Worker threads for heavy work

The main window should stay responsive. If a task is CPU-heavy or processes large blobs of text or binary data, move it to oro:worker_threads.

This pattern is useful for markdown indexing, import parsing, thumbnailing, encryption, and other work that should not block user interaction.

1) Start a worker from the main window#

src/main.js:

import { Worker } from 'oro:worker_threads'

const worker = new Worker(new URL('./analysis-worker.js', import.meta.url).toString(), {
  workerData: {
    note: 'The runtime should stay responsive even while long text is being processed.',
  },
})

worker.on('message', (message) => {
  console.log('analysis result', message)
})

2) Read workerData inside the worker#

src/analysis-worker.js:

import { parentPort, workerData } from 'oro:worker_threads'

const words = String(workerData.note)
  .split(/\s+/)
  .filter(Boolean)
  .length

parentPort?.postMessage({ words })

This keeps startup simple: pass the initial job in workerData, send later updates with postMessage.

3) Use workers for bounded jobs#

Good worker jobs:

• parse and validate imported data,
• build a search index,
• hash or encrypt large payloads,
• generate derived artifacts like previews or summaries.

Bad worker jobs:

• ordinary button click handlers,
• short UI state updates,
• tasks that mostly wait on network or filesystem I/O.

4) Use environment data for shared process context#

If every worker needs the same process-wide value, use environment data:

import {
  setEnvironmentData,
  getEnvironmentData,
} from 'oro:worker_threads'

setEnvironmentData('apiBaseUrl', 'https://api.example.com')

console.log(getEnvironmentData('apiBaseUrl'))

That is cleaner than hand-threading the same config through every constructor.

5) Shut workers down when the job is done#

Workers are not fire-and-forget:

worker.on('message', (message) => {
  console.log(message)
  worker.terminate()
})

If a job can be reused across windows, use a shared worker instead.

Next#

• Shared workers across windows
• Testing and diagnostics
• oro:worker_threads