⏱ Timeout & Retry Made Fluent

⏱ Timeout & Retry Made Fluent — AsyncWhat.self()

In asynchronous workflows, handling timeouts and retries is often verbose, full of try/catch, timers, and orchestration boilerplate. The self() method of AsyncWhat provides a declarative, fluent abstraction for these concerns, so you can focus on what should happen, not how to wire it up.

---

⏳ Timeout Mode

const fetchWithTimeout = AsyncWhat.as(() => fetch(url).then(res => res.text()))
  .self(2000);  // timeout after 2s

await fetchWithTimeout();

• self(ms) wraps the current async computation in a timeout guard.
• If the operation does not resolve within ms milliseconds, it rejects with TimeoutError.
• Timeouts stay explicit and visible in the chain, without cluttering the business logic.

---

🔁 Retry Mode

Retries are often the messiest part of async logic. self(nAttempts, baseDelay, factor, maxDelay) makes it clean:

const resilientFetch = AsyncWhat.as(() => fetch(url).then(res => res.text()))
  .self(2000)        // timeout after 2s
  .self(3, 100, 2);  // 3 attempts, 100ms base, double each time

• nAttempts → number of times to retry.
• baseDelay → initial delay before retries.
• factor → growth rate for delays (delay = baseDelay * factor^i).
• Parallel retries (default): all attempts run with overlapping time windows; the first success resolves the chain.
• Sequential retries: achieved by setting baseDelay > timeout, so each attempt waits for the previous one.

---

📊 Parallel vs Sequential

Mode	Attempt	Start (ms)	Duration (ms)	Notes
Parallel Retries	1	0	200	Attempts overlap
	2	100	200	First success resolves chain
	3	200	200
Sequential Retries	1	0	200	Each waits for timeout+delay
	2	2500	200
	3	5000	200

---

✅ Benefits

1. Readable — no try/catch pyramids or manual timers.
2. Composable — works seamlessly with .sthen(), .else(), when(), and the rest of the AsyncWhat toolkit.
3. Flexible — parallel fire-and-race or sequential backoff, just by tweaking numbers.

---

By embracing AsyncWhat.self(), you can model robust async flows declaratively. Timeouts, retries, and backoff become part of the fluent chain, keeping code expressive, fault-tolerant, and easy to reason about.

— @fizzwiz ✨