41

Slide 41 text

No content

Slide 42

Slide 42 text

{result}

; }

Slide 43

Slide 43 text

function* resourcefulOperation(value: number) { let newValue = String(value); while (true) { yield; for (let i = 0; i < 1000000; i++) { newValue = `${value} + ${i} = ${value + i}`; } return newValue; } } const initialValue = 0; const scheduler = new Scheduler(resourcefulOperation, initialValue); function ResourcefulComponent(props: { value: number }) { const { value } = props; const result = scheduler.performUnitOfWork(value); return

{result}

; }

Slide 44

Slide 44 text

{result}

; } PROMOTED TO A GENERATOR YIELDING EXECUTION DOING CONCURRENT TASKS

Slide 45

Slide 45 text

DEEP DIVING ON CONCURRENT REACT

Slide 46

Slide 46 text

enum SchedulerState { IDLE = "IDLE", PENDING = "PENDING", DONE = "DONE", } class Scheduler { state: SchedulerState; result: T; worker: (data: T) = > Generator; iterator: Generator; constructor(worker: (data: T) = > Generator, initialResult: T) { this.state = SchedulerState.IDLE; this.worker = worker; this.result = initialResult; } performUnitOfWork(data: T) { switch (this.state) { case "IDLE": this.state = SchedulerState.PENDING; this.iterator = this.worker(data); throw Promise.resolve(); case "PENDING": const { value, done } = this.iterator.next(); if (done) { this.result = value; this.state = SchedulerState.DONE; return value; } throw Promise.resolve(); case "DONE": this.state = SchedulerState.IDLE; return this.result; } } }

Slide 47

Slide 47 text

performUnitOfWork(data: T) { switch (this.state) { case "IDLE": this.state = SchedulerState.PENDING; this.iterator = this.worker(data); throw Promise.resolve(); case "PENDING": const { value, done } = this.iterator.next(); if (done) { this.result = value; this.state = SchedulerState.DONE; return value; } throw Promise.resolve(); case "DONE": this.state = SchedulerState.IDLE; return this.result; } }

Slide 48

Slide 48 text

DID WE JUST useTransition’ED? 🤔

Slide 49

Slide 49 text

{result}

; }

Slide 50

Slide 50 text

{result}

; }

Slide 51

Slide 51 text

YES, WE DID 🤓 WITH OUR OWN SCHEDULER

Slide 52

Slide 52 text

↝ A COOPERATIVE MULTITASKING MODEL ↝ A SINGLE INTERRUPTIBLE RENDERING THREAD ↝ RENDERING CAN BE INTERLEAVED WITH OTHER MAIN THREAD TASKS (AND OTHER REACT RENDERS) ↝ AN UPDATE CAN HAPPEN IN THE BACKGROUND WITHOUT BLOCKING THE RESPONSE TO NEW INPUT SCHEDULING IN REACT

Slide 53

Slide 53 text

↓ ORIGINAL RENDER TASK USER INPUT → ↑ HIGHER PRIORITY RENDER TASK ↓ RESUME ORIGINAL RENDER TASK

Slide 54

Slide 54 text

↝ IT YIELDS EXECUTION IS BACK TO THE MAIN THREAD EVERY 5MS ↝ IT'S SMALLER THAN A SINGLE FRAME EVEN ON 120FPS, SO IT WON'T BLOCK ANIMATIONS ↝ IN PRACTICE, RENDERING IS INTERRUPTIBLE HEURISTICS

Slide 55

Slide 55 text

PRIORITY LEVELS

Slide 56

Slide 56 text

PRIORITY LEVELS

Slide 57

Slide 57 text

PRIORITY TIMEOUT WHEN I m m ediate SYNCHRONOUSLY TASKS THAT NEED TO RUN SYNCHRONOUSLY UserBlocking 250MS RESULTS OF A USER INTERACTION (E.G. A BUTTON CLICK) Normal 5S UPDATES THAT DON’T HAVE TO FEEL INSTANTANEOUS Low 10S TASKS THAT CAN BE DEFERRED BUT MUST STILL COMPLETE EVENTUALLY (E.G. AN ANALYTICS NOTIFICATION) Idle NO TIMEOUT TASKS THAT DO NOT HAVE TO RUN AT ALL (E.G. HIDDEN OFFSCREEN CONTENT)

Slide 58

Slide 58 text

↝ ONE LANE = ONE BIT IN A BITMASK ↝ ONE UPDATE IN REACT = ONE LANE ↝ UPDATES IN THE SAME LANE RENDER IN THE SAME BATCH. DIFFERENT LANES, SEPARATE BATCHES. ↝ 31 LEVELS OF GRANULARITY (= ONE BITMASK) RENDER LANES 🤿

Slide 59

Slide 59 text

↝ ALLOWS TO CHOOSE WHETHER TO RENDER MULTIPLE TRANSITIONS IN A SINGLE BATCH OR RENDER THEM INDEPENDENTLY ↝ REDUCES OVERHEAD OF MULTIPLE LAYOUT PASSES, STYLE RECALCULATIONS, AND MULTIPLE PAINTS RENDER LANES 🤿

Slide 60

Slide 60 text

No content

Slide 61

Slide 61 text

DEEP DIVING ON CONCURRENT REACT

Slide 62

Slide 62 text

#QUESTION 🤔 How do we benefit from these in our everyday projects?

Slide 63

Slide 63 text

Scheduling in React [for the rest of us] DEEP DIVING ON CONCURRENT REACT

Slide 64

Slide 64 text

SCHEDULING IN REACT HANDLING LOTS OF DATA WITH THE useTransition HOOK TACKLING WASTED RENDERS WITH THE useSyncExternalStore HOOK HYDRATION IMPROVEMENTS WITH SELECTIVE HYDRATION & CONCURRENT REACT PROFILER ENHANCEMENTS INSPECT TRANSITIONS, GET WARNS, AND MUCH MORE!

Slide 65

Slide 65 text

#1 HANDLING LARGE SETS OF DATA DEEP DIVING ON CONCURRENT REACT

Slide 66

Slide 66 text

😔 NON-PRACTICAL… ↝ FINDING PRIMES ↝ CRACKING PASSWORDS ↝ SIERPINSKI TRIANGLE

Slide 67

Slide 67 text

😔 NON-PRACTICAL… ↝ FINDING PRIMES ↝ CRACKING PASSWORDS ↝ SIERPINSKI TRIANGLE 😊 PRACTICAL… ↝ RENDERING MANY DATA-POINTS ↝ RENDERING ON A ↝ PROCESSING DATA

Slide 68

Slide 68 text

DAILY VISITORS (BEFORE)

Slide 69

Slide 69 text

DAILY VISITORS (BEFORE) const DailyVisitors = () = > { const [data, setData] = useState(initialData); useEffect(() = > { setData(initialData); }, []); const onChange = (newData) = > { setData(newData); }; return ( ); }; export default DailyVisitors;

Slide 70

Slide 70 text

DAILY VISITORS (AFTER) const DailyVisitors = () = > { const [data, setData] = useState(initialData); const [, startTransition] = useTransition(); useEffect(() = > { setData(initialData); }, []); const onChange = (newData) = > { startTransition(() = > { setData(newData); }); }; return ( ); }; export default DailyVisitors;

Slide 71

Slide 71 text

DAILY VISITORS (AFTER)

Slide 72

Slide 72 text

↝ ˜100K + POINTS PLOTTED ↝ SUPPORT FOR SEARCHING AND FILTERING ↝ USED WORKERS + REDUX-SAGA UTILITIES + DEBOUNCING ↝ COULD'VE USED TRANSITIONS CASE #1: MAPS

Slide 73

Slide 73 text

CASE #2: GAME ADMIN ↝ THOUSANDS OF REAL-TIME PLAYERS MESSAGING ↝ SUPPORT FOR SEARCHING AND FILTERING ↝ USED VIRTUALIZATION AND MEMOIZATION ↝ COULD'VE USED TRANSITIONS

Slide 74

Slide 74 text

#2 TACKLING WASTED RENDERS DEEP DIVING ON CONCURRENT REACT

Slide 75

Slide 75 text

useSyncExternalStore() function useSyncExternalStore( subscribe: (onStoreChange: () = > void) = > () = > void, getSnapshot: () = > Snapshot, getServerSnapshot?: () = > Snapshot ): Snapshot;

Slide 76

Slide 76 text

DEEP DIVING ON CONCURRENT REACT

Slide 77

Slide 77 text

No content

Slide 78

Slide 78 text

#QUESTION 🤔 How do we benefit from these in our everyday projects?

Slide 79

Slide 79 text

No content

Slide 80

Slide 80 text

useLocation() function Pathname() { const { pathname } = useLocation(); return ; } function Hash() { const { hash } = useLocation(); return ; }

Slide 81

Slide 81 text

useLocation() function Pathname() { const { pathname } = useLocation(); return ; } function Hash() { const { hash } = useLocation(); return ; } OVER-RETURNING HOOK

Slide 82

Slide 82 text

SPEAKERS LIST [BEFORE]

Slide 83

Slide 83 text

useLocation() function Pathname() { const { pathname } = useLocation(); return ; } function Hash() { const { hash } = useLocation(); return ; }

Slide 84

Slide 84 text

useLocation() useHistorySelector() function useHistorySelector(selector) { const history = useHistory(); return useSyncExternalStore(history.listen, () = > selector(history)); } function Pathname() { const pathname = useHistorySelector((history) = > history.location.pathname); return ; } function Hash() { const hash = useHistorySelector((history) = > history.location.hash); return ; }

Slide 85

Slide 85 text

Slide 86

Slide 86 text

SPEAKERS LIST [AFTER]

Slide 87

Slide 87 text

#3 HYDRATION IMPROVEMENTS DEEP DIVING ON CONCURRENT REACT

Slide 88

Slide 88 text

↝ BEFORE, HYDRATION COULD ONLY BEGIN AFTER THE ENTIRE DATA WAS FETCHED AND RENDERED ↝ USERS COULDN’T INTERACT WITH THE PAGE UNTIL HYDRATION WAS COMPLETE FOR THE WHOLE PAGE ↝ PARTS OF YOUR APP THAT LOAD FAST WOULD ALWAYS HAVE TO WAIT FOR THE SLOW ONES HYDRATION

Slide 89

Slide 89 text

↝ REACT WON'T WAIT FOR A COMPONENT TO LOAD TO CONTINUE STREAMING THE REST OF THE HTML ↝ REACT PRIORITIZES HYDRATING THE PARTS THAT THE USER INTERACTED WITH BEFORE THE REST SELECTIVE HYDRATION

Slide 90

Slide 90 text

↝ COMPONENTS CAN BECOME INTERACTIVE FASTER BY ALLOWING THE BROWSER TO DO OTHER WORK AT THE SAME TIME AS HYDRATION ↝ RESULTS IN LOWER FIRST INPUT DELAY (FID) AND INTERACTION TO NEXT PAINT (INP) SELECTIVE HYDRATION

Slide 91

Slide 91 text

DEEP DIVING ON CONCURRENT REACT

Slide 92

Slide 92 text

#4 PROFILER ENHANCEMENTS DEEP DIVING ON CONCURRENT REACT

Slide 93

Slide 93 text

TRANSITIONS — INTRODUCING A NEW REACT PROFILER, BY BRIAN VAUGHN

Slide 94

Slide 94 text

— INTRODUCING A NEW REACT PROFILER, BY BRIAN VAUGHN WARNS

Slide 95

Slide 95 text

The Future DEEP DIVING ON CONCURRENT REACT

Slide 96

Slide 96 text

↝ I/O LIBRARIES LIKE react-fetch ↝ BUILT-IN FOR DATA FETCHING LIBRARIES TO INTEGRATE WITH ↝ FOR CPU-BOUND TREES TO IMMEDIATELY FALLBACK WITHOUT EVEN TRYING TO RENDER THE FUTURE 🤿

Slide 97

Slide 97 text

↝ useInsertionEffect FOR STYLESHEET LIBRARIES ↝ THE COMPONENT ↝ SERVER COMPONENTS ↝ NATIVE SCHEDULING PRIMITIVES ON THE BROWSER THE FUTURE 🤿

Slide 98

Slide 98 text

DEEP DIVING ON CONCURRENT REACT

Slide 99

Slide 99 text

SCHEDULING API ↝ A MORE ROBUST SOLUTION FOR SCHEDULING TASKS ↝ CONTROL AND SCHEDULE PRIORITIZED TASKS IN A UNITED AND FLEXIBLE WAY ↝ INTEGRATED DIRECTLY INTO THE EVENT LOOP ↝ ALIGNED WITH THE WORK OF THE REACT TEAM AND IN COOPERATION WITH GOOGLE, W3C AND OTHERS

Slide 100

Slide 100 text

scheduler.postTask() SCHEDULE AND CONTROL PRIORITIZING TASKS. scheduler.wait() YIELD AND RESUME AFTER SOME AMOUNT OF TIME OR PERHAPS AFTER AN EVENT HAS OCCURRED. scheduler.yield() BREAK UP LONG TASKS BY YIELDING TO THE BROWSER AND CONTINUING AFTER BEING RESCHEDULED. isInputPending() DETERMINE IF THE CURRENT TASK IS BLOCKING INPUT EVENTS. SCHEDULING API

Slide 101

Slide 101 text

scheduler.postTask() scheduler.postTask(() = > { console.log('React Brussels'); }, { delay: 10 }); scheduler.postTask(() = > { console.log('React India'); }); scheduler.postTask(() = > { console.log('React Alicante'); }); / / 'React India' 'React Alicante' 'React Brussels'

Slide 102

Slide 102 text

scheduler.postTask() const controller = new TaskController({ priority: "user-blocking" }); const signal = controller.signal; console.log(signal.priority); / / 'user-blocking' console.log(signal.aborted); / / 'false' scheduler.postTask(doWork, { signal }); controller.setPriority("background"); controller.abort();

Slide 103

Slide 103 text

isInputPending() while (workQueue.length > 0) { if (navigator.scheduling.isInputPending()) { / / Stop doing work to handle any input event break; } let job = workQueue.shift(); job.execute(); }

Slide 104

Slide 104 text

SCHEDULING API

Slide 105

Slide 105 text

Closing Notes DEEP DIVING ON CONCURRENT REACT

Slide 106

Slide 106 text

DEEP DIVING ON CONCURRENT REACT

Slide 107

Slide 107 text

#1 DEEP DIVING ON CONCURRENT REACT REACT IS NOT REACTIVE, BUT IT IS CONCURRENT AND THAT MIGHT BE ENOUGH FOR YOU

Slide 108

Slide 108 text

No content

Slide 109

Slide 109 text

#2 DEEP DIVING ON CONCURRENT REACT REACT HAS BEEN PUSHING WEB APIS TO THE FUTURE E.G. THE SCHEDULER API AND DISCUSSIONS AROUND EFFECT HANDLERS

Slide 110

Slide 110 text

#3 DEEP DIVING ON CONCURRENT REACT REACT TRIES TO ADDRESS THE LACK OF SOME JS/WEB PLATFORM RESOURCES E.G. EFFECT HANDLERS, CONTINUATIONS & THE SCHEDULER API

Slide 111

Slide 111 text

DEEP DIVING ON CONCURRENT REACT

Slide 112

Slide 112 text

🤿

Slide 113

Slide 113 text

export const Hello = () = > { const value = usePromise(() = > delay("Hello, React Advanced! 👋 🇬🇧 ", 5000)); return {value}; }; function Demo() { return ( }> ); } 🤿

Slide 114

Slide 114 text

DID WE JUST CREATE React.use()? 🤔

Slide 115

Slide 115 text

YES, WE DID 🤓

Slide 116

Slide 116 text

#4 DEEP DIVING ON CONCURRENT REACT UNDERSTANDING THESE INTERNALS AND THEIR RATIONALES HELPS US IMPLEMENT OUR OWN ABSTRACTIONS E.G. THE GENERATOR-BASED SCHEDULER & FIRST CLASS SUPPORT FOR PROMISES

Slide 117

Slide 117 text

#5 DEEP DIVING ON CONCURRENT REACT SCHEDULING DOES NOT NECESSARILY MEAN BETTER PERFORMANCE

Slide 118

Slide 118 text

#6 DEEP DIVING ON CONCURRENT REACT THERE'S NO SILVER BULLET. IDENTIFY YOUR CORE METRICS.

Slide 119

Slide 119 text

#7 DEEP DIVING ON CONCURRENT REACT THERE’S A LOT OF INFORMATION OUT THERE

Slide 120

Slide 120 text

No content

Slide 121

Slide 121 text

#8 DEEP DIVING ON CONCURRENT REACT ALWAYS TRY TO CORRELATE BUSINESS METRICS WITH PERFORMANCE

Slide 122

Slide 122 text

DEEP DIVING ON CONCURRENT REACT

Slide 123

Slide 123 text

DEEP DIVING ON CONCURRENT REACT THAT’S ALL, FOLKS! THANKS! 🤝 QUESTIONS? MATHEUS ALBUQUERQUE ↑ ALL THE LINKS! 🤓